Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Conference held at Anaheim, California, June 1996). IAHS Publ. no. 239, 1997.
331
DAVID F. MAUNE
Dewberry & Davis, 8401 Arlington Boulevard, Fairfax, Virginia 22031-4666, USA
INTRODUCTION
In 1994, in response to severe flooding in Georgia, Florida, Alabama and Texas,
Dewberry & Davis surveyed nearly 8000 flooded buildings to collect flood inventory
data for FEMA's Individual Assistance Program. Concurrently, certified flood
adjusters made "windshield survey" damage estimates for these buildings; these
estimates were subsequently found to be in error by 50-100%. Damage estimates can
be vital for timely and correct rebuild/buy-out decisions. These decisions often
depend on whether estimated repair costs exceed 50% of the replacement value of the
flooded building. In 1994, FEMA determined that existing computer models could
more accurately estimate flood damages by knowing three things about each flooded
building:
(a) What is the approximate area of the building's "footprint"?
(b) What is the building's estimated replacement value?
(c) How deep was the interior flooding, to the nearest whole foot (30 cm)?
The answers to (a) and (b) could be collected in advance for all flood-prone
buildings in a community; but FEMA needed a way to quickly obtain the answers to
(c) for each building actually flooded. Additionally, D&D sought other means to help
flood plain managers to be truly proactive. Central to this need was the means to
better perform flood hazard identification and risk assessment, vital for flood
mitigation initiatives.
332
David F. Maune
GPS COMPARISON
In 1995, in cooperation with the Louisville and Jefferson County, Kentucky,
Metropolitan Sewer District, FEMA sponsored a comparison of two GPS
technologies on the basis of vertical accuracy and cost/productivity. We refer to
these technologies as "GPS Backpack" operated by Larry N. Scartz, Ltd, and "GPS
TruckMAP" operated by John E. Chance & Associates. Both utilized Trimble
4000SSe receivers with real-time kinematic (RTK) and on-the-fly (OTF) reinitialization. They used alternative techniques for surveying the 3-D coordinates
(latitude, longitude, and elevation) of survey target points on buildings without
intruding on private property. D&D calls this "stand-off surveying".
For the productivity portion of the test, nearly 1300 high-density houses were
surveyed to determine if Elevation Certificates could be mass produced for US$30
per house, as opposed to the typical price of US$250 per house.
For the accuracy portion of the test, 62 of the 1300 total houses were selected to
be independently surveyed by both methods because they presented technical
difficulties in one or more of the following ways:
(a) they were located along tree-lined streets where tree canopy cover would
interfere with GPS signals and where D&D could test the OTF capabilities when
satellite lock was lost;
(b) they were located on the opposite side of hills from the GPS base station, where
RTK radio corrections would have difficulty reaching the GPS rover units; and/or
(c) they were located up to 200 feet (61 m) off the road so that elevations would be
"cantilevered" by significant distances.
These three technical challenges were considered essential to test the true
capabilities and limitations of stand-off GPS survey techniques. BackPack and
TruckMAP would independently survey these 62 houses, and correct for local
variations in gravity. D&D would then compare the two elevation data sets and
determine if FEMA's 6-inch (15 cm) vertical accuracy requirement for this test was
satisfied. If the GPS technologies performed well under these difficult conditions,
they could be relied upon also to perform well under simpler conditions.
When the two elevation data sets were laid side-by-side for the 62 difficult
houses, the results were amazing! The elevations all agreed within about one inch
(2.5 cm). The r.m.s. difference was two-thirds of an inch (1.7 cm), and the
maximum error was less than +2 inches (5 cm) at the 95% confidence level. In highdensity housing areas, both methods proved that highly-accurate Elevation
Certificates could be mass produced for less than US$30 per house.
Both BackPack and TruckMAP gave comparable results and were considered
useable. FEMA subsequently sponsored GPS elevation surveys of thousands of other
homes in 61 counties in eight states nationwide. With the best geodetic-grade GPS
receivers and exacting procedures, D&D found that survey control points and
benchmarks are typically in error by 6 to 12 inches (15 to 30 cm), and sometimes by
several feet or more. D&D found some new homes nationwide had been constructed
at elevations that make them vulnerable to predicted floods in the future. And D&D
found that approximately one-third of conventional Elevation Certificates, that
establish the cost of flood insurance for post-FIRM homes, were in error by more
than 1 foot (30 cm) when checked by more-accurate survey methods.
