Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Practical Manual
This manual contains the software instructions and exercises of the practical part of the course
'Introduction Geo-Information Science' (GRS-10306).
Although some theoretical background is given in each module, this manual is not meant to be a lecture
textbook, so for explanations about GIS-concepts we refer to the lectures (PowerPoint presentations are
available on the course website) and the lecture book and reader (see below). The practical must be
regarded as an extension of the lectures in which GIS and Remote Sensing theory is clarified by means
of exercises using the software packages ArcGIS and Erdas Imagine.
Theory
For the theoretical part of the course
'Introduction Geo-Information
Science' (GRS-10306) the following
lecture notes are used:
• Introduction to Geographic
Information Systems
K.T. Chang, 2009, 5th edition,
McGraw-Hill.
• Remote Sensing reader
J.G.P.W. Clevers (Ed.).
http://www.geo-informatie.nl/courses/grs10306/
The website contains course information, a course schedule, PowerPoint presentations of the lectures,
datasets for the practicals, examination information and marks and a link to the digital versions of the
modules including the answers to the exercises.
Practical rooms
The practical part of the course is scheduled in the practical rooms PC37 or PC38 on the second floor
of the GAIA building or in several PC-rooms in Forum. Desks and chairs are adjustable to prevent RSI
symptoms. Adjustment instructions for desks and chairs are attached to the doors of the rooms. Food
and drinks are prohibited while working behind the computers.
Computer use
All computers use the operating system Windows XP professional. You can only log in on the computer
if you have a WUR account. You have to work on the same PC during the course, since data is stored
on the hard drive of a particular computer and not on your personal (M) drive. Practical work is done
individually, in couples of two or groups of four participants.
Software
All the software you are working with during this course is already installed on the C-drive of the pc.
You will work with the software packages ArcGIS and Erdas Imagine.
Installing data
Before starting with the first module of the practical, you have to have to install data on your pc. You
can access and download data via the website. The data are packed in ZIP files. Click the link ‘Course
materials’, subsequently click ‘Download practical data’ and following the instructions on the website.
The data are will be automatically stored at the correct location on the D-drive of your pc.
Do not change folder structure and names!!!
The Morning/Afternoon folders contain two folders; one with all ArcGIS related data and one with all
Erdas Imagine related data. In the “ArcMap documents” folder all ArcMap documents are stored. In
each module you will work with one or more documents. The datasets you will work with are stored in
the “Data” folder. In the “Workspace” folder you will store all datasets you will create during the
practical so you will not change the original datasets.
4. MAP PRESENTATION
INTRODUCTION 4-1
5. QUERIES
INTRODUCTION 5-1
6. TRANSFORMATIONS
INTRODUCTION 6-1
7. RASTER OPERATIONS
INTRODUCTION 7-1
8. VECTOR OPERATIONS
INTRODUCTION 8-1
9. SURFACE ANALYSIS
INTRODUCTION 9-1
REFERENCES
Practical Manual
Module 1
‘Getting to know ArcGIS’
1. GETTING TO KNOW ARCGIS
INTRODUCTION 1-1
Introduction
People have used maps for thousands of years to present and analyze geographic information. The
ArcGIS software package is one of the latest extensions to this ancient tradition. The package is
developed by the Environmental Systems Research Institute (ESRI), one of the world’s leading
companies in the field of geographic information system (GIS) technology.
ArcGIS is not a GIS program itself. ESRI describes the software package as an ‘integrated collection of
(more than fifteen) GIS software products with which the user or company can build a complete GIS’.
During this course you will get acquainted with three of these ArcGIS software products. These are the
desktop applications ArcCatalog, ArcToolbox and ArcMap. You can find more information about the
ArcGIS software package on the ESRI website: www.esri.com.
ArcCatalog
The ArcCatalog application helps you to organize and manage all your spatial datasets. It includes tools
that allow you to explore, find and view your geographic information in the form of maps or tables.
You can replace, rename or delete your spatial datasets and view, create and edit the metadata that is
linked to your data.
Warning: NEVER use the Windows Explorer for geodata management, e.g. renaming, replacing or
deleting your geodata. Instead, use ArcCatalog for these purposes.
ArcToolbox
The ArcToolbox application offers you a comprehensive set of GIS tools for geoprocessing and
analyzing your spatial data. Some common geoprocessing tasks include:
• Converting data (such as converting a vector to a raster dataset);
• Overlaying data (by merging or intersecting datasets);
• Extracting data (by clipping a subset of data or selecting data with certain characteristics);
• Finding what's nearby (by buffering data or finding points near other features);
• Managing data (by joining tables, copying datasets, or creating new datasets).
ArcToolbox is not a stand-alone GIS application; it is integrated in both ArcCatalog and ArcMap. You
can only access ArcToolbox via these two programs.
ArcMap
ArcMap is the central GIS desktop application of the ArcGIS software package. ArcMap gives you the
power to access, visualize, edit and process geographic data stored in various data formats. During this
course you will mainly work with data in shapefile format (*.shp), ArcMap’s own data type for vector
datasets.
It is also possible to create your own geographic data in form of maps. Once you have made the map
you want, it is easy to add tabular data, such as dBase files and data from database servers, to your map
so that you can display, query, process and organize your data geographically. In fact, while ArcMap
lets you create great maps to show off your data, you will find the software’s true power lies in how
easily it enables you to solve simple problems by uncovering and analyzing trends and patterns.
In this module:
Exploring datasets and metadata in ArcCatalog.
Working with data frames, layers and tables in ArcMap.
Symbolizing your data: choosing the correct map type.
Getting help from ArcGIS’ extensive help system.
3
4
1
2
quick access to various controls and tools. Note that there are buttons to start ArcMap and
1.
Start ArcCatalog and select the folder that contains your spatial data in the catalog tree
D:\IGI\...*…\ArcGIS\Data (*morning or afternoon). Set the view window setting to Contents.
b. The shapefile is ArcMap’s storage type for vector datasets. In both the catalog tree and the view
window you see an icon in front of each dataset name. There are three different icons related to the
shapefile storage type (check this). What do these icons represent?
INSTRUCTIONS:
2.
Select the ‘Netherlands’ dataset in the catalog tree. Set the view window setting to Preview.
A map of the Netherlands appears showing all features of this dataset. Click the Preview dropdown
arrow and choose Table view.
a. How many records does the table contain? And how many fields?
b. Add a field to the table by using the Options button. Enter capital in the Name field of the window.
Choose Text from the dropdown menu of the Type field. Change the field length to 20.
d. Sort the province names (field: PROVNAME) ascending. How many records belong to the
province of Noord-Brabant?
e. What is the total area of the Netherlands? Hint: use the Statistics function by clicking on the column
heading with the right mouse button .
f. Which province contains a feature that has a perimeter of 219934.719 meters? Hint: use the Find
function by clicking on the Options button. Uncheck search only selected fields. What is the value
of the attribute FID of this feature?
Click ‘View’ in the menu bar and then Toolbar Metadata. The Metadata toolbar will
appear in the application window.
5. Click the Edit Metadata button on the metadata editor toolbar.
6. The information in a metadata document is divided into seven sections (1), displayed across
the top of the metadata editor. When you click a section title in the editor, several tabs appear
that represent the different groups of metadata elements that are defined within the section (2).
1
2
7. You can type or edit information in text boxes that are provided for most elements.
3.
Select the ‘Netherlands’ dataset in the catalog tree. Set the view window setting to Metadata. The
metadata document belonging to this dataset appears in the view window.
a. What kind of information is stored under each of the three tabs of the metadata document?
d. What are the coordinates of the most South-East part of the Netherlands (lower right corner)?
If you browse through the metadata of the ‘Netherlands’ dataset you will notice that the metadata
document is far from complete. You will now use the metadata editor to create some information that is
missing.
a. Click on the Description tab in the metadata document. An abstract for this dataset is still required
(check this).
b. Open the metadata editor. Click on the Identification section and select the General tab. Go to the
Abstract text box and delete its red content. Now you can type your abstract, something like: ‘The
dataset supplies information about the 12 provinces of The Netherlands’. Save your edits.
d. Keywords are very useful when you are searching for a dataset in a database on the web. The
‘Netherlands’ dataset is lacking keywords in the metadata document. Check this. If keywords are
present, they can be found under the Description tab. Use the metadata editor to add some
keywords. Hint: the metadata element Keywords can be found in the Identification section. Save
your edits.
f. Close ArcCatalog.
4
5
1
3
2
1
The properties of a data frame define the context for the data
with which you work; these include the coordinate system,
measurement units, scale, the drawing order of datasets,
and so on. Data frames and the table of contents are the
primary ways to interact with geographic data in ArcMap. Figure 3. Detail of the TOC. The
dataset “Land use” is part of
Links to the data frames and datasets of an ArcMap session data frame “Wag_south1”.
can be conveniently stored in one file called an ArcMap Document (*.mxd). An ArcMap document is
nothing more than a desktop screen of your computer. If you remove, add or change the way the data is
presented (other colors in the legend) you don’t alter or remove data sets.
Note that the datasets in the table of contents are sorted by drawing
order. Datasets at the top of the list are drawn on top of datasets
lower in the list. Figure 4. The ArcToolbox tree
structure.
Toolbars (5)
These bars, located beneath the menu bar in the ArcMap window, contain buttons that give you quick
access to various controls and tools. The tool remains selected until you choose another one. You can
add toolbars via the view pull down menu in the Menu bar
INSTRUCTIONS:
a. Start ArcMap. Click “An existing map”, select “Browse for maps”, press “OK” and browse to
D:\IGI\...*…\ArcGIS\ArcMap documents (*morning or afternoon). Open the ArcMap document
called ‘Intro_AM.mxd’.
Warning: an ArcMap document does NOT store the actual spatial datasets contained in your table of
contents. It only stores references to the location of the datasets contained in an ArcMap document,
(such reference can for example be: D:\IGI\Morning\ArcGIS\Data\Landuse.shp). When you open a
document, ArcMap looks for the datasets on the hard drive of the pc using these references. If it cannot
find the dataset, for example when you have moved, renamed or deleted it with ArcCatalog, the dataset
cannot be drawn in the view window; it has a broken link. You can immediately tell whether a dataset
has a broken data link because it will have a red exclamation mark (!) next to its name in the table of
contents and the check box next to the layer will be unavailable.
If the data is temporarily unavailable, you can ignore the broken link and display the content of the
ArcMap document without the dataset. The dataset will still be part of the map and listed in the table of
contents; it simply will not display. However, broken links can be easily repaired.
INSTRUCTIONS:
1. Locate the layer with the broken link in the table of contents.
2. Right-click the layer and select Data Repair Data Source.
3. Navigate to the source dataset and click Add.
4. Click OK.
5. ArcMap repairs the link to the dataset.
6.
a. Repair the broken link of dataset ‘Soil_types’. The source dataset is stored as ‘Soil_types.shp’ in
folder D:\IGI\...*…\ArcGIS\Data (*morning or afternoon).
Datasets are often referred to as layers in ArcMap. Layers are, in this context, visual representations
of spatial datasets. However, layers do not necessarily refer to spatial datasets. In this practical manual,
the ESRI term ‘layer’ is therefore only used when referred to layer files (see section ‘Working with
layer files’), otherwise the generic term ‘dataset’ is used to refer to digitally stored spatial data.
All datasets present in an ArcMap document are listed in the table of contents, under one or more data
frames. The table of contents also shows the symbols and colors used to draw the features of each
dataset.
Activate a data frame and turn the visibility of a dataset on and off
The order of datasets within the table of contents determines the drawing order; a dataset is drawn on
top of those below them. Thus, you have to put the datasets that form the background of your map, e.g.
the ocean, at the bottom of the table of contents. Note that for a dataset to be visible in the view
INSTRUCTIONS:
1. To activate a data frame, right-click the data frame you wish to activate, and select Activate
from to appearing context menu.
2. A dataset can be turned on and off by clicking the check box.
3. To add a data frame click Insert Data Frame in the menu bar.
4. You can rename the inserted data frame. Right-click the data frame and select Properties. The
Data Frame Properties window opens. Select the General tab and type in the name of the data
frame in the Name text box.
7.
The ArcMap document ‘Intro_AM.mxd’ contains a number of data frames and datasets, which are
listed in the table of contents.
c. Turn the visibility of the datasets contained in the data frame on and off.
e. Insert a new data frame. The inserted data frame is immediately activated. Name this data frame
‘Soil and Land use’.
Adding datasets
INSTRUCTIONS:
8.
Layers store symbology display and labeling. They define how spatial data is drawn in the view
window. For example, a layer might select specific cities from a shapefile, draw them as blue squares,
and label them with text stored in a related table. Layers are not spatial datasets!!
INSTRUCTIONS:
1. Right-click the dataset in the table of contents you want to save as layer file. Click Save as
Layer File from the context menu. Layer files have the ‘.lyr’ extension.
9.
a. Activate the data frame ‘Wag_south1’. Add layer ‘Landuse.lyr’ and dataset ‘Landuse.shp’ to the
data frame.
You can see that the land use features of dataset ‘Landuse.shp’ are displayed in single color. The layer
file ‘Landuse.lyr’ contains an appropriate symbology to display land use types.
b. Open the Layer properties window of the land use layer file. Click the Source tab. What is the data
source of this layer file?
c. What does this mean when you change something in the attribute table of ‘Landuse.lyr’?
To keep your table of contents organized you can group (thematically) related datasets in one data
frame into a group layer. For example, suppose you have two datasets in a data frame, one
representing railroads, the other highways. You might choose to group these datasets and name the
resulting group layer ‘transportation networks’.
Once you have grouped related datasets, you can rename the group layer and subsequently save it as a
layer file, which you can add to other data frames or ArcMap documents.
INSTRUCTIONS:
• You can also create a group layer by right-clicking the data frame. Select New Group
Layer. A new group layer appears in the table of contents. Now you can add layers to
your group layer. If the layer you want to add to a group is already present in the
table of contents, you can drag and drop it in the group layer.
3. You can rename a group layer. Right-click the group layer and select Properties. The Layer
Properties window opens. Select the General tab and type in the name of the group layer in
the Name text box. Note that only the name as it appears in the table of contents is
changed, NOT the name under which the layer is stored.
4. To ungroup the group layer, right-click the layer name and select ungroup from the context
menu.
10.
a. Activate the data frame ‘Wag_south1’. Group the layers ‘Soil_points’ and ‘Soil_types’. Name the
group layer ‘Soil data’ and save the group layer as a layer file in the workspace directory. Name
the layer file ‘soil data’.
b. Create another group layer in data frame ‘Wag_south1’. Use this time the function New Group
Layer. Name the layer ‘Infrastructure’.
c. In the group layer you just created you want to organize datasets that are related to infrastructure. In
this case these are the datasets ‘Roads’ and ‘Trails’. These datasets are already present in your data
frame so you can drag these layers to the ‘infrastructure’ group layer. Save the group layer as a
group layer file in your workspace. Name the layer file ‘Infrastructure’.
d. Drag the group layer ‘Soil data’ to the top of the table of contents.
e. What has happened to the other layers in the data frame? Explain your answer.
f. Turn off the visibility of the group layer ‘Soil data’. What happens with the visibility of the layers
that are contained in this group layer? Explain your answer.
h. Activate the data frame ‘Soils and Land use’ and add the layer file ‘Soil data’.
i. Remove data frame ‘Soils and Land use’ from your ArcMap document.
INSTRUCTIONS:
1. Click the Zoom In button or Zoom Out button on the Tools toolbar.
2. Move the mouse pointer over the map display and click once to zoom around a point.
Alternatively, click and drag a rectangle defining the area on which you want to zoom in or
out. The view redraws to show you the area of the rectangle you defined.
4. Click the Zoom To Full Extent button to zoom to the full spatial extent of all the layers
in your data frame.
5. If you want to zoom out, use the
6. Use the Fixed Zoom In or Fixed Zoom Out buttons for stepwise zooming.
7. Click the Pan button to pan around the dataset.
11.
a. Zoom in to the ponds in the top-left corner of dataset ‘Landuse’ and zoom back to ‘full extension’.
Attributes describe the characteristics of geographical phenomena in a GIS. For example, attributes of a
river might include its name, length and average depth. They are usually stored in a table and are linked
to the spatial features of a dataset by a unique identifier. The attribute table is arranged so that each
row, called RECORD, represents a feature and each column, called FIELD, represents one attribute.
INSTRUCTIONS:
3. For quick access to the attribute values of one particular feature, click the Identify tool
on the Tools toolbar and click the feature in the view window whose attributes you want to
examine.
4. The feature's attribute values are presented in the Identify Results window.
12.
Activate data frame ‘Wag_south1’. Open the attribute table of dataset ‘Roads’.
2. How many attributes does this table contain? Write down the attributes of this table.
d. Use the Identify tool to view the attributes of one of the features of the ‘Soil_types’ dataset. Note
that the attributes are displayed of the Top-most layer, which is the default setting. This means that
when features from different datasets overlap, only the attribute information of the feature that is on
top is displayed. You can change the default setting by clicking the dropdown arrow in Identify
Results window.
Activate the data frame ‘Wag_south2’. Open the attribute table of dataset ‘LU_raster’.
a. How many attributes does this table contain? Write down the attributes of this table.
d. What is the size of one raster cell? Zoom in until you are able to distinguish individual raster cells.
Use the Measure button on the Tools toolbar. Check your result by looking at the properties of
‘LU_raster’ (Right-click the dataset, open the Properties window and select the Source tab).
The basic options of the Symbology editor (Figure 5) are described in this section. Note that this is a
description for symbolizing vector datasets. Raster datasets have limited options for visualization!
