Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: www.elsevier.com/locate/rser

Surface measured solar radiation data and solar energy resource assessment MARK
of Pakistan: A review
⁎
Z.R. Tahir , Muhammad Asim
Department of Mechanical Engineering, University of Engineering and Technology Lahore, Pakistan

A R T I C L E I N F O A BS T RAC T

Keywords: Over the past decade, Pakistan has been facing a shortage of between 3000 MW and 6000 MW in the supply and
Global horizontal irradiation demand of electricity, leading to several hours of power cut. Sixty-three percent of current electricity generation
Surface measured data is based on fossil fuels, leading to cost vulnerability, whereas only one percent is based on renewable resources
Solar resource assessment (non-hydroelectric). The geographical location and climatic conditions of the country offer a high solar energy
World Radiation Data Centre
potential, which signifies the need for solar resource assessment for the planning of solar energy projects. The
Pakistan
article provides an overview of the solar resource assessment based on surface measured data and satellite data
carried out by researchers. Various researchers have developed solar energy potential maps for the country
based on solar insolation data measured by Pakistan Meteorological Department (PMD). The data measured by
PMD reported in the previous literature have been compared with the long-term data from the World Radiation
Data Centre (WRDC), and both data show good agreement. The approximate annual mean daily global
horizontal insolation for the whole country, based on surface measured data by WRDC, is 5.30 kW h/m2 (19.0
MJ/m2 or 221 W/m2). The government of Pakistan sought to exploit the solar and wind energy potential during
the last decade with the help of international agencies. Based on satellite data, the solar resource assessment
was done by National Renewable Energy Laboratory, USA (NREL) in 2007, and the Energy Sector Management
Assistance Program (ESMAP) of the World Bank in 2015. The solar atlas prepared by NREL was not validated
by the surface measured data, whereas the solar atlas prepared under ESMAP is to be validated in 2017. A
comparison of NREL estimated data and WRDC data shows that NREL's atlas overestimates solar insolation up
to 25% in some regions. The annual mean daily solar global horizontal insolation estimated under ESMAP is
5.67 kW h/m2.

1. Introduction
Pakistan has been facing severe electricity crisis since 2007 due to
the enormous difference in demand and supply leading to load
shedding of several hours daily across the country. The installed
generation capacity of the country is about 23,600 MW, and the power
shortage is approximately 3000 MW to 6000 MW during peak hours
[1–3]. The energy shortage becomes worse during the summer, leading
to prolonged load shedding of between eight and eighteen hours, which
varies significantly between urban and rural areas [1–3]. The country's
energy mix is based on 63% of fossil fuels, out of which 37% is fuel oil,
which means the cost of generation is expensive and highly dependent
on oil price changes, leading to cost vulnerability. The energy produced
by renewable sources (other than large hydroelectric) contributes less
than 1% of the current energy mix [1,4]. The Alternative Energy
Development Board (AEDB) Pakistan, reports that 28 solar power
projects with a capacity of 956.8 MW are under development within
the framework of the AEDB policies and procedures [5]. The solar
market in the country is at a nascent stage, and the country has
tremendous potential to meet its power needs from solar energy
resources. According to AEDB [6] the wind energy has commissioned
project capacity of 308.2 MW, under construction project capacity of

Abbreviations: AEDB, Alternative Energy Development Board Pakistan; CSR, NREL's Climatological Solar Radiation; DHI, Diffused Horizontal Irradiation; DNI, Direct Normal
Irradiation; ESMAP, Energy Sector Management Assistance Program; GHI, Global Horizontal Irradiation or Insolation; GOES, Geostationary Operational Environmental Satellite; GTI,
Global Solar Irradiation on Flat Plate Tilted at Latitude; IEA, International Energy Agency; kW h, Kilowatt Hour; MAPE, Mean Absolute Percentage Error; MBE, Mean Bias Error; MJ,
Mega Joule; MPE, Mean Percentage Error; MW h, Megawatt Hour; NASA, National Aeronautics and Space Administration; NREL, National Renewable Energy Laboratory, USA; NWFP,
North-West Frontier Province of Pakistan (now Khyber Pakhtunkhwa); PMD, Pakistan Meteorological Department; RMSE, Root Mean Square Error; SARI/E, South Asia Regional
Initiative for Energy Cooperation; SSE, Solar Meteorology and Solar Energy service; SUNY, State University of New York at Albany; TW h, Terawatt Hour; UNDP, United Nations
Development Program; USAID, U.S. Agency for International Development; WMO, World Meteorological Organization; WRDC, Mean Absolute Percentage Error
⁎
Corresponding author.
E-mail address: ziartahir@uet.edu.pk (Z.R. Tahir).

http://dx.doi.org/10.1016/j.rser.2017.06.090
Received 10 April 2016; Received in revised form 19 May 2017; Accepted 22 June 2017
Available online 06 July 2017
1364-0321/ © 2017 Elsevier Ltd. All rights reserved.
Z.R. Tahir, M. Asim Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861