333
5. Conventional Elevation
Certificates: costly (typically
US$250), less accurate.
6. Pre-FIRM buildings currently
don't require Elevation Certificates
to identify actual flood risks.
Subsidy is expensive; Congress
directed 1996 subsidy restudy.
7. Difficult to predict candidate
buildings for retrofit/floodproofing.
8. Post-flood "windshield"
estimate errors of 50-100%.
9. Lengthy delay for disaster
inventories and "rebuild/buy-out"
decisions for buildings
substantially damaged.
334
David F. Maune
Inability to quantify flood hazards and risks. Challenge 3 indicates the dilemma
in being unable to accurately quantify hazards and risks from 500-year, 100-year,
50-year, and 10-year floods. By pre-flood surveying the elevation of the reference
level of each building in or hear a Special Flood Hazard Area (SFHA), the
community can estimate, on a house-by-house basis, the depth of interior flooding
that would be caused by the standard flood events. The computer models, cited
above, can then compute the estimated damages to each building, and to the
community as a whole, as a result of the standard flood events (500-year, 100-year,
50-year, and 10-year floods). Such hazard identifications and risk assessments are the
key to all mitigation efforts, and the community can then be aggressive and proactive
in taking mitigation initiatives to reduce future flood losses, and in promoting flood
insurance to owners of at-risk homes.
Difficulty in justifying drainage improvement projects. Challenge 4 indicates that
it is difficult to justify drainage improvement projects without detailed elevation data
on individual buildings in the drainage basin. For example, how does one prove
whether or not it is worth US$2 million to construct a drainage improvement project
that will lower the base flood elevation (BFE) by 2 feet for an area that includes 400
flood-prone homes? By knowing the elevation of the lowest floor of each home, its
"footprint" area, and its replacement value, computer models can accurately estimate
expected damages from standard flood events prior to drainage improvements, and
then recompute the expected damages with drainage improvements that lower the
BFEs. The drainage improvement project benefits can be determined in terms of the
damage avoided.
Limitations in conventional Elevation Certificates. Challenges 5 and 6 pertain to
conventional Elevation Certificates, which are sometimes considered to be an
impediment to the sale of flood insurance. All Elevation Certificates (conventional or
GPS) are expensive when not mass-produced. Although Elevation Certificates aren't
required for pre-FIRM buildings (constructed prior to publication of Flood Insurance
Rate Maps of the area), the US Congress has directed a 1996 study of the current
subsidy for pre-FIRM homes. Without Elevation Certificates, it is difficult to identify
candidate buildings for retrofit/flood-proofing.
Challenges 5 and 6 can be solved by producing highly-accurate GPS Elevation
Certificates, mass-produced and quality-controlled, for all buildings in or near flood
plains, providing them free to pre-FIRM and post-FIRM homeowners, and
encouraging them to purchase flood insurance. A sample GPS Elevation Certificate is
at Fig. 1. In addition to the individualized photograph of the building in question, the
background map pinpoints the building's geographic location centred on the base
map road network and also its position in or near the SFHA shown in blue. The BFE
is interpolated to the nearest 0.1 foot (3 cm), and the elevation of the "target point"
surveyed on the house is also shown to the nearest 0.1 foot. Target points are most
typically the bottom of front door (BFD) or the top of foundation (TOF). Offsets to
below-ground floors are estimated, based on standard 8-foot (244 cm) basement
foundations, or 9-foot (274 cm) standard offsets between floors. Corrections can be
made by the insurance agent and owner if the offset distance error is significant for
insurance rating purposes. The estimated depth of interior flooding from the 100-year
l-*1
!id
5.-S
.! fc
Qi
< I
^v--
.,:^|lF^
mm^Mtm
ii
335
Ml ?
Hi
I 111
I
336
David F. Maune