There are many ways to display the features of a dataset in a map. The choice of a map type depends
on the nature of the spatial data (e.g. categorical, classes, quantitative) and what you want to show the
user with your map. You can define the map type in the Show field of the symbology editor (box 1,
Figure 5). The most important ones are:
Features
Single symbol All the features in the dataset are drawn with the same color and
symbol. This is useful when you only need to show where a
dataset’s features are located.
Categories
Unique values Each unique value (class) of a dataset’s attribute is displayed with a
different color or symbol. This is the most effective method for
displaying categorical data. For example, a land use map shows
each unique land use type with one a specific color.
Quantities
Graduated color The color of the features changes (e.g. from light red to dark red)
according to the values of a particular attribute. This is useful when
you want to map quantities or for showing data that is ranked.
Graduated symbol The size of the symbol representing a feature is drawn with
changes according to the values of a particular attribute. This map
type is the best way to symbolize data that expresses size or
magnitude. Graduated symbol is only available for point and line
data.
2 3
Figure 5. The symbology editor; where you can change the way your data is displayed.
When one of these map types is chosen, which can be done by clicking on it, different sets of options
become available.
Once you have chosen the attribute that is to be displayed, you can adjust the symbol size range. The
central part of the symbology editor contains three columns: Symbol, Range and Label. The Range
defines the range of attribute values that fall within a symbol class.
Within Quantities ‘Proportional symbols’ is a more advanced symbology.
1. Double-click the dataset for which you want to change the symbology
2. Click on the Symbology tab, click Categories in the Show field and select Unique values.
3. Click Value Field dropdown arrow; choose the attribute that you want to display in the map.
4. Click the Add all values button and uncheck the <all other values> check box.
5. If you want to change the labels that will appear next to the symbols in the table of contents,
type label text into the Label field.
6. You can choose a pre-defined color ramp, or double-click the rectangular colored box in the
Symbol field to set the color of each attribute value manually.
7. Click Apply to redraw the map using your new legend.
14.
Activate the data frame ‘Wag_south1’. Open the Layer Properties window of dataset ‘Soil_types’ and
click the Symbology tab.
b. Display the attribute ‘Soilname’. Which map type did you choose? Explain your answer.
e. In which part of the study area are the highest pH values found?
15.
c. Open the attribute table of dataset ‘Roads’. How many road features does the table contain? When
the answer is different from the answer of 15.b, explain the difference.
Help topics
To browse through the contents of ‘ArcGIS Desktop Help’, go to
the menu bar click Help ArcGIS Desktop Help (or press the
F1 key). Then click the Contents tab. Use the open button to
expand the contents of a book; use the Display button to display a
Help Topic.
To search the Index of ArcGIS Desktop Help, click the Index tab.
Type a keyword in the text box and the help system displays a list
of index entries for that word. Choose the entry you want, and then
click the Display button or double-click the entry to view the Help
Topic for that entry.
• To get help on a command in a context menu (menu launched by right-clicking), highlight the
command and press Shift + F1.
• To get help on a control in a dialog box, click the What's This? button at the top of the
dialog box and click the control. On some dialog boxes there are also About or Help buttons,
both of which provide additional Help information specific to the dialog box.
• To get help on a window, such as the table of contents or the Identify Results window, click
inside the window, then press Shift + F1.
Online help
You can access additional online help resources provided by ESRI: the GIS dictionary and ESRI’s
support center. You can find shortcuts to the sites in the Help pull down menu in the menu bar.
16.
a. Try to find out what the definition of a join is in ArcMap, using the ArcGIS Desktop Help system.
b. Try to find out information about the ArcToolbox button using the What’s This? button.
Practical Manual
Module 2
‘Data storage: digitizing and data structure’
2. DATA STORAGE: DIGITIZING AND DATA STRUCTURE
INTRODUCTION 2-1
Introduction
This module deals with spatial data storage in GIS. During this course you will mainly use the relational
data model as a method of structuring data as collections of tables that are logically associated to each
other by shared attributes.
The first part of this module deals with data storage in vector which is based on tables. You will create
a new vector dataset by digitizing points as a method of (secondary) data capture. Digitizing is the
process of using a mouse to automatically store locations of geographic features by converting their
map positions to series of x, y coordinates in computer files or database with an associated table.
Subsequently, you will view, create and fill tables. In addition you will also join tables.
The second part deals with data storage in raster. You will see that spatial data is structured differently
in a raster environment.
In this module:
Creating a new dataset.
Adding attributes to tables of vector data.
Calculating the area of polygon features.
Joining tables.
Comparison of the vector and raster data structures.
Objectives
After having completed this module you will be capable:
to create a new dataset using ArcMap;
to create and fill a table using ArcMap;
describe the different data types;
to distinguish the difference between discrete and continuous rasters and between zones and regions;
to understand the data structure of vector and rater datasets.
INSTRUCTIONS:
10. Select the Sketch tool . Now it is possible to ‘draw’ point features within the new dataset.
Make sure the check box in front of the dataset in which you are editing is checked.
11. When you are finished, click the Editor dropdown arrow and click Stop Editing.
DON’T FORGET TO SAVE YOUR EDITS!!!
13. Use the Pointer tool to select the point you want to move. To select a point you click on
the point to be moved. Selection handles will appear around the selected features.
14. Move the selected point by dragging it to its new location.
15. When you are finished moving new points, click Stop Editing and save your changes.
1.
In this exercise you are going to create a new point dataset by digitizing the soil point observations as
presented in Figure 3. Activate data frame ‘Wag_south’ and display the datasets ‘Soil_types’ and
‘Roads’.
a. Create a new point dataset. Make sure that the Feature Class Location is your workspace
D:\IGI\...*…\ArcGIS\workspace (*morning or afternoon) and give the new dataset the name
‘Soil_pts’ in the Feature Class name text box.
b. Digitize the locations of the soil point observations as drawn in Figure 3 in order of profile code:
start with point X1 and finish with X9!
c. Open the attribute table of the new point dataset. How many records does the attribute table of
dataset ‘Soil_pts’ contain?
Initially the attribute table will contain three fields, called FID, ID and Shape. The FID field contains
unique identifiers of the features in the vector dataset. The unique identifier links the thematic
(attribute) data to the geometry of a geographic feature. These unique identifiers cannot be changed. In
the ID field a user-defined identifier can be stored. The Shape field stores the feature type of the
geographic feature (points, lines or polygons). This field is maintained by ArcMap and cannot be
edited.
You can add new fields to this table at anytime to store additional attribute data for the point features.
When you create new fields in an attribute table you must select a data type for that field. The data
type determines the kind of data (e.g. text, number) that can be stored in a field. ArcMap supports
several data types, but the most important ones are: float, integer, text, and date.
Important:
• NEVER use blank spaces and symbols in attribute names, only letters, numbers and
underscores. Attribute names can only be 10 characters long.
• Adding and deleting fields should be done using ArcToolbox as described below.
• When adding or deleting fields or rows in a table, make sure that the table itself is NOT
open and NOT in edit mode!
INSTRUCTIONS:
An alternative way of adding and deleting fields is using the ‘Options’ button on the bottom of an
attribute table for adding a field and right click in the field with the field name to delete this field.
INSTRUCTIONS:
a. Write down the meaning of the data types integer, float, text and date. Use the ArcGIS Desktop
Help System to find the definitions. Hint: use the keyword ’'add field’ in the Index field of the help.
b. You are going to add six attributes to the table of the dataset ‘Soil_pts’:
Prof_code, pH, Clay, Silt, Sand and Soilcode. Use the field definitions presented in Table 1.
c. When you have defined the new fields of attribute table ‘Soil_pts’ you can enter the attribute values
according to Figure 4. When you have entered all attribute values stop editing and save your edits.
Then save the ArcMap document.
Prof_code Text 2 - -
pH Float - 3 1
Clay Short - 3 -
Silt Short - 3 -
Sand Short - 3 -
Soilcode Text 8 - -
Important!! When you have saved a field definition (e.g. data type), it is not possible to change this
field definition afterwards!! When you want to change the field definition afterwards, you have to delete
the field and add a new field to the attribute table.
1. Activate the data frame that contains the dataset you want to calculate the area for.
2. Right-click the dataset you want to calculate the area for and click Open Attribute Table.
3. Click on the Options button in the right bottom corner of the attribute table and choose Add
field.
5. Right-click the field heading for this new area field and click Calculate Geometry.
3.
INSTRUCTIONS:
The table has now been created, but it doesn’t appear in the table of contents. To make it visible, select
the Source tab underneath the table of contents. When the table is opened, you can see that it contains
two fields: Rowid, OBJECTID and FIELD1.
1. Adding fields to this table works the same as adding fields to the other datasets. You can add
fields to your table at any time, so you don’t have to add them all at once.
2. You can remove the ‘FIELD1’ field and ‘OBJECTID’ field using the Delete Field option in
the ArcToolbox: Data Management Tools Fields Delete Field. Note that a table must
always contain two fields. So you can only delete attribute “Field1” once you have added a
new field to the table.
3. To add data to the table, click Start Editing and open the table. If you edit a table you should
select the proper ‘source’. In this case the blue highlighted line in the figure below
4. Add data to the table by clicking any blank cell in the table.
5. When you stop editing, save your changes.
You can make further edits to a table at any time to change data values, add or delete fields, and add or
delete records. Make sure that the table is not open and in edit mode when you add or delete fields!!
4.
a. Create a new table called ‘Landscape.dbf’. Save it in your workspace directory. Add two fields to
the new table: LU_CODE and LU_DESCRIP. The field definitions for these fields are presented
in the table below.
b. When you have defined the fields of table ‘Landscape.dbf’, you can enter the attribute values.
Complete the new table according to Figure 6 and save this the edits.
Joining tables
You can add tabular data to an existing dataset by joining it to its attribute table. When you join a
table to a dataset’s attribute table, all fields (attributes) from the join table are appended to the attribute
table of the dataset. You can use any of these fields to symbolize, label, query, or analyze the dataset’s
features.
A join is based on the values of a similar defined attribute that can be found in both tables. The
name of the attribute does not have to be the same in both tables, but the data type and the attribute
values have to be the same (Figure 7).
INSTRUCTIONS:
1. To make join, go to ArcToolbox and click Data Management Tools Joins Add Join.
2. Select the dataset (Layer name) to which the table will be joined and select the field on which
the join is based (Input join field).
5.
a. Join the fields of the table ‘Landscape’ to the attribute table of dataset ‘Soil_types’ by common field
‘LU_CODE’.
A second method to establish a link between attribute tables is with the Relate functionality. The
difference with Join is that one attribute table is not appended to the other. Relate simply defines a
relationship between two tables based on a shared attribute. You can only access related data by
working with the attributes of a dataset. ‘Relate’ is treated in more detail during the following-up course
Geo-information Tools (GRS 20806).
Discrete rasters represent geographic features that have definable boundaries, sometimes referred to as
categorical or discontinuous data. Examples of discrete terrain object are: lake, forested land,
buildings, roads etc.
Continuous rasters represent geographic phenomena that vary spatially without discrete steps. Each
cell value is a measure of the concentration or level of that location. Continuous geographic
phenomena, in general, do not have distinct boundaries like discrete geographic features. A geographic
feature, such as a lake, has a real and definable boundary. However a geographic phenomenon, like lake
depth, continuously changes. Potentially, each cell in a continuous raster can have a different value.
Examples of continuous data include contamination levels, heat from a fire, elevation, or a
concentration diminishing from a source.
Important: Rasters are always rectangular. Every cell location in a raster has a value assigned to it.
When information is insufficient or unavailable for a cell location, the location will be assigned the
value of NoData. NoData and 0 are not the same: 0 is a valid value that can be used in geoprocessing
whereas NoData is excluded from geoprocessing.
The colors are assigned to the raster cells based on the cell value. Each value is symbolized
with one color.
6.
Activate the data frame ‘Vector vs. Raster’ and display the raster dataset ‘LU_raster’.
a. Explain the meaning of the attributes of the dataset ‘LU_raster’ as they are displayed in the Identify
Results dialog.
c. Open the attribute table of ‘LU_raster’ (right click on dataset). How many different values
(i.e. land use classes) does the dataset of ‘LU_raster’ contain?
Note that a region is the raster equivalent of a vector point, line or polygon feature: a discrete object
that represents one geographic feature.
Figure 9. Raster cells belong to zones and regions. This raster contains five zones. The zone with
value 4 is made up of three regions.
7.
a. Do you select a zone or region when you select one record in the attribute table of ‘LU_raster’?
Explain your answer!
8.
a. Compare the attribute tables of the datasets ‘LU_raster’ and ‘Land_use’ (Figure 10) and write down
the main differences in data storage between the two tables.
(1)
(2)
Practical Manual
Module 3
‘Map projections’
3. MAP PROJECTIONS
INTRODUCTION 3-1
From the Earth’s surface to a 3D reference surface 3-1
From a 3D reference surface to a 2D map projection plane 3-1
Introduction
The locations of spatial features on the Earth’s surface are described by a three-dimensional coordinate
reference system. The spherical reference system that is already in use for more than 200 years is known
as the Geographic Reference System that describes locations on the Earth’s surface by latitude and
longitude (Lo and Yeung, 2002).
If you want to produce a map of features on the Earth’s surface, you need to transform the spherical
surface to a flat map. This transformation from three-dimensional surface onto a two-dimensional map
is called projection. Mathematical expressions convert data from the angular, geographical coordinate
system of a sphere to a linear, orthogonal projected coordinate system of a flat map. This transformation
requires several steps.
If you want to establish a 3D reference surface for a particular area (state, country, continent) you have
to adjust the ellipsoid for these discrepancies so that the reference surface closely fits the actual surface
of the Earth. A datum is an ellipsoid that is adjusted so that it matches the actual shape of the Earth of a
particular region as well as possible. The datum serves as the 3D reference surface for calculating the
geographic coordinates of a location.
From the 19th until halfway the 20th century ellipsoids were only fitted to the Earth’s shape over a
particular country or continent (Snyder, 1987). These ellipsoids were determined by ground
measurements. Datums based on these local ellipsoids fit the Earth’s surface only in a particular area.
Examples include the datum ‘Amersfoort’, which closely fits the Earth’s surface in the Netherlands and
datum ‘NAD27’, which fits to the Earth’s surface in the North American continent. These datums are
based on the Bessel 1841 and Clarke 1866 ellipsoids respectively.
Since the space age, satellite-determined ellipsoids have become available that represent the closest fit
of the entire surface of the Earth. The most widely used ellipsoid is WGS84. Because this is already the
best approximation of the shape of the Earth no adjustment based on the difference between the
ellipsoid and actual shape is needed. WGS84 is both ellipsoid and datum. Be aware that although global
ellipsoids give a better overall approximation of the shape of the entire Earth than local ellipsoids, they
do not generally give the best fit for a particular region (Snyder, 1987). This makes the WGS84 datum
less suitable for regional mapping purposes.
Remember that ArcGIS has an extensive help system that contains much information about projections
and coordinate systems, which can help you to clarify concepts discussed in this module if necessary.
In this module:
Defining a projection for an unprojected dataset
Reprojecting a dataset
On-the-fly projection
The effect of different map projections on geometric properties
Objectives
After having completed this module this part you will be capable:
to understand the basic theoretical framework of map projections;
to project and reproject datasets in ArcMap;
to understand the influence of the data frame on map display in ArcMap;
to describe the geometric distortions that are associated with different projections.
Figure 1. Under the Source tab you find information about the map projection of the dataset.
If you define a projection, you can choose between two coordinate systems:
If your dataset has a geographic coordinate system but no project coordinate system, you can still
display it on your map. ArcMap draws the data by simply treating the latitude/longitude coordinates as
planar (x,y) coordinates.
INSTRUCTIONS:
a. Open the Layer properties of dataset ‘Netherlands_rd’. Click the source tab and confirm this dataset
does not have a projection.
b. Project this dataset. Choose the projection ‘Rijksdriehoekstelsel’, the Dutch reference system. This
is a projected reference system. The projection file is located in the ‘National Grids’ folder.
c. Open the Layer properties window of dataset ‘Netherlands_rd’. The source information contains
now a list with projection parameters (Figure 2). Confirm this dataset references locations using a
projected coordinate system. On which geographic coordinate system (datum) is this projected
coordinate system based?
Figure 2. When a dataset has a map projection defined, projection parameters can be found under the
source tab.
INSTRUCTIONS:
2.
In this exercise you will use the Project tool to transform the map projection of dataset
‘Netherlands_rd’ from ‘Rijksdriehoek’ (‘RD’) to ‘UTM’ and ‘Mercator’.
c. Display the three datasets in the view window. Zoom to full extent . Why do the 3 datasets not
overlap? Explain your answer.
d. Explain the large difference in size and area between the Mercator projection and the other two
projections.
e. Which projection gives the best representation of the length, width and area of the Netherlands?
Explain your answer.
When you add a new data frame to an ArcMap document, the coordinate system of the data frame is not
yet defined. The data frame takes automatically the coordinate system of the first dataset added to
the data frame. If you choose to add more datasets after the first, then these datasets are displayed as if
they have the same coordinate system as the data frame. For example, if the first dataset has the
Mercator projection, then all other datasets present in the same data frame, are displayed as if they have
the Mercator projection. Even if this dataset has another map projection defined! This means that
datasets that represent the same area with different projections, still can overlap.
The first dataset added to data frame ‘unprojected’ (exercises 1 and 2) was not projected, which means
that the data frame remains unprojected (data frame coordinate system is set to Unknown) so that
datasets with different projections are displayed at different locations in the view window.
INSTRUCTIONS:
3.
Activate data frame ‘Unprojected’. Zoom to full extent . Open the data frame properties window.
Set the projection of this data frame to ‘Rijksdriehoekstelsel’.
a. Explain what happened to the orientation of the datasets in the view window.
b. Open the data frame properties window. What is the current coordinate system of this data frame?
c. Check if the Display button of the Table of Contents is activated. Drag dataset
‘Netherlands_mercator’ from data frame ‘Unprojected’ to the empty data frame. What is the
coordinate system of the data frame now?
Add datasets ‘World’, ‘Alterra’ and ‘Cities’ to the data frame. Click OK when a Warning appears.
Change the data frame projection to Robinson (Projected coordinate systems \ World \ Robinson).
e. Measure again the distance between Alterra and Buenos Aires. Does the distance differ from the
previous measurement? Explain your answer.
• Conformal projections
Conformal projections preserve local shape. To preserve individual angles describing the spatial
relationships, a conformal projection must show the meridians and parallels intersecting at 90-
degree angles on the map.
• Equivalent projections
Equivalent (equal-area) projections preserve the area of displayed features. To do this, the other
properties—shape, distance and direction—are distorted. In equivalent projections, the meridians
and parallels may not intersect at right angles.
• Equidistant projections
Equidistant maps preserve the distances between certain points. Distance is not maintained
correctly by any projection throughout an entire map. Distance is often only true when measured
parallel to the meridians.
• True-direction projections
True-direction maps preserve direction, which means that direction measurements made on the
ground are the same as direction measurements made on the map. Azimuthal (planar) projections
are always true-direction. Conformal projections, such as the Mercator projection, are also true-
direction.
Figure 5 shows the effect of different projection types on the shape and size of The Netherlands. Table
2 gives projection details, the length, width and area of The Netherlands for each projection. Figure 5a
represents The Netherlands in RD projection. This is a local projection, specially developed to
represent The Netherlands on a flat map with minimal distortion. The RD projection is a planar
projection which preserves shape and direction. Distance and area are distorted. However, because The
Netherlands covers a relatively small area, distortions are very small (e.g. maximum miscalculation of
distance is 10 centimeter per kilometer). We assume that the RD projection gives the most realistic
estimation of the size and shape of The Netherlands. Figures 5b, 5d and 5f represent The Netherlands
using conic projections designed for use on continental (European) scale level. Figures 5c, 5e en 5g
represent The Netherlands using cylindrical projections designed for use on global scale level.
Figure 5c shows a conformal projection (Mercator), designed for use on global scale level. The shape of
the country is preserved. Size distortion is severe compared to the RD projection: the area increase is
268%, the length and width increase with 162%. Figure 5b shows also a conformal projection of The
Netherlands, designed for use on continental scale level. The shape is preserved, length, width and area
are still distorted but the distortions are less severe compared to the Mercator projection.
Figure 5e (equivalent projection on global scale level) shows a deformed country: squeezed in north-
south direction and stretched in East-West direction. The area however, almost equals the RD area.
Figure 5d shows The Netherlands using a continental equivalent projection. The area is almost identical
to the area of Figure 5e, but shape and distances are better preserved.
Figure 5g shows a global, equidistance projection of The Netherlands. The length of the Netherlands is
given accurately. Shape and area are distorted. The country appears squeezed in east-west direction and
the area is 19.5% smaller than the RD area. Figure 5f shows an equidistance projection on continental
scale level.
This example shows map projections are scale level dependent. This means that you have to be careful
when projecting your dataset. If you working on local scale level, then do not choose a map projection
that is designed for use on global scale level. Distortions can be severe. Second choose a map
a.
b. c.
d. e.
f. g.
Table 2. Seven different projections of the Netherlands result in seven different distances and areas.
Figure Projection Projection Projection Preserved Length Width Area
type class Scale property (km) (km) (km2)
True Shape,
5a Azimuthal National 311.6 264.8 34,749
direction direction
Continental
5b Conformal Conic Shape, direction 304.7 273.6 33,810
(Europe)
5c Conformal Cylindrical Global Shape, direction 506.3 431.0 93,357
Continental
5d Equivalent Conic Area 313.2 273.8 34,752
(Europe)
5e Equivalent Cylindrical Global Area 191.3 431.0 34,751
Continental Distance (in N-
5f Equidist. Conic 309.0 273.7 34,281
(Europe) S direction)
Distance (in N-
5g Equidist. Cylindrical Global 311.3 215.5 28,335
S direction)
4.
For this exercise you need data frames ‘Mercator’, ‘Equivalent’, ‘Equidistance’. Each data frame
contains a world map with a different projection, the location of the Alterra building and the locations
of the world’s largest cities. When you retrieve area information make sure you use the F_Area
field!!!
a. Measure the distances between: Alterra – Johannesburg and São Paulo – Jakarta. Retrieve the areas
of the USA, the Dem. Rep. Congo (Zaire) and the (continental part of the )world. Write down your
findings in Table 3.
b. Measure the distances between: Alterra – Johannesburg and São Paulo – Jakarta. Retrieve the areas
of the USA, the Dem. Rep. Congo (Zaire) and the (continental part of the )world. Write down your
findings in Table 3.
c. Measure the distances between: Alterra – Johannesburg and São Paulo – Jakarta. Retrieve the areas
of the USA, the Dem. Rep. Congo (Zaire) and the (continental part of the )world. Write down your
findings in Table 3.
5.
a. The equidistant projection gives the best estimation of the distance between Alterra and
Johannesburg. But the distance between São Paulo – Jakarta is heavily underestimated. The other
two projections give a better estimation. Explain why.
c. Try to explain the difference in distance between measured distance (of the Mercator and equivalent
projections) and true distance of São Paulo – Jakarta.
6.
b. Uncheck dataset ‘World_eqdist’. Click the symbol of dataset ‘World_eqarea’, select Hollow from
the Symbol Selector. Set Outline Width to 2. Zoom in to the equatorial regions (Northern South
America, Central Africa, Indonesia). Write down your findings.
Practical Manual
Module 4
‘Map presentation’
4. MAP PRESENTATION 4-1
INTRODUCTION 4-1
USING THE SYMBOLOGY EDITOR TO SYMBOLIZE DATA 4-2
The symbology editor 4-2
Map types and data scales 4-4
12 STEPS TO PRESENT A MAP 4-5
Preparing the map legend (steps 1 - 2, 4 - 9) 4-7
Laying out and printing maps (steps 3 & 10 - 12) 4-9
Creating a layout 4-9
Adding a data frame to a layout 4-9
Adding a legend to a layout 4-10
Adding a north arrow to a layout 4-10
Adding a scale bar to a layout 4-10
Adding a title to a layout 4-10
Adding text and other graphics to a layout 4-11
Exporting and printing a map 4-11
ADDING GRAPHICS, TEXT AND LABELS TO A MAP 4-13
Adding and modifying graphics 4-13
Adding and modifying text 4-14
Adding and modifying labels 4-14
4. MAP PRESENTATION
Introduction
Maps in ArcMap are based on spatial data. Spatial data refer to information about the locations and
geographic features on the Earth’s surface and the relationships between them, along with attribute
information describing what these features represent. You can communicate complex information more
effectively using maps than tables or lists, because maps take advantage of our natural abilities to
distinguish and interpret colors, patterns and spatial relationships. When you display your data properly
on a map you’ll see spatial distributions, relationships and trends that you couldn’t see before. Your
maps will help you make decisions and solve problems. They also help to communicate your
information and results more effectively to others.
Choosing how to represent your data on a map may well be your most important mapmaking decision.
Symbolizing your data involves choosing appropriate colors and symbols to represent features. This
also involves grouping or classifying features according to their attributes and attribute values.
In this module:
Symbolizing your spatial data.
Adding text and graphics to a map.
Labeling a dataset’s features.
Graphic symbols and attributes (Bertin).
The 12 steps to present a map.
Objectives
After having completed this module you will be capable:
to use the 12 steps to present a map;
to symbolize your data: choosing the correct graphic symbols and attributes;
to create a complete layout of a map.
You use the Symbology tab in the Layer properties window (see Module 1) to control how a dataset is
drawn in the view window. When the symbology editor is opened, you can choose for each dataset a
map type on basis of the type of information your map represents. ArcMap has various map types to
choose from.
In some cases like in Figure 1 the symbology editor gives you the option to label items in the legend. In
this way the values in the legend are displayed different. For example you could rename the soils with
the value “Kalkhoudende ooivaaggrond” for soilname to “Calcareous ooi vague soil” if you want to
present the map to a English speaking audience.
Figure1. The symbology editor; where you can change the way your data is displayed.
Categories:
• Unique values of one field: displays categories/classes using values of one attribute (attribute
= field);
• Unique values of many fields: displays categories/classes using unique attribute combination
values up to 3 attributes;
• Match to symbol in a style: displays categories/classes by matching attribute values to
symbols a symbols list in a style.
Quantities:
• Graduate color: displays quantities using color scale according to an attribute with quantity
values;
• Graduate symbol: displays quantities using a changing symbol size defined by a symbol size
attribute;
• Proportional symbol: displays quantities using a changing symbol size defined by a symbol
size attribute to show exact values;
• Dot density: displays quantities using dot densities according to attribute values.
Charts:
• Pie: displays a pie chart according to multiple attributes for each feature;
• Bar/Column: displays a bar or column chart according to multiple attributes for each feature;
• Stacked: displays a stacked chart according to multiple attributes for each feature.
Multiple attributes:
• Quantity by category: Draw quantities for each category/class of one or more attributes.
1.
Open ArcMap document ‘Map presentation1.mxd’. Activate data frame ‘Wag_south1’ and open the
Symbology editor of dataset ‘Soil_types’:
Note: it is also possible to change the basic colors by clicking on the symbol in the table of contents.
2.
In this exercise you will change the display of a dataset by choosing different map types.
a. Create the following legend for dataset ‘Soil_types’: Map type: Unique value
Value field: Area
Is this map type a logical choice to symbolize the attribute ‘area’? Explain your answer.
b. Change the map type to ‘Graduated color’; choose for Value field ‘Area’. Explain why it is not
possible to represent the attribute ‘Soilcode’ with graduated colors.
d. Type short names for the soil features in the Label field of the Symbology editor. Set the colors for
the different classes individually; choose your own colors and apply the legend to the map.
nominal: Qualitative attribute values are different in nature, without one aspect being more
important than another (e.g. forms of land use).
ordinal: Qualitative attribute values are different from each other, but there is one single
way to order them, as some are more important/intense than others (e.g. importance
of roads).
interval: Quantitative attribute values are different and can be ordered. The distance between
individual measurements can be determined (e.g. Celsius temperature scale).
ratio: Quantitative attribute values are different and can be ordered. Distances between
individual measurements can be determined and these individual measurements can
be related to each other. They also have an absolute zero (e.g. number of visitors).
• size
• (grey) value
• grain / texture
• color hue
• orientation
• shape
Graphic attributes cannot be used randomly. Each one is perceived in a different way by the map user
and they vary in their perception properties. For example, a difference in color, orientation or shape
will be interpreted as a difference in properties or quality. But a difference in size or (grey) value will
be interpreted as a difference in the amount or quantity. The choice of any particular graphic attribute
thus depends on the data scale of the attribute that is to be presented in the map. Figure 4 lists the
graphic attributes that most suitably represent the various perception properties and shows the
relationship between data scale and graphic attributes.
DATA SCALE
During the next exercises you will practice with the 12 steps to present a map approach with an
existing map, made in ArcMap. ArcMap provides two ways to view a map:
• Data view
• Layout view
Each view lets you look at and interact with the map in a specific way.
When you want ’quick look’ your datasets, choose data view. You have been working in data view
until now. Data view is an all-purpose view for exploring, displaying, and querying your data. This
view excludes all the map elements on the layout, such as titles, north arrows, and scale bars, and lets
you focus on your datasets destined for a single map data frame.
When you're preparing your map to hang on the wall, put in a report, or publish on the web, you'll
want to work with the layout view. In layout view, you'll see a virtual page upon which you can
place and arrange map elements. You can do almost everything you can in data view, plus design
your final map.
INSTRUCTIONS:
1. You can switch between data view and layout view in the View menu in the menu bar.
2. Or use the two buttons in the lower left corner of the view window to switch between the two
views.
3.
Open ArcMap document ‘Map presentation2.mxd’. This document contains a soil map of the
Wageningen South area, called Soil map – Wageningen south. Set the view window to layout view.
a. Give an example of the purpose and user group of this soil map.
Based on the purpose and user group of your map, you decide which spatial components (datasets)
your map (and thus your legend) should contain, e.g. roads, land use, soil types, and the hierarchy of
the components in the legend. The legend of a soil map should begin with the soil types and not with
secondary information as build-up areas or roads. The hierarchy in your legend differs from the
hierarchy in your Table of Contents (TOC). The latter is used to draw the features in the correct order.
As an example the soil map might include build-up area. This will probably be drawn on top of your
soil map (highest element in your TOC). In your legend you should mention it as one of the last items.
4.
In this exercise you will practice the “12 steps to present a map” approach with a map made in
ArcMap.
Make sure the view window of ArcMap document ‘Map presentation2.mxd’is set to layout view. If
you want to look at a larger (full screen) version of the map, go the menu bar and select: File Print
Preview.
A soil map of the study area Wageningen South is displayed with legend.
a. Repair the link of the missing information about Clay% (Data Source is file soil_points.shp).
b. Describe in Table 1 the steps 4 to 9 how this map has been created.
c. What is your opinion about the chosen graphic attributes according to figures 3 and 4?
When you have finished this exercise, close the ArcMap document without saving.
A layout lets you assemble the spatial components and map elements you want to represent in your
map, arrange them to get the design you want, and then print it when it is ready. Layouts make it easy
to produce presentation quality maps with ArcMap. You can place more than one data frame onto your
layout along with any of the charts and tables in your ArcMap document. If you change your mind
later, you can add, remove, resize and move each element in a layout as required.
This section now continues with a description how to create a layout and how to add various map
elements to it and is concluded with an exercise in which you will create a layout yourself. The ArcGIS
Desktop Help system contains also an elaborate description of the map layout. Use the keywords
‘layout’, ‘map elements’ or ‘legend’ in the search index.
Creating a layout
INSTRUCTIONS:
ArcMap uses all data frames present in the table of contents in the layout. If you have two data frames
in the table of contents, you will get automatically two data frames in your layout!!
Warning: it is not possible to simply delete a data frame from the layout. When you do this, you also
delete the corresponding dataset from the table of contents, and at the same time from the entire map
document!!! If you do not want to have multiple data frames in your layout, you should rearrange the
datasets in the table of contents into one data frame and then delete the other frames.
It is possible to alter the properties of the frames in layout view: right-click the frame and select
Properties. You can for example change the border of the frame, set a scale or add a reference grid to
your map.
INSTRUCTIONS:
1. In the menu bar select: Insert Legend (this is only possible when you are in layout view).
2. It the appearing window, select the dataset(s) you want to add to the legend. You can also
change the order of the datasets in the legend (think about hierarchy!!) with the up and down
arrows. Click next.
3. In the following windows some options appear to customize your legend. For now, just click
next and finish. Of course, you are free to play around with the options if you wish!
4. The legend is placed in the layout, where it is possible to move it around and change its size.
5. If you are not satisfied with the result you can always modify your legend. Right-click the
legend in the layout and select Properties from the menu that appears. The Legend Properties
window opens in which you can adjust some legend settings.
It is always possible to change the name of one of the layers by clicking on it in the table of contents.
INSTRUCTIONS:
1. In the menu bar select: Insert Legend (this is only possible when you are in layout view).
2. In the Textbox you can define your title
3. Click OK and move the title to an appropriate place within the layout.
INSTRUCTIONS:
1. In the menu bar select: Insert North Arrow (this is only possible when you are in layout
view).
2. In the North Arrow Selector menu you can choose the north arrow style.
3. Click the Properties button and select the North Arrow tab to set the calibration angle. You
can also change the size, style and font of the arrow.
4. Click OK and move the north arrow to an appropriate place within the layout.
Add a scale bar after the datasets and a legend are added to the layout.
Note: Before you add the scale bar to your layout, you have to check if your data frame has a
coordinate system (see Module 3) and if the map units of the data frame are set to meters!!! To check
this, go to the menu bar and select: View Data Frame Properties General.
1. In the menu bar select: Insert Scale Bar (this is only possible when you are in layout
view).
2. Choose the scale bar you want to insert and click OK.
3. The Scale bar is added to the layout where you can change its size and move it. Make sure it is
not too small!
4. Double click on the scale bar, to open the Properties window.
5. In this window the Scale and Units and the Numbers and Marks tabs are important.
6. In the Scale and Units tab it is possible to set the Division value (not to small!), number of
divisions, number of subdivisions and what the scale bar should do when it is being resized
(don’t use the ‘adjust division value’, because the division value should be a ‘round number’,
like 1km, 500 m etc.).
7. Choose the scale bar units.
8. In the Number and Marks tab it is possible to define how often a numerical value appears on
the scale bar, and where this is placed in relation to the scale bar.
9. Click apply and OK.
INSTRUCTIONS:
1. You can use the Text tool to add text for titles and other descriptive text (e.g. the source)
and you can use the drawing tools to add graphics such as boxes, circles and arrows anywhere on
your layout.
2. The drawing tools are in a dropdown list in the Draw tool bar
3. You always can change the properties of a frame: right-click the frame in the layout and select
properties. You can’t change the font size directly. If you wan’t to change it you should enlarge the
text box.
INSTRUCTIONS:
For the inhabitants of the Hinkeloord area a map should be created that gives information about
nature in this area. You are free to choose information out of the available datasets.
a. Have a look at the different datasets and write down the 12 steps to present the map.
1. Purpose
2. User group
3. Title
10 North Arrow
12 Source
b. Create the map using ArcMap according to the “12 steps to present a map” guideline.
c. Before finishing your map / layout perform the exercises 6 – 9. Store the map as a .jpg file with
the name ‘Nature map Hinkeloord’. Discuss your map together with the table above with one of the
supervisors.
1. Open a document, click the New Graphic dropdown arrow in ArcMap’s Draw toolbar
and select the type of graphic.
2. Position the cursor where you want the graphic to start, hold down the left mouse and drag to
where you want the graphic to end, then release the button. For the polygon type, click
once to start, and the click once for each corner. Finish by double-clicking.
3. Adding a line or a curve, click where you want the line to start, click each point along the line
and double-click where you want the line to end.
4. To add a circle, click the circle tool , position the cursor where you want the center of the
circle to be.
5. To change the appearance of a graphic, select a graphic in a view with the Pointer tool .
When a graphic is selected, selection handles appear around it. To change the individual
vertices, click on the edit vertices tool.
6. To change the appearance of the graphic, double-click it in the view window.
7. The Graphic Properties window appears. Click the Symbology tab and choose how you want
the selected graphic to look.
Note: to make a graphic transparent, open the Symbology Selector. Click on Fill Color and choose No
Color.
6.
a. Change the color of one of the soil polygons with the Symbol selector. Create a polygon with a red
outline and a green raster fill.
b. Change the color of one of the soil polygons with the Symbol selector. Create a light gray polygon
without a border (outline).
c. Draw 4 points using the Draw toolbar and display each with a different size and color.
7.
In the map you’ll find a well and some locations with rare vegetation near the river Rhine.
a. Draw a circle around the well. Keep the circle within the borders of the feature. Make this area
recognizable by changing the outline, color, etc.
b. Add an irregularly shaped graphic around the locations with rare vegetation. Make it recognizable
on the map by changing the appearance of this graphic.
8.
a. Add the following text to the graphics in data frame ‘Wag_south2’: ‘Well’, ‘Nature Reserve’.
b. Change the size and style of the text and drag it to another position if it’s necessary.
INSTRUCTIONS:
9.
a. Label the features of dataset ‘Soil_types’ based on the field ‘Soilcode’. Set the label properties to:
Verdana, 18, Bold Italic, red.
b. What will happen to the labels when you turn the ‘Soil_types’ dataset off? Explain your answer.
c. Turn the labels off. You can uncheck the Label features in this layer box or right-click the dataset
in the table of contents and uncheck the Label features option.
d. Add a label to the most northern soil feature using the label function in the draw toolbar.
e. What will happen to this label when you turn the ‘Soil_types’ dataset off? And when you deactivate
the data frame? Explain your answers.
Practical Manual
Module 5
‘Queries’
5. QUERIES
INTRODUCTION 5-1
Introduction
When you want to set up a GIS application to solve a spatial problem, you start with defining a
conceptual model of reality and collecting spatial data. The next logical step involves spatial and
tabular operations using the data you collected. You need to analyze, combine and integrate your
datasets, the so-called data handling, to extract the spatial information you need to solve your spatial
problem. The query (abbreviation of ‘question and reply’) is the first data handling class that will be
discussed during this ArcGIS practical. Queries are processes that extract information from a GIS by
making selections from tabular and spatial datasets. These selections do not change the geometry and
attribute information of the datasets. These processes do not change the thematic and geometric
meaning of the original data. With queries you can select a set out of the whole dataset. You can
discover and analyze new spatial relationships and solve problems when you start asking questions such
as: Where is...? , Where's the closest…? What's inside…? What intersects…?
ArcMap provides a number of tools to help you find answers to these types of questions. Queries can be
divided into tabular and spatial dataset queries.
The structure of a query depends on the query language used. ArcGIS uses SQL (Standard Query
Language). The basic syntax of SQL is:
Select <features/records>
From <dataset>
Where <condition> ;
Example of a tabular dataset query: ‘Where are the soil points located with pH > 6?’
Select Point features
From Soil_points
Where pH-value > 6
Example of a spatial dataset query: ‘How many soil points are located within 500 m1 from the river
Rhine?’
Select Point features
From Soil_points
Where within a distance of 500 m1 from the river Rhine
In this module:
Querying a vector dataset:
• tabular dataset queries: selecting features by using a query expression;
• selecting features based on spatial relationships;
• saving selected features into a new vector dataset.
Querying a raster dataset:
• selecting raster cells by using a query expression;
• selecting raster cells with ArcMap’s ‘Con’ tool.
Objectives
After having completed this module you will be capable:
to describe the difference between tabular dataset and spatial dataset queries;
to perform various queries with ArcMap tools;
to create new datasets containing selected features or raster cells.
When you select features, they are highlighted. Features remain highlighted until you make a different
selection, or when you deselect them. When you open the attribute table you will see that the feature
table records for the features you selected are also highlighted.
INSTRUCTIONS:
To find out how many features fulfill the conditions in your expression, or to review the attributes of the
selected features, open the attribute table. The number of records that are selected is displayed at the
bottom of the attribute table!!
a. Select all features of dataset ‘Soil_types’, which have an Area that is larger than 150000 m2, by
using the query builder. Make sure that the method used is: Create a new Selection. How many
records are selected?
b. Select all features of dataset ‘Soil_types’ which have an Area that is larger than 150000 m2 AND
which are classified as: ‘kalkhoudende poldervaaggrond’. How many records are selected? Clear
the selected set of features so none are selected.
c. Write the query of exercise 1b down according to the formal structure of a SQL query.
Select:
From:
Where:
Sorting attributes
By sorting an attribute table you can list the features in order of importance.
INSTRUCTIONS:
1. In the table, right-click the header of the field you wish to sort the attribute values for.
2. Select the Sort Ascending button to sort the attribute values, smallest values first or in
alphabetical order.
3. Select the Sort Descending button to sort the attribute values, highest values first or in
opposed alphabetical order.
2.
a. Use the Sort Descending button to order the attribute values of field ‘ polygon_co’ of the dataset
‘Soil_types’, the highest value first, the lowest last.
Refining a query
Until now we have seen how to create a new selection set of features based on a certain query condition.
Once you have selected a set from a dataset, you can refine the selection. For example, you can reduce
the set or expand the number of features. ArcMap offers three possibilities to refine your selection.
INSTRUCTIONS:
1. To refine a query click the Method dropdown arrow in the Select by Attributes window and
select one of the options.
2. Proceed with the query as usual.
3.
Select features from dataset ‘Soil_points’ that have a clay % of more than 10%.
b. The clay % of point X9 appears to contain a measurement error. Remove that point feature from the
selection.
2. Click the Select Feature tool and click the feature(s) you want to select. Hold SHIFT to
select multiple features or to drag a box around them. Features that fall partly or wholly
inside the box are selected. The outline of the selected features turns light blue to indicate they
are selected.
3. Open the attribute table.
4. Click the Show: Selected button to display only the records of the
selected features.
2. Open the attribute table, click the Show: Selected button to display the
records of the selected features. Click the All button to display all the records of the attribute
table again.
4.
a. Display only dataset ‘Soil_types’. Select the soil feature with soil name “Kalkloze
poldervaaggrond”. Open the attribute table and display only the records of the selected features.
b. Write this query down according to the formal structure format of a SQL query.
S(elect):
F(rom):
W(here):
c. Remove the most southern soil feature from the selection and add the most northern one to the
selection. Afterwards display the selected records in the table.
For example, suppose you have two datasets displayed in the view window: ‘soil types’ and ‘land use’,
with the ‘land use’ dataset on top of the ‘soil types’ dataset. If you select a land use feature with the
Select Feature tool , you will also select the underlying soil feature at that location. This can be
inconvenient.
To overcome this inconvenience you can specify the datasets you want to select from with the
‘selectable layers’ setting.
INSTRUCTIONS:
• Click the Selection tab at the bottom of the table of contents. Now the table of contents lists all
layers in your data frame.
• You can only select features, using the cursor or a graphic, from datasets that are checked.
• Uncheck the datasets you do not want to be able to select from. Now these datasets can be
drawn in the view window but you do not select features from these datasets when you use the
Select Feature or Select by Graphic tools.
• A dataset turns bold in this list when a set of features is currently selected from it. The number
of selected features is also shown between brackets.
5.
a. Display the datasets ‘Land_use’ and ‘Soil_types’. Select the mathematics building (Figure 2) with
Mahematics Building
Now you see that besides the mathematics building also the soil feature in which the building is located
is selected (turn off and on the Land_use dataset).
b. Click the Selection tab at the bottom of the table of contents. The datasets in the data frame appears
are listed here. You can see that there is one feature selected in both datasets layers. Uncheck the
box of the dataset ‘Soil_types’ and click the Display tab at the bottom of the view window.
c. Select the mathematics building again with the Select Feature tool. Now you see that only the
mathematics building is selected in the Land_use dataset and not the soil feature on which this
building is located.
INSTRUCTIONS:
1. Make sure that the Draw toolbar is activated. Select from the menu bar: View Toolbars
Draw.
2. Click the New Graphic dropdown arrow and click a graphic tool, such as the circle
tool .
3. Move the cursor to a location where you want the circle to be centered, hold down the left
mouse button and drag out to define the circle. You can set the diameter of the circle by right-
clicking the graphic. Select Properties from the pull-
down menu and click the Size and Position tab. Here
you can adjust the with and height of the circle. With
anchor point (see Figure 3) you can define the fixed
point of the circle. By default this is the lower left
corner, but since you want to keep the circle in the
center you should use the middle as the anchor point
Figure 3. Choosing Anchor Point.
4. Optionally, click the Fill Color dropdown arrow and click No Color to make the
graphic. When the features are selected, you can see the selected features under the graphic.
5. To select features with the graphic go to the menu bar and click: Selection Select by
Graphics.
6. The features that fall within or intersecting the graphic will be highlighted to show they are
selected.
7. Now that you’ve found the features, you can look at their attributes by opening the attribute
table.
8. You can use the Pointer tool to move the drawn shaped around the map or resize them by
dragging one of the handles. To delete a graphic, select it and choose Delete from the Edit
menu in the menu bar (or press ‘Delete’).
6.
a. Draw two circles centering on soil points X1 and X2, with a radius of 400 m. What are the soil
codes of the selected soil features?
This query method selects features based on their spatial relationships to one or more source features.
Features to be selected may be in the same dataset as the source features or they are in different datasets
(Chang, 2006).
The Select By Location functionality can be used to select features (target features) that are within a
specified distance to user-defined source features in the same or a different dataset.
Examples include finding a forest (target feature) adjacent to a city border (source feature) and finding
a pub (target feature) within 200 meters from a train station (source feature). When you use the Select
by Location functionality, select first the source features with a tabular data query before you
start to use the tool.
INSTRUCTIONS:
5. Specify the distance within which to search for features. For example when you set the distance
to 200 meters, all features will be selected that are located within 200 meters from the source
feature(s). If you want to select features that are adjacent to other features like in the example
above, set the distance to 0!
6. Click Apply.
Note: If you would follow these instructions you would end up with a selection of land use features
adjacent to a certain forest feature AND the forest feature itself! The Select by Location dialog
also allows the user to refine a query. Click the ‘I want to:’ dropdown arrow and choose one
of the options offered.
a. Select the features from the dataset ‘Soil_types’ adjacent to features with soil name ‘kalkloze
poldervaaggrond’. What will be the source feature of this selection? How many features are
selected?
b. You used the spatial relationship “Are within a distance of” but you can also use another spatial
relationship to make same selection. Which one? Check it by comparing both results.
c. Select the features from the dataset ‘Land_use’ that are within 250 meters from the soil feature with
soil name ‘kalkloze poldervaaggrond’. How many features of the land_use dataset are selected?
INSTRUCTIONS:
8.
a. Write down the codes of the soil points (dataset ‘Soil_points’) that fall completely within the soil
features classified as ‘kalkhoudende ooivaaggrond’.
A spatial join is a ‘join’ the attribute tables of two datasets using the ‘Shape’ fields in these tables to
join them (Figure 4). A spatial join creates a new dataset, which contains both sets of attributes (like a
normal join) and the new attribute ‘distance’.
This is can be the distance from a line feature to the nearest point feature or vice versa.
Spatial join
Figure 5. Spatial join based on the shape-field of both tables. In the new table a ‘distance’ field is
added in which the distance from each polylines to its nearest point are given.
INSTRUCTIONS:
The procedure, you have to follow to find the nearest line feature to a point feature using spatial join, is
the following:
1. Right-click the dataset to which you want to join attributes, click Joins and Relates Join.
2. Select Join data from another layer based on spatial location from the first dropdown list.
3. Click the layer dropdown arrow and click the name of the dataset whose attributes you want to
join.
4. Click Each line will be given all the attributes of the point that is closest to it, and a distance
field showing how close that line is.
5. Define the name and output location of the created new dataset. Make sure you save the
dataset in your Workspace.
6. Click OK.
7. A new dataset is added to the map.
Note: Whenever a new dataset is added, do not forget to save the ArcMap document!
9.
a. What is the code of the nearest soil point to the road named ‘Veerweg’? To answer this question
you will have to create a spatial join between datasets ‘roads’ and ‘Soil_points’.
b. Write down the distance between the ‘Veerweg’ and its nearest soil point.
INSTRUCTIONS:
10.
INSTRUCTIONS:
Unfortunately you cannot export the selection to a new raster dataset. With the Select by Attributes
functionality you can only visualize your selection. To create a new dataset from a subset you have to
use the CON tool, which is located in the Spatial Analyst toolbox.
1. Activate the extension ‘Spatial Analyst’ by choosing Extensions of the Tools menu.
2. Check the extension ‘Spatial Analyst’. This extension allows you to use the Spatial Analyst
tools for geoprocessing of raster datasets!
If the evaluated cell values of the input raster are TRUE, then user-defined value(s) are assigned to
these cells in the output raster. User-defined values can in this case be the original cell value, a constant
or the value from another raster dataset.
The cell values that are evaluated as FALSE can be set to a different set of user-defined values. For
example, if the cell value in the input raster is greater than 10, then return 1, otherwise return 100.
A selection is where the geometry and attribute information do not change. This is the case where the
original cell values are saved in the output raster. If other values are saved in the output raster, then not
only a selection takes place, but also an operation because the thematic (attribute) data of the raster
changes.
A great advantage of the use of the Con tool is that you can retain the original cell value in the
output raster! In module 7 you will see that you can also make selections with the Raster Calculator
functionality but then you will loose the original values. A disadvantage of the Con tool is that you can
only use it for discrete rasters. For raster with a floating data type you can not use the con tool.
11.
a. Select all cells for which ‘value = 5’. How many cells are selected?
b. Clear the selection. Select only those land use classes which cell count is higher than 4000. Which
classes are selected?
c. Create a new raster dataset that contains all raster cells from ‘LU_raster’ classified as forest
(value=3). Retain the original cell value. Save the new dataset in your workspace under a logical
name.
d. Create a new raster dataset that contains all raster cells from ‘LU_raster’ classified as buildings
(value=1) and roads (value=2). Assign a constant value to the output raster cells. Save the new
dataset in your workspace under a logical name
Practical Manual
Module 6
‘Transformations’
6. TRANSFORMATIONS 6-1
INTRODUCTION 6-1
PART 1: GEOREFERENCING AN IMAGE 6-2
Overall procedure .................................................................................................6-2
Georeferencing in ArcMap...................................................................................6-4
Geometric transformation and resampling ....………………………………...6-4
Image registration: obtaining control points .....................................................6-7
Validation .............................................................................................................6-9
PART 2: DATASET STRUCTURE TRANSFORMATION 6-10
Transforming datasets from vector to raster.......................................................6-10
Weight tables to determine the cell value after transformation ......................6-10
Vector-raster transformation in ArcMap.........................................................6-13
Transforming datasets from raster to vector.......................................................6-15
Raster-vector transformation in ArcMap ........................................................6-15
Changing the cell size of a raster dataset............................................................6-16
6. TRANSFORMATIONS
Introduction
The second data handling class comprises Transformations. Transformations can be divided into three
groups:
• Projection transformation: the mathematical conversion of a map from one projected
coordinate system to another.
• Georeferencing: the geometric transformation from digitizer units or image coordinates to a
projected coordinate system using a set of control points.
• Data structure transformation: conversion from one data structure into another, for example
vector to raster.
The first transformation group was discussed in module 3 ‘Map projections’: you have projected and
reprojected vector datasets. In this module the other two groups of transformations are treated.
Part 1 of this module discusses georeferencing. You will perform an image-to-map transformation,
which means that you reference image coordinates to map projected coordinates. The second part deals
with data structure transformation: vector data to raster data or raster data to vector data
conversions.
In this module:
Image registration and geometric transformation of an image file.
Resampling raster values.
Assessing positional accuracy by RMS error: validation of the georeferenced raster.
Vector-raster transformations.
Raster-vector transformations
Objectives
After having completed this module you will be capable:
to interpret and explain the meaning of the terms: image registration, image rectification,
georeferencing, (ground) control points, validation, spatial resolution and spatial accuracy within
the context of geometric transformation of raster data;
to enumerate which actions and data are required to carry out an image rectification;
to argue your choices for methods of image rectification and resampling method;
to reason the spatial accuracy obtained.
to define decision-rules that determine which attribute is stored after data structure transformation;
to perform data structure transformations in ArcMap using ArcToolbox.
However, this part of the ArcGIS practical concerns the recurring problem where (1) coordinates of a
newly digitized map are represented by digitizer units from a point of reference on a sheet of paper or
(2) coordinates of an image are represented by column and row pixel indices. You have to
georeference the digitized map or image to a map projected coordinate system in order to use this data
in GIS analysis (Figure 1).
Figure 1. An image of the ‘Droevendaal Experimental Farm’ (left) has to be georeferenced in order to
align it with a topographical map in a GIS.
Overall procedure
The georeferencing of raster data is often referred to as image rectification and can be divided into
three steps:
• Image registration
Image registration refers to the process of the identification of corresponding points in an input
image and a reference dataset in a known map projection. These points are called (ground)
control points.
• Geometric transformation
The control points are used to determine the coefficients for two (polynomial) transformation
functions that describes the relationship between image coordinates and map coordinates
(Figure 2).
x = f1 ( X , Y ) y = f2 ( X ,Y )
x = b01 + b1 X + b2 Y
Affine transformation
y = b02 + b1 X + b22Y
with b01…b22 = linear transformation coefficients
(x,y) = input image coordinates (column, row)
(X,Y) = map coordinates
Figure 3. Transformation equation coefficients are determined on basis of the control points.
While you might think each cell in an image is transformed to its new location in a projected raster
dataset, the process actually works in reverse. During georeferencing, a matrix of "empty" cells is
computed using the map coordinates. Then, each empty cell of the output raster is given the value
of the corresponding cell (determined by the transformation function) or cells in the input image,
based on a process called resampling.
• Resampling
The two most common resampling techniques are nearest neighbor assignment and bilinear
interpolation. These techniques assign a value to each empty cell by examining the cell values
in the unreferenced raster dataset (the input image). A third common resampling technique is
cubic convolution. This technique and the effects of the different techniques on the output
raster are discussed in more detail in the course Remote Sensing (GRS-20306).
Transformation functions
Reference Image Output raster
(a)
+ + + + +
+ + + + +
+ + + + +
(b) Input raster
1.
Open ArcMap document ‘Transformations part1.mxd’. Activate data frame ‘Scanned image’. This data
frame contains an image of the ‘Droevendaal Experimental Farm’.
a. What is a nominal data scale? How do you know that the values of raster layer ‘Scanned map’ are
on nominal scale?
b. What is the unit of the coordinates of the scanned image coordinates (meters or pixels)? How many
rows and columns does the image have?
c. Add vector layer ‘Top10vct.lyr’ to the data frame. This layer can be found in the data folder.
INSTRUCTIONS:
1. Open the Georeferencing toolbar. Click View in the menu bar, point to Toolbars and click
Georeferencing.
1 2
2. In the table of contents, right-click the referenced dataset (in this case the ‘Top10vct’) and
click Zoom to Layer.
3. Click the Georeferencing dropdown arrow in the Georeferencing toolbar, click Fit to display.
Make sure the layer box is set to “scanmap”.
4. Now both layers are displayed in the view window.
Figure 6. Collecting control points. A detail of the image and reference datasets showing a
control point that links an identical locations in image (top) and reference (bottom) datasets.
a. What are control points and where do you need them for?
Start collecting control points. Answer exercises c and d when you have collected 2 and 4 points,
respectively. Collect 12 control points in total. Note that during the collection, the scanned map is
adjusted (based on the transformation function) to match the reference dataset after each collected
point. Make sure your points are well distributed among the map.
c. When you have collected 2 points, open the Link Table. Are there already any residuals
calculated. Why (not)?
d. Open the table again when you have collected 4 points. What is the Total RMS error (Figure 7)?
e. What Total RMS error did you achieve after the collection of the 12 control points? A Total RMS
error between 2 and 4 is acceptable. Save the set of control points (open the Link Table and click
Save). Place the file in your workspace!
f. Replace control points with high residuals if necessary. What happens with the Total RMS error if
such control points are removed?
g. What does a low RMS value indicate; does it imply an accurate registration and hence good
rectification results? Why (not)?
Now you will perform the geometric transformation and resampling of the input image. In ArcGIS
these two steps are integrated into one operation!
INSTRUCTIONS:
1. First select the transformation function. Click the Georeferencing dropdown arrow in the
toolbar, point to Transformation and select a transformation polynomial.
2. Click the Georeferencing dropdown arrow and click Rectify. A dialog box opens (Figure 8).
3. Set the Cell Size for the output raster dataset.
4. Select a Resample Type.
5. And specify the name and location of the output raster.
6. Set format to GRID
7. Click Save.
3.
Rectify the input image of the experimental farm. Use a 1st order polynomial (Affine) transformation
function. Use both the Nearest Neighbor and the Bilinear Interpolation resampling types. Name the
output rasters ‘farm_1st_nn’ and ‘farm_1st_bl, respectively. Save the output rasters in your
workspace. Change the cell size to 2 meters and set the format to GRID.
a. What is the meaning of the Nearest Neighbor and Bilinear resampling type options?
Activate data frame ‘Referenced image’ and add both output rasters to it. Change the symbology of the
two raster datasets. Open the Symbology editor and import farm.lyr
a. What pattern in the displacements between the rectified images and the ‘top10vct’ layer can be
observed? Hint: zoom in.
5.
Activate data frame ‘Scanned image’. Repeat the rectification using a 2nd order transformation function.
Name the output raster ‘farm_2nd’. Choose the appropriate resampling type.
a. Open the Link Table. What Total RMS error have you achieved now?
b. Did the spatial match between the rectified image and the reference layer ‘top10vct’ improve?
c. How does this result relate to your answer to exercise 5a? Is there any reason to use a 2nd order
(quadratic) transformation function?
6.
1. Open the Excel workbook ‘Validate.xls’. This file can be found in folder
D:\IGI\...*…\ArcGIS\data\georeferencing (*morning or afternoon).
2. Measure the 15 coordinates of the northeast corner of the validation plots in the map “farm_2nd“.
3. Fill in the coordinates in the columns X_image and Y_image of the Excel sheet.
4. The RMS E(rror) of the residuals is automatically computed.
5. Repeat the measurements and calculations with the raster dataset ‘Farm_1st_nn’.
a. What overall RMS error did you achieve for the 1st and 2nd order polynominal transformations
(give unit)?
b. How does this compare to your answer(s) to exercises 5a and if applicable 2d?
7.
a. Draw a situation showing that it is necessary to make decisions about the value labeling of raster
cells, when transforming: 1. point features to raster cells;
2. polygon features to raster cells.
Sketch 1:
Sketch 2:
In most GIS systems, the user is able to influence the outcome of a decision in case of ambiguity
(multiple vector features in one raster cell). A common used approach to define a decision rule is by
means of a weight table. These weights are used to resolve cases when a single cell contains more than
one vector element. In this case the vector elements with the highest value in the weight value will be
assigned to this cell.
As a remark, in this example it is important to know which the wells with toxic drinking water are. For
further analysis, the point elements have to be converted to raster cells.
b. Write down the relation between the above mentioned descriptions about wells by giving it a value
and a weight. The weight value has to ensure that the worst water quality is represented in a raster
cell after transformation.
c. During the vector-raster transformation, the coordinates which describe the position of a geometric
element disappear. Describe in your own words what the reason is that this happens. Use your
lecture book to support your answer.
d. In the figure below, a raster is draped on the points of the wells described earlier. Fill in the empty
raster at the right with the code number of the well which is assigned to the raster cell according to
your weight table given.
Figure 10. Wells described in a vector structure as point elements, with a raster along those points
(left) and an empty raster (right).
In the next example line objects have to be transformed into a raster environment (Figure 11). The line
objects represent different types of roads. The road with the highest traffic capacity is the most
important for further analysis, so it has the highest weight value.
a. Fill in the empty raster presented below with codes for the different road types according to their
traffic intensity.
Figure 11. Roads described in a vector structure as line elements, with a raster along those lines (left)
and an empty raster (right).
In the next example polygon objects have to be transformed into a raster environment (Figure 12). More
than one polygon object is able to share the area of a raster cell. Many GIS systems determine which
partial area is the largest in the cell. The value of this area object will be stored in the raster cell. This
means that very small polygons, which are situated completely within a raster cell, are lost. These areas
could be very important to the user.
Once again a weight table is used as a rule of decision to make sure that this information is not lost. In
the example presented below, heather areas play the most important role for future analysis.
a. Fill the empty raster with a code number, make sure that all heather areas maintain in the final result.
Figure 12. Land use described in a vector structure as polygon elements, with a raster draped on the
polygons and an empty raster.
All GIS programs have default decision rules settings. The most common rules are now discussed.
When you convert polygons to raster cells, cells are given the value of the polygon with the largest area
within the raster cell. During transformations of line features, cells are given the value of the line feature
that is found within each cell. When more than one line feature intersects a raster cell, the first line
feature value that is encountered during processing is given to the cell. Cells that are not intersected by
a line feature are given the value of NoData. When you convert point features, cells are given the value
of the point that is found within each cell. If more than point is found in a cell, then the cell is given the
value of the point it first encounters when processing. Cells that do not contain a point feature are given
the value of NoData.
It is important to realize that all raster cells get a value after the transformation from vector. This can
either be a value based on the vector feature or the NoData value if the raster cell does not intersect
with a vector feature. There will be no empty cells in a raster.
1. Activate the data frame that contains the vector dataset you want to convert to raster
2. Open ArcToolbox; select Conversion Tools to Raster Feature to Raster
3. In the dialog box, define the input dataset, the field which will be used to assign the values to
the raster, the output dataset (select a proper location) and the cell size. Click OK.
11.
Activate the data frame ‘Wag_south’ of ArcMap document ‘Transformations part2.mxd’. Convert the
features of dataset ‘Soil_types’ into raster cells. The cell size of the new raster dataset ‘Soilraster1’ has
to be 100 m. For cell values choose field ‘Soilcode’. Store dataset ‘Soilraster1’ in your workspace
directory.
a. Overlay ‘Soil_types’ with ‘Soilraster1’. Explain why some parts of the polygon features are not
given a cell value based on the soil code (e.g. along the borders of the river Rhine).
b. Which value has the highest number of cells? What is the total number of raster cells?
c. Select the Identify tool and click in the Rhine. Are these cells as empty as they appear in the
map? Or do they also have a value?
d. Give the cells with the value of NoData a color. Open the Layer properties (double-click the layer).
In the lower right corner of the Symbology tab there is the Display NoData as dropdown list. Click
on the arrow and select a color. Explain what happens.
f. The total number of raster cells you calculated earlier, is the number of raster cells that have a value
based on the attribute ‘Soil code’. Calculate the total number of raster cells again, taking into
account the cells with value NoData. Hint: you can find the number of rows and columns of the
raster in the Layer properties. Open the layer properties window and select the Source tab.
Convert the features of dataset ‘Soil_types’ to raster cells again, but now the cell size of the new raster
dataset ‘Soilraster2’ has to be 10 m.
g. Which value has the highest number of cells? What is its total number of raster cells (including the
NoData cells)?
h. Which of the two vector-raster transformations gives the best results? Explain your answer.
12.
Determine the individual features in Figure 13. Give each new feature a new identifier number in the
empty raster.
Figure 13. Land use described in a raster structure (left) and an empty raster (right).
INSTRUCTIONS:
1. Activate the data frame that contains the raster dataset you want to convert to vector.
2. Open ArcToolbox; select Conversion Tools From Raster Raster to polygon.
3. In the dialog box, define the input dataset, the field which will be used to assign the values to
the vector dataset, the output dataset (select a proper location). Click OK.
Convert raster dataset ‘Soilraster1’ into the polygon feature dataset ‘Soilvector1’. Dataset
‘Soilvector1’ has to be stored in your workspace directory. Choose the field ‘Soilcode’ to be used in the
transformation.
b. What are the differences between the attribute table of dataset ‘Soilvector1’ and the attribute
table of the original dataset ‘Soil_types’?
c. Where are the biggest discrepancies located in comparison to the original dataset ‘Soil_types’?
14.
Convert raster dataset ‘Soilraster2’ into the polygon feature dataset ‘Soilvector2’. Dataset
‘Soilvector2’ has to be stored in your workspace directory.
a. What are the major differences between datasets ‘Soilvector1’ and ‘Soilvector2’, and between
datasets ‘Soilvector2’ and ‘Soil_types’? Explain these differences.
b. Where are the biggest discrepancies between datasets ‘Soilvector2’ and ‘Soil_types’?
a
1 1 1 2 2 2
1 2 1 1 1 2 2 2
1 1 1 2 2 2
2 2 2 1 1 1
2 1 2 2 2 1 1 1
2 2 2 1 1 1
b
1 2 2
1 1 2 ?
1 2 1
1 ? 2
1 2
? ? ?
2 1
2 ? 1
Figure 14. Different resampling options. (a) going to a smaller cell size, without changing the
geometric meaning, (b) going to a larger cell size, (c) going to a smaller cell size and changing the
geometric meaning.
15.
a. Convert soilraster1 to a dataset soilraster3 this time with a cell size of 50m with the nearest
neighbor resampling technique. Is the new raster dataset geometrically different and more accurate?
b. Convert soilraster1 to a dataset soilraster4, this time with a cell size of 30m with the nearest
neighbor resampling technique. Is the new raster dataset geometrically different and more accurate?
c. Should you use the bilinear resampling technique for transforming the soilraster dataset? Explain
your answer.
d. Suppose you have to compare to land use rasters, one from 1950 (with a cell size of 50m) and one
from 2000 (30m) which cell size and resampling technique will you use for your analysis?
Practical Manual
Module 7
‘Raster operations’
7. RASTER OPERATIONS
INTRODUCTION 7-1
Introduction
The first two data handling classes, queries and transformations, were discussed in modules 5 and 6.
This module and the next two modules deal with the third data handling class: operations. This module
focuses on operations in raster environment, so called cell-based analysis. Raster operations are
powerful tools in spatial modeling.
ArcGIS offers the user a comprehensive toolset for cell-based GIS operations: ArcGIS Spatial
Analyst. These operations can be divided into five types: local, focal, zonal and global operations and
operations that perform a specific application (e.g. hydrologic runoff analysis functions and cost or
resistance functions). The specific application operation type will not be discussed in this module.
Cell-based operations use Map Algebra (Tomlin, 1990). Map Algebra is a computational language that
models the surface of the earth as a multitude of independent, coincident dataset layers using operators
or functions to one or more (coinciding) raster dataset layers (Bruns and Egenhofer, 1997); see
examples below. The operations are performed on individual cell value(s) of the input layer(s)
(Heywood et al., 2002). Map Algebra output is always a new raster layer.
Note that in order to use Map Algebra with more than one input raster dataset, the geometry (cellsize,
extent, orientation) of the input raster datasets the geometry must match.
In this module the following GIS tools and operations are used:
The Spatial Analyst Toolbox
Local operations with the raster calculator and the combine function
Focal operations: focal (neighborhood) statistics
Zonal operations: zonal statistics and tabulate areas
Global operations: Euclidean Distance (Buffers)
Objectives
After having completed this module this part you will be capable:
to give definitions of the different groups of operations;
to describe the possibilities of each group of operations;
to perform various operations with ArcGIS using the Spatial Analyst Toolbox.
We will discuss four types of local operations in this module. There are more operations available in
ArcGIS. You can find them in the Local toolset of the Spatial Analyst Tools in ArcToolbox. Feel free
to explore them yourself if you desire. It is adivised to explore ArcGIS Desktop Help to gain more
information about these operations.
• if a cell in an input raster dataset has the value of NODATA, the value of the cell at the
same location in the output raster is always NODATA (Figures 2-6)!!
• the cell values of the input raster dataset(s) are not added to the ouput table as
attributes, except when using the ‘combine’ function.
You can imagine that this is not always desireable. More background information about how the value
of NoData is treated during operations is given in the course Geo-information Tools (GRS 20806).
1. The first type of local operations calculate output values for each location as a function of the cell
values in a single raster dataset, using mathematical functions.
For example, the output value is the base e of the input value (figure 2).
Mathematical functions (f) have the syntax: output raster = f(input). They are applied to the
values in a single input raster. There are four groups of mathematical functions:
• Logarithmic;
• Arithmetic (add, substract, divide, multiply);
• Trigonometric (sin, cos, tan);
• Powers.
These functions can be carried out with the following ArcGIS tools:
• Spatial Analyst toolbar: Raster Calculator.
• Spatial Analyst toolbox: Math toolset.
Figure 2.The logarithmic function ‘Exp’ (base e). Syntax: ouput raster = Exp(Inlayer1).
Arithmetic operators. Arithmetic operators allow for the addition, subtraction, multiplication, and
division of two rasters, numbers, or a combination of the two.
Figure 3. Using arithmetic operators to calculate the cell average of two raster datasets. Syntax:
ouput raster dataset = (Indataset1 + Indataset 2) / 2.
Logical operators. Logical operators evaluate the values of an input rasters using a conditional
statement. Logical operators can be Boolean (AND, OR, etc.) or Relational (<, >, <>, etc.). One
conditional statement can contain one or multiple Boolean and relational operators. The outcome
of a logical operation is always TRUE or FALSE.
• Spatial Analyst toolbox: Math Logical. The logical toolset contains tools for
Boolean end relational evaluation. The logical tools use one operator and have the
syntax: output raster = (input1 ‘logical opr’ input2).
Figure 4. The ‘Boolean AND’ tool finds cells that have non-zero values in both input raster
datasets. Syntax: ouput raster dataset = (Indataset1 & Indataset 2).
• Spatial Analyst toolbar: Raster Calculator. With the raster calculator you can combine
multiple Boolean and relational operators within one conditional statement. For example
‘output raster dataset = (soil type = 1 AND elevation >=40)’.
4. The fourth type of local operations is the raster overlay. In an overlay the cell values of two or
more input raster datasets are combined into one output raster dataset using the combine
function of Spatial Analyst.
The Combine function assigns a new unique value to each unique combination of values at each
location. The original Value items, or the alternative field values if specified, are added to the
output rasters' attribute table: one for each input raster (Figure 6).
Figure 6. Overlay of two input raster datasets. Syntax: outgrid = combine(ingrid1, ingrid2, ingrid3).
First activate the Spatial Analyst extension. Click Tools in the menu bar, click Extensions and check
Spatial Analyst. This extension allows you to use the Spatial Analyst tools for geoprocessing of raster
datasets!
a. Are local operations geometric (spatial) or thematic (non-spatial) operations? Explain your answer.
b. What geometric criteria should be fulfilled in order to execute local operations that involve more
than one dataset?
Warning: the output raster dataset of the Raster Calculator is a TEMPORARY raster dataset, which is
stored in the ‘TEMP’ folder on your harddisk and NOT in your workspace. This raster dataset will be
lost when you restart your ArcMap document. If you are satisfied with the output result of your
operation, you can make the raster dataset permanent. You can do this by right-clicking the raster
dataset and selecting Data Make Permanent. Give the raster dataset an appropriate name and save
the file in your workspace folder. Make sure the type is set to ESRI GRID.
INSTRUCTIONS:
1. Open the Spatial Analyst toolbar: in the menu bar select: View Toolbars Spatial
Analyst.
2. Click the Spatial Analyst dropdown arrow, click Raster Calculator.
3. The raster calculator dialog box aids in the creation of an expression that produces a new
output raster dataset. The expression can be based on a single raster dataset or multiple raster
datasets.
4. There are seven sections in the Raster Calculator dialog box (Figure 7).
1. The layers box lists the available raster datasets in the active data frame;
2. Arithmetic operators;
3. A keypad of numbers;
4. Relational operators;
5. Boolean operators;
6. Mathematical functions;
7. The expression box.
5. Type your expression in the expression box or create your expression by clicking the operator
buttons. You can insert raster datasets in your expression by double-clicking a raster dataset in
the layers list.
6. To access recently entered expressions, right-click in the expression box and click Recent
Expressions. Then copy and paste the expression into the expression box.
7. Click Evaluate.
1 2 3 4 5 6
2.
Activate data frame ‘Local’. The dataset ‘dem’ (digital elevation model) contains elevation values,
expressed in meters. Create a new raster where elevation is expressed in centimeters.
a. Write down the created expression which you formulated in the Raster Calculator dialog box.
b. What is the name of this new raster dataset? Make the raster dataset permanent. Save it in your
workspace folder and change the name to something more appropriate.
3.
You can also select raster cells with the Raster Calculator using relational and Boolean operators in the
expression. Note: when you press the‘=’ button in the raster calculator, a double equal to ‘= =’ appears
in the expression. This is normal raster calculator syntax.
a. Create an expression in the raster calculator with which you select buildings (Lu_raster value 1)
located on sites that have an elevation of more than 30 meters. Write down the expression.
b. What is the meaning of the values ‘0’ and ‘1’ in the output raster dataset?
e. You can consider the expression you used as a query: you made a selection. However, there is a
difference between this selection and the queries you made in module 5. What is this difference?
Raster overlay
The output rasters you created during the previous exercise were binary: a raster cell did or did not
fulfill the condition you stated in the expression. The attribute table of the output raster contains only
the values ‘0’ and ‘1’. The original values of the soil code or land use code were lost which can be
inconvenient if you want to do further analysis.
A raster overlay by using the combine function avoids this problem. The original values of the input
rasters are kept in the attribute table of the output raster.
INSTRUCTIONS:
Note: the values in the output attribute table are integer. Values that contain decimals (such as elevation
values) will be rounded!!
4.
Overlay the rasters ‘soil_raster’, ‘lu_raster’ and ‘dem’ using the combine function.
a. Select from the attribute table pasture areas (value 5) on soils with soil code ‘Rn95C’ (value 3).
What is the elevation range of the selected raster cells?
b. What part (in %) of the soils with soil code Rn95C is covered by pasture?
Figure 8. Focal operations affect the source cell and a user-defined neighborhood.
Two important types of focal operations are focal statistics and focal filters. The focal filters are beyond
the scope of this course. This subject is treated in the course Remote Sensing (GRS 20306). However,
feel free to consult ArcGIS Desktop Help for more information about filters. In this module we will
focus on the focal stastics, also referred to as neighborhood statistics.
How does a focal operator work? We know that focal operators use the value of a source cell and the
values of the cells in a user-defined neighborhood around the source cell to calculate a new cell value
for the output raster cell at the same location as the source cell.
Suppose we define a rectangular neighborhood focus of 5x5 raster cells (Figure 9) and a focal function,
for example the neighborhood statistic ‘focal sum’. You can consider the neighborhood focus as a
window that moves over your raster dataset layer. In this example the moving window is located at the
top left corner of the raster dataset layer (Neighborhood A, figure 9).
Figure 10 illustrates the calculation of a neighborhood statistic ‘focal sum’. The upper part of Figure 10
shows the function for a 3x3 rectangular neighborhood. The processing cell has value 2. The same cell
in the output raster will get the value of the sum of the processing cell value plus the values of the
neighboring cells; in this case the sum equals 21. So a value of 21 is assigned to the cell in the output
raster at the same location as the processing cell in the input raster. The lower part of Figure 10 shows
the result for an entire raster dataset.
Figure 10. Illustration of the calculation of the neighborhood statistic ‘focal sum’.
If a cell with the value of NoData is present in the neighborhood, it will be ignored in the processing,
on contrary with NoData cells in local operations. However, if the entire neighborhood consists of cells
of NoData, the output cell value will be NoData!!
It is beyond the scope of this course to explain why ArcMap treats the value of NoData different during
focal operations than during local operations. The course Geo-information Tools (GRS 20806) will
elaborate on the treatment of NoData values during processing.
1. Open ArcToolbox. Click the Spatial Analyst Tools Neighborhood Focal Statistics.
2. Choose the Input raster dataset you want to calculate a statistic for.
5.
a. Calculate for raster dataset ‘dem’ the neighborhood statistic ‘mean’. Use the following setting:
b. Choose the same symbology for the output raster as the elevation raster. Explain the effects of this
‘focal mean’ operation.
c. Calculate the neighborhood statistic ‘maximum’ for raster dataset ‘lu_raster’. Use the same
neighborhood settings described above.
d. Do the calculated values have any meaning? Use the word ‘data scale’ or ‘measurement scale’ in
your answer.
ArcMap offers a variety of zonal operations which can be found in Spatial Analyst toolbox (Zonal
toolset) of ArcToolbox. In this module we will discuss two zonal operations: Zonal Statistics and Zonal
Area.
Zonal statistics
The Zonal Statistics function calculates statistics for each zone (see Module 2 for a definition fo
‘Zone’) defined by a zone dataset, based on values from another dataset (the “value” raster) found
within a zone (Figure 11). This could be average population density per zone of pollution or most
common vegetation type per zone of elevation. The output of this operation can be presented in a (dbf)
table or as raster dataset. During this course only tabular output is created. The zone dataset can be
vector or raster data. The value raster must be a raster dataset.
With an integer input value raster the statistics are: area, minimum, maximum, range, mean, standard
deviation, sum, variety, majority, minority, and median. If the input values are of floating-point data
type, the zonal calculations for majority, median, minority and variety are not available.
4 2 1 3 3 3
Zonal mean
5 5 3 4 4 4
2 5 5 4 4 3
a b c
Figure 11. The ‘zone dataset’ (a) defines zones, in this case there are two zones. Zones can for example
be land use types. The ‘value raster’ (b) contains the input values used in calculating the output for
each zone. These can for example be elevation values. The zonal function ‘mean’ would return output
(c). The average elevation value is calculated for each zone (check this). All cells in the output raster
that belong to the same zone receive the same value, in this case the average elevation within each
zone. Note that in this example raster output is given. In the following exercises tabular output is
created.
INSTRUCTIONS:
1. Open ArcToolbox. Click Spatial Analyst Tools Zonal Zonal Statistics as Table.
2. Click the Input raster or feature zone data dropdown arrow and choose the vector or raster
dataset that defines the zones. Statistics will be calculated for each zone of this raster dataset,
based on the information in the ‘value raster’ (step 4).
3. Click the Zone field dropdown arrow and choose the field containing the attribute values
which define the zones.
4. Click the Value Raster dropdown arrow and choose the input raster. This raster supplies the
values that are used to calculate statistics for each zone of the zone dataset (step 2).
5. Specify the name and location (your workspace) of the output table.
6. Click OK.
7. To view the table, click the Source tab at the bottom of the table of contents.
Activate data frame ‘Zonal’. Calculate elevation statistics for every soil zone (soil type). Use
SOILCODE as Zone field. Chart the mean statistic.
c. What data scale or measurement scale should data have to calculate statistics?
Calculate landuse statistics for every soil zone. Use again SOILCODE as Zone field.
d. What does a record in the statistic table represent?
e. Which of the attributes of the statistic table gives relevant information? Explain your answer.
Zonal geometry
The Zonal Geometry tool calculates several geometric properties of zones, including the area. The
output is in form of a (dbf) table.
INSTRUCTIONS:
1. Open ArcToolbox. Click Spatial Analyst Tools Zonal Zonal Geometry as Table.
2. Click the Input raster or feature zone data dropdown arrow and choose the dataset that
contains the zones you want to calculate the areas for.
3. Click the Zone field dropdown arrow and choose the field that contains the values of the
zones.
4. Specify name and location of the output table.
5. Specify the processing cell size (optional) and click OK.
6. Pay attention to data management. The output tables often provide more information than you
are interested in. Delete the fields that contain information that is not directly relevant for your
application (see Module 2 how to delete fields)!
7. You can join the attribute table that contains the area values to the original raster dataset.
7.
Calculate the areas of the land use zones (‘Lu_raster’) by using ‘Zonal Geometry as Table’ and join the
output table to original raster dataset.
According map algebra, Euclidean distance is calculated from the center of the source cells to the
center of each of the surrounding cells (Figure 12) that contains the value of NoData. Each cell in the
output raster contains a distance value which is the distance to its nearest source cell. Source cells are
given the value of zero. Be aware that the output raster is a continuous dataset with floating-point
distance values.
Figure 12. Calculating the Euclidean distance from a raster cell to its nearest source cell.
In this module we focus on the use the Euclidean distance function for proximity analysis. Proximity
analysis is used to determine the proximity of spatial features within a dataset layer or between two
dataset layers. A widely applied proximity function in GIS analysis is the buffer. Buffers can be applied
to both raster and vector data. The output dataset that contains the buffer zone can also be raster or
vector.
Note that if you calculate the Euclidean distance to vector features, a transformation must take place
because the Euclidean distance function can only be applied to raster cells. However, you do NOT have
to do this yourself. The transformation is done automatically by ArcMap during the process of
Euclidean distance calculation!!!
INSTRUCTIONS:
1. Display the dataset that contains the source cells or source features.
2. Select the zone(s) or feature(s) of interest. If you do not select particular feature or zone, all
will be used for as sources for Euclidean distance calculation.
8.
Create a new dataset ‘Dist_roads’ that contains Euclidean distance values to all road features. Do not
specify a maximum distance and choose 10 m as output cell size.
a. How many distance classes are distinguished in the new raster dataset ‘Dist_roads’? The number of
classes is determined by an ArcMap default setting.
Change the symbology with the Symbology editor. Choose map type ‘Stretched’ and a color ramp to
visualize the Euclidean distance raster.
b. Does the Euclidean distance raster contain discrete or continuous data? Explain your answer.
Use the Raster Calculator to create a buffer zone of 100 meters around the roads.
Note that the output extent and mask settings have effect on all raster operations. However, they are
frequently used during buffer and reclass operations. We will therefore discuss these settings in this
context. During the course Geo-information Tools (GRS 20806) the mask and output extent are
discussed in more detail.
1. Click the Tools in the menu bar. Click Options Geoprocessing Environments
General settings.
2. Click the Output Extent dropdown arrow and select an extent.
Note that this is an ArcMap setting. It does not change the extent of datasets that have already been
created. Only datasets that are created after you changed the output extent setting, will have the
specified extent.
9.
a. Create a new dataset ‘Dist_extent’ that contains Euclidean distance values to all road features. Do
not specify a maximum distance and choose 10 m as output cell size.
b. Give the NoData values of dataset ‘lu_raster’ a color. You can see that the extent of the distance
raster coincides with the extent of dataset‘lu_raster’.
You created a distance raster that has the same extent as dataset ‘Lu_raster’. However, the land use
dataset has an irregular shape. Parts of the distance raster fall outside the borders of the ‘Lu_raster’
dataset. If you want your distance raster to have a specific spatial area, for example, the same area as
the ‘Lu_raster’ dataset, you have to set the Mask. The mask identifies those locations within the output
extent that will be included during geoprocessing. This can be an area, like in this case, but it can also
be a selection of cells or features. Cells that fall outside the mask will be assigned the value of
NoData.
INSTRUCTIONS (Mask):
1. Click the Tools in the menu bar. Click Options Geoprocessing Environments
Raster analysis settings.
2. Click the Mask dropdown arrow and select a dataset that will be used as mask.
3. You can also set the output cell size. This cell size is automatically chosen for all raster output
datasets.
10.
a. Create a new dataset ‘Dist_mask’ that contains Euclidean distance values to all road features. Do
not specify a maximum distance and choose 10 m as output cell size.
b. Display the ‘Dist_mask’ dataset in the view window and compare it with the dataset ‘Dist_extent’ to
see the result of a mask.
There are several approaches to reclassify your data; by individual values, by ranges, by intervals or
area, or through an alternative value.
You can reclassify the continuous Euclidean distance raster to a discrete raster.
INSTRUCTIONS:
Be aware that you lose the original distance values during reclassification. They are replaced by a
new value.
11.
a. Reclassify raster dataset ‘Dist_mask’ into distance classes of 50 meters. Name the dataset
‘Recl_dist’ and save the dataset in your workspace.
Practical Manual
Module 8
‘Vector operations’
8. VECTOR OPERATIONS
INTRODUCTION 8-1
Introduction
In the previous module, operations with raster data were discussed. Although vector operations are not
divided into local, focal, zonal and global operations, there are vector counterparts of the raster local,
focal and zonal operations. Focal operations do not exist in vector environment. Raster operations are
cell-based; the raster cell is the smallest geometric object to which operations are applied. The smallest
geometric object to which vector operations are applied is a point, line or polygon feature. Vector
operation can therefore be considered feature-based.
All ArcGIS tools for raster operations are conveniently stored in the Spatial Analyst toolbox. The
vector counterpart of this toolbox is the Analysis toolbox. The Analysis toolbox provides a powerful set
of tools to perform various geoprocessing operations for all types of vector data. With these tools, you
can perform overlays, create buffers, calculate statistics, perform proximity analysis, and much more.
When you need to solve a spatial or statistical problem, you should always look in the Analysis
toolbox.
As with raster datasets, vector datasets should always be projected in the same reference system in
order to apply geoprocessing when multiple datasets are involved!!
In this module:
Recalculating attribute values with the field calculator.
Vector overlays: the union, intersect and identity functions.
Deriving attribute statistics.
Vector buffers.
Objectives
After having completed this module you will be capable:
to calculate new attribute values with the field calculator and field statistics;
to describe and apply the most important vector overlay types;
to create vector buffer zones around features.
Note that field calculations and vector overlays (next section) are vector counterpart of the local raster
operations. Local operations are performed on single raster cells, the smallest geometric object. Field
calculations and overlays are performed on single features, the smallest geometric object in a vector
environment.
INSTRUCTIONS:
Note: Adding and deleting fields should be done using ArcToolbox, as described in module 2.
Whenever problems (errors) occur, close ArcMap and re-open it.
1.
Open ArcMap document ‘Vector operations.mxd’. Activate data frame ‘Wag_south’. In the attribute
table of dataset ‘Soil_points’, the values of attribute Clay% and Silt% are stored. During the texture
analysis of the soil samples something went wrong. The clay percentage of all the soil points is 2 % too
low, and the silt percentage is 2 % too high.
a. Correct these clay and silt fractions for this soil sample stored in the attribute table ‘Soil_points’.
Suppose you want to analyze the Sand% / Clay% - ratio (SC-ratio) of the soil samples.
c. Describe how to implement this in ArcMap and write down the soil sample with the highest and the
soil sample with the lowest ratio-value.
(hint: use field setting: Data type float, Precision: 4, Scale: 2).
So, “an overlay can be considered as the process of stacking digital representations of various spatial
datasets on top of each other so that each position in the area covered can be analyzed in terms of these
datasets”(Burrough, 1986). From the references of Heywood and Burrough two types of overlays are
distinguished:
1. Optical overlays;
2. Topological overlays.
In optical overlays, spatial datasets are visually superimposed; there is NO integration of geometry and
attribute data involved. Thus, strictly speaking, optical overlays are not operations. Optical overlays
can be used to query; selections based on spatial relationships (this was discussed in module 5).
In a topological overlay, the geometry and attribute data of two datasets are integrated. Where lines or
polygon borders cross each other, new intersection points are calculated. By definition, the topological
overlay always creates a new dataset.
Figure 2 shows the results of a topological overlay of two thematic datasets. In this example, the output
dataset contains 8 polygon features. The number of polygons of the final map is not only dependent on
the numbers of polygons in initial datasets, but also dependent on the form of the boundaries. The more
complicated the boundaries in the source datasets, the more polygons features in the output datasets.
Note that this is the vector counterpart of the raster combine function.
Topological overlays
With topological overlays you integrate geometry and attribute data spatial datasets. In this way you
create new spatial data that gives you new information. ArcGIS offers a variety of overlay tools (the
Analysis toolset in ArcToolbox). We will discuss three overlays. Consult the ArcGIS Desktop Help
system for more background information about the different types of overlays.
Intersect
The Intersect function computes a geometric intersection of the input features and a tabular join of the
attribute tables. Features or portions of features which overlap in all dataset layers will be written to the
output dataset (Figure 3). Note that this is a different intersect than the optical intersect (Select by
Location) you used in exercise 3.
Union
Union is a topological overlay of two or more polygon spatial datasets that preserves the features that
fall within the spatial extent of either input dataset; that is, all features from both datasets are retained
and extracted into a new polygon dataset and a tabular join of the attribute tables is computed (Figure
4).
Identity
Identity is a topological overlay that computes the geometric intersection of two datasets. The output
dataset preserves all the features of the first dataset plus those portions of the second dataset that
overlap the first and a tabular join of the attribute tables is computed. For example, a road passing
through two counties would be split into two arc features, each with the attributes of the road and the
county it passes through (Figure 5).
a. Figure 6 illustrates an intersect overlay of two datasets. Write down the attribute table of the output
dataset that results from this overlay.
E
A C
B
F
INSTRUCTIONS:
The union and identity overlays can also be found in the Overlay toolset of the Analysis toolbox.
Intersect dataset ‘Land_use’ with the soil feature(s) classified as ‘gY30’. Save the output dataset in
your workspace.
b. The attribute table of the output dataset contains two 'area' fields. Explain the meaning of both
fields.
c. Recalculate the area of the features of the intersect output dataset. What is the total area of the of
the output dataset?
d. Calculate the total residential (C2=1) area within the intersected area.
Statistics function
When exploring the contents of a dataset you can derive statistics describing the attribute values in
numeric fields. You'll see how many values the column has, as well as the sum, minimum, mean,
maximum, and standard deviation of those values. A histogram is also provided showing how the
field's values are distributed. Statistics are calculated for all numeric fields in the table. Be aware that
only statistics calculated for ratio and interval (quantitative) data are meaningful!!
INSTRUCTIONS:
1. Open the attribute table of the dataset from which you have selected features.
2. Right-click the header of the field you wish to calculate statistics for.
3. Click Statistics. The window that appears shows the sum, mean, maximum, minimum, range,
variance and standard deviation of the attribute which you have chosen (or which you have
selected from the drop down menu). The count shows how many features are currently
selected. You can only calculate statistics of fields that contain numeric data.
4. To see a description of another field's values, click the Field dropdown arrow and select
another field.
5. If you want to calculate statistics for a subset of your data, make a selection first!!
4.
a. Write down the values of statistics Sum, Count, Mean, Maximum, Minimum and Standard
Deviation of the Area field.
Summarize function
Sometimes the attribute information you have about map features is not organized the way you want—
for instance, you have population data by municipality when you want it by province. By summarizing
the data in a table, you can derive various summary statistics—including the count, average, minimum,
and maximum value—and get exactly the information you want.
You can do this using the Summarize function. ArcMap creates a new table containing one record for
each unique value (class) of the selected field (e.g. ‘C2’), along with statistics summarizing any of the
other fields (e.g. ‘Area’) in the attribute table. You can then join this table of summary statistics to the
Note that this has some similarities to the zonal statistics function (module 7). With zonal statistics you
could calculate for example the average elevation of each soil type. This function used two raster
datasets as input: one dataset contained the zones (soil_types); the other datasets contained the values
from which the statistics were derived (elevation). The summarize function does basically the same.
Instead of two datasets, summarize uses two fields of an attribute table as input. One field functions as
‘zone’, the values in the other field are summarized for each ‘zone’. In tabular statistics the ‘zone’
attribute is often referred to as ‘case field’ or ‘grouping attribute’. In ArcGIS it is referred to as ‘case
field’.
INSTRUCTIONS:
1. Open the attribute table of the dataset for which you want to summarize statistics.
2. Right-click the heading of the field you want to summarize and click Summarize (this function
can also be found in ArcToolbox: Analysis Tools Statistics Summary statistics).
3. In the dialog box that appears:
• select the field to summarize. Summary statistics are calculated for each unique attribute
value that this field contains (for example each land use ‘zone’ of field ‘C2’);
• choose one or more summary statistics for the fields you want to include in the output table;
• specify the name and location of the output table.
4. It is possible to only use the selected features, by checking the box Summarize on the selected
records only.
5. To view the table, click the Source tab at the bottom of the table of contents.
6. Click OK.
7. Click Yes.
5.
Summarize the statistics minimum, maximum, sum and average of the field area for the seven land use
types (zones) that are present in the Wageningen South study area.
b. Write this statistical query down according to the formal structure format of a query. So, select
attributes S from database F, which fulfill the condition W (see module 5).
S:
F:
W:
c. Join the summary statistics table to the ‘Land_use’ attribute table. Where is the largest building
located? If you don’t remember how to join tables, consult module 2 again.
d. Select buildings that are larger than the average? Write down the number of selected polygons.
e. What data scale must the ‘zone’ field have? And the field that contains the values that are used to
calculate the statistics?
INSTRUCTIONS: (buffer)
1. Make sure the map units are defined in your dataframe and/or dataset.
2. Select the features you want to buffer.
3. Open ArcToolbox, click Analysis Tools Proximity Buffer.
4. Specify the input features from the dropdown list or browse to the dataset.
5. Define the output feature class (save the dataset in your workspace).
6. Fill in the buffer distance (or choose a field which provides buffer distances).
7. Optional you can choose to dissolve the buffers (Dissolve: all), the result will then be one
single feature instead of one buffer feature for every input feature.
8. To calculate the area of a vector buffer: see module 2 (page 2-6).
6.
Create a buffer zone of 500 meters around the Gen. Foulkesweg. Use a query to make sure you select
all features with street name ‘Gen. Foulkesweg’. Set Dissolve to all to dissolve the borders between the
buffers around the different features.
b. Is the river ‘Rhine’ located within the zone of 500 m around the Gen. Foulkesweg?
7.
Create a vector buffer zone of 400 meters, divided into 4 sub zones of 100 m each, around soil point
X7.
a. How many other soil points are located within 400 m of this soil point?
b. Calculate the areas of the different sub zones of this multiple ring buffer.
Step 1 Calculate the Euclidean distance from soil profile point X7.
Set Maximum distance to 400, Output cell size 5 meters.
Set your extent to Lu_raster if you don’t get a full circular distance raster
Step 2 Reclassify the output distance raster to 4 buffer zones of 100 meters.
c. Compare the result with the multiple ring vector buffers you just created. What is noticeable?
d. Calculate the areas of the four raster buffer zones (see module 7 how to do this). Join the output
table to the attribute table of the buffer dataset. Compare these areas with the areas of the vector
buffer rings.
INSTRUCTIONS:
1. Select the features which define your clip feature
2. Open ArcToolbox, click Analysis Tools Extract Clip
3. In the next dialog screen select your input
4. Select your clip features
5. In the box “output feature class” you can specify the name and location of the output.
6. Press OK
8.
Activate data frame ‘Create subset’. This data frame contains the ‘Land_use’ dataset and a dataset with
the postal code zones of Wageningen.
a. Select the feature with postal code ‘6703’. Take a subset from the ‘Land_use’ dataset that contains
the area that has postal code 6703.
Consult the ArcGIS Desktop Help to learn more about these tools.
Practical Manual
Module 9
‘Surface analysis’
9. SURFACE ANALYSIS
INTRODUCTION 9-1
Introduction
This last module of the ArcGIS part of the course ‘Introduction to Geo-information Science’ focuses on
operations involving surfaces.
A surface can be defined as a geographic phenomenon represented as a set of continuous data, for
example rainfall, temperature, ice thickness of a glacier or organic matter content of a soil. A
continually varying surface can be represented by isolines (contour lines), and these contours can be
effectively regarded as sets of closed and nested polygons. Although contours are very suitable
presentation form of a continually varying surface, they are not particularly suitable for numerical
analysis (or modelling). So other data formats were developed in order to be able to represent and to
use effectively in spatial analysis where a continuous phenomenon is involved.
A surface model is an approximation of a surface. Surface models are stored and displayed as rasters
or TINs (Triangular Irregular Networks; vector format). Because a surface contains an infinite number
of points, it is impossible to measure and record the value at every point. A surface model approximates
a surface by taking a sample of the values at different points on the surface and interpolating the values
between these points.
A widely used surface model in GIS is the Digital Terrain Model (DTM). This is a digital
representation of the continuous variation of topography over space. A well known example of a DTM
is the Digital Elevation Model (DEM): a digital model of a topographic surface using information on
elevation of the land’s surface. Other examples of DTMs are models of slope or aspect. Note that these
terrain models are derived from a DEM.
Digital elevation models have many uses. Among the most important are the following:
1. For hydrological analysis.
2. Three-dimensional display of landforms for landscape design and planning.
3. For planning routes of roads, location of dams, etc.
4. For statistical analysis and comparison of different kinds of terrain.
5. For computing slope maps, aspect maps, and slope profiles that can be used to prepare shaded
relief maps, assist geomorphological studies, or estimate erosion and run-off.
6. Provide data for image simulation models of landscapes and landscape processes.
ArcMap’s Spatial Analyst offers the user a wide range of functions to analyze surfaces in the Surface
toolset.
In this module:
Interpolation of height points.
Deriving slope and aspect from a DEM.
Objectives
After having completed this module you will be capable:
to understand different interpolation methods;
to apply an interpolation method to derive a elevation surface from observation points;
to derive other terrain models from a DEM.
Spatial interpolation
Visiting every location in a study area to measure a phenomenon is usually nearly impossible and very
expensive. Instead, you can measure the phenomenon at strategically dispersed sample locations, and
predicted values can be assigned to all other locations by means of an interpolation technique (Figure
1). Input points can be either randomly or regularly spaced or based on some sampling scheme.
Spatial interpolation functions create a continuous (or prediction) surface from observation point
values (Figure 1). They make predictions from observed values for all locations in a raster dataset,
whether a measurement has been taken at the location or not
Figure 1. Observation points (dark dots) with interpolated surface DEM and from the DEM computed
contour lines.
Why interpolation? The assumption that makes interpolation a viable option is that spatially distributed
objects are spatially correlated; in other words, things that are close together tend to have similar
characteristics: the values of points close to sampled points are more likely to be similar than those that
are farther apart. This is the basis of spatial interpolation.
There are a variety of ways to derive a prediction for each location; each method is referred to as a
model. With each model, there are different assumptions made of the data, and certain models are more
applicable for specific data, for example, one model may account for local variation better than another.
Each model produces predictions using different calculations. Based on how the sample points are
distributed, each interpolation method will compute a different result. No matter which interpolation
method is used, the more input points and the greater their distribution, the more reliable the end result.
In this module we focus on the deterministic functions. Kriging is beyond the scope of this course. Feel
free to consult ArcGIS Desktop Help to learn more about Kriging.
Note: the distribution of the observation points is very important when applying an interpolation.
Depending upon the interpolation when making a DEM, observation points must be well distributed
over the whole surface according to the relief of the terrain. Sharp changes in height over a short
distance requires more observation points than relatively flat areas.
Figure 2: Inverse distance weighted interpolation method (IDW): the values of points closer to the
observation point (points in the circle) are more similar to the value of the observation point than the
points further away.
IDW allows you to control the influence of the observation points upon the interpolated values, based
upon their weight and their distance from the output point:
• The power parameter in the IDW interpolation controls the significance of the
surrounding points upon the interpolated value. Power is the exponent of distance. A
higher power results in less influence from distant points: distant points receive lower
weights. However, lower powers tend to treat all the sample points equally resulting in
a smoother surface. When the power is 0 there is no decrease in weight with distance
which means that all observation points get the same weight. The prediction will be the
mean of the measured values.
• A specified number of points or all points within a specified radius can be used to
determine the output value for each location.
The IDW interpolation method can be accessed using the Spatial Analyst toolbar or the IDW tool in
the Interpolation toolset of the Spatial Analyst toolbox.
There are two Spline methods: Regularized and Tension. The Regularized method creates a smooth,
gradually changing surface with values that may lie outside the sample data range. The Tension method
controls the stiffness of the surface according to the character of the modeled phenomenon. It creates a
less smooth surface with values more closely constrained by the sample data range.
Further control of the output surface is accomplished through two additional parameters: weight and
number of points. Consult ArcGIS Desktop Help to learn how these two parameters influence Spline
interpolation (use keyword ‘Spline interpolation’, subsequently choose the described subentry).
In Figure 3, Spline estimates the value of the selected cell at 23. Spline tries to fit a curve using the
selected subset of samples, in this case 6 samples. The curve would start at one of the cells with a value
of 10, start up to a cell with a 20, continue up or overshooting, then come back down to another 20 and
back down to a 10. The estimate cell, 23, may have been on the upswing of the curve.
INSTRUCTIONS:
1.
b. Is this a vector or raster dataset vector or raster structured? Explain your answer.
2.
In this exercise you have to interpolate the elevation point observations using different interpolation
methods, types and parameter settings.
a. Set the Output Extent to ‘Same as Layer lu_raster’ and set the Mask to ‘lu_raster’ (see Module 7
how to set the output extent and mask).
b. Create 5 digital elevation models by interpolation of the elevation points. Use the interpolation
parameters as given in Table 1.
c. Display the DEMs you created. The DEMs must be displayed with the same symbology in order to
compare. Open the Symbology editor, right-click the box that contains the ranges and labels,
choose ‘Load class breaks’ from the menu and select ‘Elevation_classification.xml’ from the data
folder.
d. Fill in Table 2. You cannot open de attribute table of the DEM (because this is a continuous raster
dataset). You must use dataset statistics to retrieve minimum, maximum and average elevation.
These statistics can be found under the Source tab in the Layer properties window.
e. How is it possible to get lower and higher elevation values after interpolation than the minimum
and maximum values contained in dataset ‘Height_observations’?
IDW_4_5
IDW_4_25
IDW_12_5
IDW_50_5
Spline_4_5
3.
a. Describe in your own words how the Variable IDW-interpolation method works.
b. What is the effect of cell size when using IDW as an interpolation method? (compare IDW_4_5
with IDW_4_25)
c. What is the effect of increasing the number of nearest points when using IDW as an interpolation
method? (compare IDW_4_5 with IDW_12_5)
d. Describe in your own words the differences between the Fixed Radius IDW-interpolation method
and the Variable IDW-interpolation method. (compare IDW_4_5 with IDW_50_5)
e. Write down the effect of a 50 meter radius on the spatial coverage of the interpolation. What could
you do to improve the coverage of the interpolation ‘IDW_50_5’.
g. For which type of terrain is it better to use the Spline interpolation method instead of IDW? Zoom
in to different locations of the Wageningen South area and compare IDW_4_5 with Spline_4_5
h. Write down what the differences are between the different interpolation methods IDW and Spline.
Use interpolation Spline_4_5 for the next exercises. Remove the other created datasets from the data
frame!
IDW and Spline interpolations are basic methods. The more advanced interpolation method ‘Topo to
Raster’ is discussed in the follow-up course Geo-information Tools (GRS 20806).
When analyzing surfaces, you perform a specified calculation that results in different representations of
a surface or that derives patterns not readily apparent in the original surface, using a continuous raster
dataset. With surface-analysis operations, you can derive additional information by producing new data
and identifying patterns in existing surfaces.
Slope analysis
Slope is defined by a plane tangent to the surface as modelled by the DEM at any given point and
comprises two components:
• Gradient: the maximum rate of change of elevation from each cell to its neighbours;
• Aspect: the compass direction of this maximum rate of change.
The Slope function calculates the maximum rate of change between each cell and its neighbors, for
example, the steepest downhill descent for the cell (the maximum change in elevation over the distance
between the cell and its eight neighbors). Every cell in the output raster has a slope value. The slope
gradient can be calculated in percentage or degrees.
INSTRUCTIONS:
4.
Derive the slope gradient from the DEM you have created. Improve the display; change the number of
classes to 15.
a. What type of raster operation (module 7) is the calculation of slope gradient? Explain your answer.
b. Write down the maximum, minimum and mean of the slope gradient dataset.
Maximum:
Minimum:
Mean:
INSTRUCTIONS:
Cells with a zero slope (flat) receive the value ‘-1’. The slope aspect in the output raster is represented
in 8 cardinal directions, e.g. East [67.5 - 112.5 degrees], Southeast [112.5 - 157.5 degrees].
5.
a. What is the slope aspect of the Rhine facing side of the Wageningen hill?
b. Add layer landuse.lyr to the data frame. What is the average slope aspect of the feature
with FID 445?
You can access the contour function in ArcToolbox and in the Spatial Analyst toolbar. The latter is the
most convenient for mapping contour lines. The contour function does not connect cell centres; it
interpolates a line that represents the most likely location with the same height. Since the lines are
smoothed, an idealistic representation of the surface is produced. You can also create an individual
contour line by clicking the Contour tool , and then selecting a location in the view. The function
traces the contour of the value that the chosen point represents.
INSTRUCTIONS:
Smaller contour intervals create more contours. The optimum contour interval is a matter of the realistic
elevation change in the DEM, the size of the area covered by the DEM, the analyses to follow, and the
desired aesthetics of the contour line dataset.
6.
Derive 3 contour datasets from the DEM. Use the following contour parameter combinations.
________________________________________
Contour dataset 1 2 3
________________________________________
Contour interval 2 4 10
Base contour 0 6 6
________________________________________
Practical Manual
Module 10
‘Digital image processing’
11. DIGITAL IMAGE PROCESSING 10-1
INTRODUCTION 10-1
START THE PROGRAM ERDAS IMAGINE 10-2
PART 1: DISPLAYING AN IMAGE DATA FILE 10-3
Display of DN-range 0 ... 255 (no stretch).........................................................10-3
Display after linear stretch of DN-range minimum ... maximum.......................10-6
Display after linear stretch of DN-range 40 ... 90 .............................................10-6
Display after standard deviation stretch............................................................10-7
Display of color composites ...............................................................................10-8
PART 2: SUPERVISED CLASSIFICATION 10-9
Examining land cover types using spectral profiles...........................................10-9
Digitizing training areas & estimation of signatures.......................................10-10
Collecting signatures.....................................................................................10-10
Evaluating Signatures ...................................................................................10-11
Land cover classification .................................................................................10-13
Minimum distance classification ..................................................................10-14
Maximum likelihood classification...............................................................10-14
Updating a color palette ................................................................................10-15
Exporting to ArcGIS file format .......................................................................10-15
LGN Database..................................................................................................10-15
IMAGE SOURCES 10-16
RELATED INTERNET SITES 10-17
10. DIGITAL IMAGE PROCESSING
Introduction
The aim of the exercises in this module is to acquire a first experience in understanding remote sensing
data by handling multi-spectral image data with the GIS and Remote Sensing package Leica Erdas
Imagine for Windows.
For the exercises, we will use image data of Wageningen and its surroundings (Figure 5). This is a
subset of a much larger scene taken by the remote sensing satellite Landsat-5 TM (Thematic Mapper)
on 11 July 1995. A spatial subset of the entire scene (185x185 km2), with seven spectral bands is
available (Table 1).
The image covers an area of 15.3×15.3 km2 and consist of 510 columns × 510 rows. Each pixel
represents an area of 30×30 m2. The sensor of band 6 observes pixels with a size of 120×120 m2.
In this module you will practice with different image processing techniques including different display
methods, the use of color composites and supervised classification.
In this module:
An introduction to the software package Erdas Imagine.
Displaying an image data file: stretching and color composites.
Selecting training sites for classification.
Collecting spectral signatures of training sites.
Three supervised classification methods.
Objectives
After having completed this module you will be capable:
to understand the principle behind various image display techniques;
to perform a supervised classification with Erdas Imagine;
to describe the differences between three supervised classification methods.
1. Start the Erdas Imagine package. Click start, select Programs ERDAS Geospatial
Imaging 9.3 ERDAS IMAGINE 9.3.
NOTE: The first time you start Erdas Image the program might give some errors, Ignore them
and start the program again.
2. Click Session in the main menu bar (Figure 1), click preferences.
3. Set Default data directory: to: D:\IGI\...*…\Erdas_imagine\data (*morning or afternoon).
4. Set Default output directory to: D:\IGI\...*…\Erdas_imagine\workspace (*morning or
afternoon).
1. Set your default Data directory and Output/Workspace directory
On top of the Erdas Imagine window you see the main menu bar (Figure 1). Clicking one of the items
of the menu bar gives a pull down menu with a number of options.
Just below the menu bar you see the viewer (Figure 2). The menu and icons in the viewer can be used
to open an image and applying basic viewer functions. If you move the cursor over the icons, you see a
short indication of the function in the lower left corner of the viewer.
In order to get familiar with image processing and remote sensing data we start with displaying and
processing of one image data file.
We use the image data of a Landsat-5 TM recording of band 4 (see Table 1) during this exercise. This
band contains spectral information of a near-infrared band: 0.76-0.90 µm. You will find that different
image stretching techniques of the same image data file produce different pictures on the screen.
This way, the Grey Scale Palette produces a picture on the screen with 256 grey tones. The range in
grey tones is a linear scale from black (DN-value 0) to white (DN-value 255); each DN-value of the
image will basically have its own grey value (Figure 4).
Since not all DN-values from 0 up to 255 are present in the original image, not all grey tones are used
in the picture. Although you will recognize Wageningen and surroundings, the picture can be made
brighter. But first you examine the DN distribution of the image. In order to examine the DN
distribution of an image, display the histogram.
INSTRUCTIONS:
1. Click the ImageInfo button in the standard toolbar or click Utility Layer info.
2. The ImageInfo window opens, showing file, layer, statistics and map information.
3. Select layer 4.
4. Click the Histogram tab or the histogram button in the toolbar of the ImageInfo window.
5. If the cursor is placed inside the histogram, three vertical lines are displayed showing the
minimum, maximum and mean values.
2.
a. What is plotted at the horizontal axis and what at the vertical axis?
b. Write down the values for minimum, maximum, mean and standard deviation for band 4.
When an image is displayed on the screen, the DN-values (File Pixel values) are translated to a grey
tone (Lookup Table (LUT) -value). In case of an image displayed without stretch, the DN-value is the
same as the LUT-value (Figure 4). You can view these values with Inquire cursor (click or
Utility Inquire cursor). It is important to zoom in to a level where you can distinguish the
individual pixels.
3.
a. Check the DN-values and LUT-values of water, grass, forest and heath land. You can find the
location of these objects in Figure 5. Write down the values in Table 2 in the ‘no stretch’ columns.
b. Which cover types has a DN-value of less than 30 in band 4? Explain this in terms of
absorption/reflectance.
c. Which two factors determine the grey tone of a pixel on the screen?
Figure 5. Selected training fields in the Landsat TM band 5 scene of 11 July 1995 of the area around
Wageningen.
Figure 6. Principle of linear stretch of image values (DN) into display levels. The linear relationship
between DN-value and LUT-value is in this example: LUT = 2.60*DN-156.
INSTRUCTIONS:
1. To apply a linear stretch to your image with the instructions mentioned on page 10-3, click in
the viewer menu bar: Raster Data scaling.
2. Make sure Linear selected in selected in the Binning field.
3. Replace the values for Min and Max with the minimum and maximum DN-values from the
image info.
4. Click OK.
4.
Open a new viewer (click the viewer button in the main menu bar), but don’t close the viewer where
you showed your image with “no stretch”. In this new viewer display the same image, but now with
min..max linear stretch.
a. Check the DN-values (FILE PIXEL) and LUT-values of the four land cover types again.
Add these values to Table 2 in the ‘linear stretch’ columns.
b. Can you explain the changes in LUT-values? Explain why some land cover types get a higher LUT
value, while other land cover types get a lower LUT-value.
5.
Open a new viewer (click the viewer button in the main menu bar), but don’t close the viewers where
you showed your image with “no stretch” and with “linear stretch DN-range min..max”
Use the data scaling function to apply a linear stretch of the DN-range from 40 to 90.
a. Which land cover types can you distinguish now with more grey tones in a smaller DN-range?
b. Investigate the DN and LUT-values of the four land cover types. Explain the linear stretch
principle.
INSTRUCTIONS:
1. Open a new viewer, select band 4 for a display in grey scale, but do not switch the ‘no stretch’
button on this time. The image will now be opened with Standard Deviation Stretch.
2. This stretch function can also be assessed through the menu bar. Click Raster Contrast
Standard Deviation Stretch.
3. You can use Tile Viewers to put the viewers easily in one screen. Click in the viewer menu
bar: View Tile Viewers.
6.
Open a new viewer (click the viewer button in the main menu bar), but don’t close the viewers where
you showed your image with “no stretch” and with “linear stretch DN-range min..max”.
a. Compare layer 4 of ‘wag95.img’ with the three other different stretching options. Which stretch
function gives in your opinion the best picture?
b. Which grey tone has grassland (see e.g. the meadows near the river) in band 3; is this different
from band 4? In what way? Explain the difference (remember the typical spectral signature of
green vegetation).
7.
Landsat-5 TM band 6 contains the thermal-infrared image data. Display layer 6 of ‘wag95.img’. You
can re-open the image, or change the band which is displayed with
b. Which cover type has a relative low temperature and which one has a relatively high temperature?
INSTRUCTIONS:
1. Open a viewer.
2. Add a raster layer to the viewer. Click the Raster Options tab.
3. Display as: True Color (even if you want to display False or Pseudo colors).
4. Attach bands to the Red, Green and Blue colors.
5. Click OK.
6. When a color composite is opened, you can always change the band combination. Click
Raster Band Combinations….
7. Change the spectral bands for the three channels. If the Auto Apply box is ticked, band
changes appear immediately on screen.
Note: the terminology used by Erdas Imagine may be confusing. The fact that you use the option ‘true
color’ in the selection menu does not mean that you display a true-color image. This depends on the
spectral bands you attach to the Red, Green and Blue band respectively.
8.
Open three color composites of image ‘wag95.img’ with band combinations as described in Table 3.
b. Check the colors for the cover types water, forest and bare soil in each composite. Write your
findings down in Table 4.
c. Which band combination or color composite shows the largest contrast between the different land
cover types? Why?
Table 3. Band combinations of three types of color composite for Landsat TM5..
Red Green Blue
True Color 3 2 1
False Color 4 3 2
Pseudo Color e.g. 4 e.g. 5 e.g. 3
As a preparation for supervised classification, one decides beforehand which cover types must be
classified and one selects proper training areas. These training areas are known cover types, based on
field visits or general knowledge of parts of the area. Since we assume that you have some knowledge
of the area around Wageningen, you will make several classifications without extensive fieldwork.
Statistical characteristics of the spectral data of the selected training areas are set down in signature
files. These signature files are then used by the classification method to derive the class boundaries for
each cover type in the feature space. The actual classification of all pixels is performed in this feature
space.
9.
INSTRUCTIONS:
1. Open the spectral profile tool. Click in the viewer menu bar: Raster Profile Tools….
2. Click Spectral and click OK. The Spectral Profile window opens.
3. Click to activate the inquire tool
4. Click with the inquire cursor a land cover type in the image. The spectral profile of this pixel
will be drawn in the graph. The line represents the value of the selected pixel for each band
(Figure 7).
5. To display wavelength on the x-axis click Edit Use Sensor Attributes…. Click the Sensor
type dropdown arrow and select landsattm.
10.
Try to locate a few different land cover types (water, forest, agricultural land, and town) and show their
spectral profiles in the graph.
a. Which two bands show the largest difference in pixel value between water and vegetation?
Representative examples of these cover types are shown in figure 5. You will use user-defined
polygons in the image for the selection of training samples
Note: The training areas are in general small areas with at least 25 pixels. These areas should be chosen
as pure (homogenous) as possible, so if you digitize e.g. a training site of water in the river, do not
include the river borders!
Collecting signatures
INSTRUCTIONS:
1. Open a new viewer and display your most expressive composite (see your answer to exercise
8c) and zoom in to get a more detailed look at the picture during digitizing.
2. Click in the main menu bar and click Signature Editor.... A new window will be
opened, move it so the area with the training fields can be seen clearly.
3. Click in the viewer menu bar AOI Tools...
4. Click the AOI Tool palette button to create a polygon. Draw a polygon in one of the
training areas (see figure 5). Digitize polygon points by clicking the LMB (Left Mouse
button) and finish it by double clicking the LMB.
11.
Digitize the 8 training areas (7 indicated in figure 5 and the class town) according to the steps described
above, and add the signatures to the signature list.
Save the signature file in the workspace folder located in the Erdas Imagine folder. Name the signature
file ‘wag95_your_name.sig’.
Evaluating Signatures
Before you perform a classification you need to study the signatures to get an accurate idea about the
position and size of the classes in the feature space. You can present the results of the signature
computation in a mean plot or histogram. You can compare the signatures of the different cover types;
see if they are well separated. If not, then perhaps you did not choose the correct training area or it is a
matter of different growth conditions or a registration error is made during field visit at the time of
image recording. This way you can also get an idea if it is useful to perform the classification with all
available bands.
For this exercise you need a viewer with the source image wag95.img and the Signature Editor with
wag95_your_name.sig. Mark the signature you want to investigate by clicking the row in the column
with the > mark. In the ERDAS IMAGINE package the signatures can be studied in different ways.
INSTRUCTIONS:
1. Click in the Signature Editor window View Columns…, the Viewer Signature Columns
window opens.
2. Select all rows except red, green and blue, click Statistics… and click min, max and mean in
the Column Statistics window.
3. Click Apply in the View Signature Columns window; close this window and the Column
Statistics window.
4. If you move the slide bar in the Signature Editor window to the right and you will see that all
statistical values appear.
12.
a. Which spectral bands show the clearest (spectral) distinction between land use classes?
INSTRUCTIONS:
1. Click in the Signature Editor window View Mean Plots..., the Signature Mean Plot
window opens. Depending on the option you choose you can display either the marked
signature or selected signatures or all signatures. You can select more than one signature by
keeping the shift key down during selection in the signature editor.
Show histograms
INSTRUCTIONS:
1. Click in the Signature Editor window View Histograms..., the Histogram Plot Control
Panel opens and simultaneously the histogram of the first band of the marked signature
appears.
2. Select the classes you want to display in a histogram in the Signature Editor if you want to
visualize multiple classes in one plot.
3. The chosen options in the Histogram Plot Control Panel are activated when you click the
Plot... button.
13.
a. Check the separability of the classes in all spectral bands by examining the histograms.
b. Which bands can be used to differentiate between deciduous and pine forest?
c. Suppose you could only use three spectral bands for land use classification. Which three bands
would you choose?
For the classifications in this module, you will use the parametric decision rules ‘Minimum distance’
and ‘Maximum likelihood’.
INSTRUCTIONS:
1. Click in the main menu bar and click Supervised Classification. The Supervised
Classification window opens (Figure 8).
2. Select the Input Raster File, this is the image you want to classify.
3. Select the Input Signature File, this is the file in which you stored the spectral signatures of
the training areas. It can be found under your ERDAS workspace directory.
4. Give the output image a name in the Classified File box.
5. Select classification decision rules: Non-parametric Rule, Overlap Rule, Unclassified Rule
and Parametric Rule.
6. Click OK.
Figure 8. The supervised classification window where you name the output files and set the decision
rules.
14.
a. Open the Supervised Classification window. Which Parametric Rules are available?
15.
Carry out a supervised classification of ‘wag95.img’ with the Minimum Distance (MD) rule.
Name the output image wag95-MD.img. Use the following classification setting:
a. Display the classification result in a new viewer and notice that all pixels are classified. Note this
image has nothing to do with spectral reflectance. You are looking at a land cover map, where pixel
values indicate a land cover class.
b. Table 5 lists four control points. Write down the land cover class of each control point in the ‘MD’
column. Use the inquire cursor tool to retrieve the value of the control point. The value corresponds
to a land cover class.
16.
Carry out a supervised classification of ‘wag95.img’ with the Maximum Likelihood (MLHD) rule.
Name the output image wag95-MLHD.img. Use the following classification setting:
b. Write down the land cover class of each control point in the ‘MLHD’ column of Table 5. Use the
inquire cursor tool to retrieve the value of the control point. The value corresponds to a land cover
class.
INSTRUCTIONS:
1. Click in the viewer menu bar Raster Attributes…. Select the layers (spectral bands) you
want to use for classification. The Raster Attributes Editor opens.
2. Click a colored cell in the color column and select a color from the list.
17.
Change the colors of one of the classification results to land use map colors (town = red, forest = dark
green, water = blue, etc…).
INSTRUCTIONS:
1. Click in the main menu bar Import. Select the classification result you want to use for further
analysis. Export it to GRID format. Save it in the proper location.
2. Open the exported file in ArcGIS (ArcMap or ArcCatalog).
18.
Export one of the land cover classification results to GRID format. Save the dataset in the ArcGIS
workspace folder.
LGN Database
Many remote sensing satellites collect Remote Sensing data. Image data from the Landsat-5 TM
satellite is used, a comparable way you just did, to make land-use classifications for the "Landelijke
Grondgebruiksclassificatie van Nederland" (LGN).
The LGN database covers The Netherlands and is created and updated at the Centre for Geo-
Information within the Environmental Sciences Group of the WUR. The data can be obtained from the
Geodesk of the Centre for Geo-Information. The LGN database is updated on a regular basis and is
very useful for all kinds of applications, e.g. for planning and for environmental scenario studies.
A part containing the surroundings of Wageningen is copied from LGN4 (2000) and available in wag-
lgn4.img.
19.
a. Give a few reasons why wag-lng4 and your classification results are not exactly the same?
Compare three images of different sensors and describe the strong and weak points of each data-set.
The available images are a MERIS, Landsat TM and Quickbird image, all covering the area around
Wageningen.
MERIS has a high spectral and radiometric resolution and a dual spatial resolution; 1200m and 300m.
Quickbird images can be either panchromatic with a spatial resolution of 0.61m, or multi-spectral,
which results in a pixel size of about 2.5m.
INSTRUCTIONS:
1. Open a false color composite of the three images (Meris_wag.img, wag95.img and
quickbird_27032003_rd.img) in separate viewers.
2. Investigate the spectral profile of several landcover types for the three images.
3. To display wavelength on the x-axis click Edit Use Sensor Attributes…. Click the Sensor
type dropdown arrow and select MERIS, landsattm, or QuickbirdMS respectively.
20.
a. Which bands did you select for each false color composite?
b. Are the images geometrically and atmospherically corrected? How did you determine this?
d. Why is it not possible to build a space-born remote sensing sensor, which has a good spatial,
spectral and temporal resolution?
For more concerning the basic principle and display of color composites, check:
http://chesapeake.towson.edu/data/all_composite.asp
Bruns, H.T. and M.J. Egenhofer, 1997. User interfaces for Map Algebra. Journal of
the Urban and Regional Information Systems Associations. Vol 9, No. 1, pp
44-54.
Lillesand T.M., R.W. Kiefer and J.W. Chipman, 2004. Remote Sensing and Image
Interpretation (5th ed). John Wiley & Sons, Inc., Hoboken, NJ, USA.