477 MW and projects of 663 MW capacity are in the pipeline. The

energy production for wind has been demonstrating a viable market for
non-hydro renewable energy in the country.
The first ever solar power plant was inaugurated in Islamabad on
May 29, 2012, under the project titled ‘Introduction of Clean Energy by
Solar Electricity Generation System’, which is a special funding from
the Japan International Cooperation Agency (JICA) under the Cool
Earth Partnership. Two photovoltaic power (PV) plants each of 178 kW
capacity were installed at the premises of the Pakistan Engineering
Council and the Planning Commission of Pakistan under this project
[7]. Quaid-e-Azam Solar Park, the first ever photovoltaic power station
in Pakistan of 1000 MW capacity is under construction in Bahawalpur
[8]. The solar power plants are under development in Punjab, Sindh,
Balochistan and Kashmir under a development program by the
International Renewable Energy Agency with the cooperation of
China and private sector energy companies in Pakistan [7].
The utilisation of solar energy began in Pakistan in the 1980s. Pallet
and Brabben [9] reported the installation and operation of twenty solar
powered micro-irrigation units (solar water pumps) at research and
private agriculture farms throughout the country with the support of
the Pakistan Agriculture Research Council (PARC) and Agriculture
Development Bank of Pakistan (ADBP). The Directorate General of
Energy Resources, Ministry of Petroleum and Natural Resources
started a program of solar villages, and the first solar energy system
was installed in 1981; this solar village had a power capacity of 5 kW
and the electric power was utilised for domestic purposes and water
pumping [10].
In Pakistan, PMD started measuring solar radiation data for five Fig. 1. (a) World map of global horizontal irradiation (kW h/m2) [26] (b) World map of
meteorological stations and sunshine duration data for 37 stations direct normal irradiation (kW h/m2) [26].
from 1957. WRDC centrally collects and archives the global solar
radiation data, which is the only reliable source to access long-term
measured solar radiation data for Pakistan; the other source is PMD. This article presents the review of the surface measured data of
The solar radiation data for Pakistan measured by PMD have been solar radiation for Pakistan, and the solar energy potential assessment
reported by many researchers [11–22] to estimate the solar energy based on surface measured data and satellite data. The articles related
potential of Pakistan. The solar energy potential maps (monthly and to renewable energy for Pakistan, especially solar energy, from inter-
annual) for the country were first developed by Raja and Twidell nationally recognised peer-reviewed journals to date were collected
[13,14] using surface measured data of global solar insolation and initially, and articles containing measured solar radiation and solar
sunshine duration of 30 years, the annual mean daily solar irradiation energy potential especially were selected. The articles were divided into
(GHI) in the major parts of the country was estimated from 4.4 to 6.0 three main sections:
kW h/m2 with mean of 5.3 kW h/m2. Ahmad et al. [20] developed
monthly maps of solar radiation (GHI) and reported that the annual • Articles related to surface measured data of solar insolation,
mean daily solar insolation for the whole country is 5.5 kW h/m2. The estimation of solar insolation based on measured solar insolation
minimum annual mean daily solar radiation (4.3 kW h/m2) of Pakistan and sunshine duration, estimation of solar radiation based on
is higher than the global annual mean daily (3.6 kW h/m2), which empirical relationships, and solar insolation maps.
reflects the excellent solar energy potential of the country [21]. • Articles related to the estimation of solar insolation based on
The research work performed to estimate the solar resource satellite data and solar insolation maps, including solar resource
assessment of Pakistan based on satellite data is limited [23–25] and assessment performed by NREL and ESMAP.
not a single attempt has been made at the national level to estimate • Articles related to the description of the solar energy potential,
solar resource potential of the whole country using long-term satellite including solar energy and the solar energy potential of the country
data. The National Renewable Energy Laboratory, USA (NREL), reported by different sources are compared, validated and pre-
developed solar resource maps for Pakistan under the South Asia sented.
Regional Initiative for Energy Cooperation (SARI/E) program of the
U.S. Agency for International Development (USAID), but these maps The long-term solar radiation data from the WRDC for five
were not validated against surface measured data. The Energy Sector meteorological stations of Pakistan was analysed to obtain monthly
Management Assistance Program (ESMAP) of the World Bank devel- mean daily values. The monthly mean daily data measured by PMD
oped maps for the solar resource estimate for Pakistan in the year 2015 and reported by researchers for the different duration for five meteor-
under a three-year project, which will complete in 2017. The SolarGIS ological stations were compared with the WRDC data. The monthly
[26] is a geographical information system designed to meet the needs mean daily solar insolation from NREL's CSR model, NREL's SUNY
of the solar energy industry; it provides worldwide maps for solar model and NASA-SSE was also compared with the long-term surface
global horizontal irradiation, direct normal irritation and diffused measured data. The solar energy potential of the country reported in
horizontal irradiation. The world maps of global horizontal irradiation the previous literature was analysed and critically reviewed. The annual
and direct normal irradiation (daily mean and annual sum) are mean daily global horizontal irradiation from the literature was
presented in the Fig. 1(a) and (b) respectively [26] to show the solar analysed, and the solar resource potential for the whole country was
potential of Pakistan compared to the world. The solar potential of calculated.
Pakistan (especially Sindh and Baluchistan provinces) is among the
second highest around the world.

2840
Z.R. Tahir, M. Asim
2. Solar resource assessment based on surface measured
data
This section presents an introduction to solar radiation terminology
and the review of the assessment of solar energy potential based on the
surface measured data. PMD began measuring global solar insolation
data in 1957 for five meteorological stations and stopped around 2000
in most of the stations [27]; the sunshine duration data for 37 stations
were also recorded from 1957 till 2000 [22]. The major research work
was reported from 1986 to 1996, and there was no significant work
reported after 1996 due to discontinuation of measurement of the
surface data. The solar energy potential in the form of solar radiation
maps was prepared by estimating the solar insolation from sunshine
durations by the majority of researchers, and the maps were presented
in the form of iso-radiation lines. The solar insolation data was
presented in various units in the previous literature, which are MJ/
m2, cal/cm2, kW h/m2 and W/m2; for consistency, the data presented
in the literature was converted into kW h/m2, whereas the data with
original units is presented in parentheses.
2.1. Solar radiation
Solar radiation or irradiation is the incident energy received per
unit area of a surface for a particular period (either hour or day),
measured in kW h/m2 or J/m2. Solar insolation is the term specifically
used for solar energy irradiation. Solar irradiance is the rate at which
the incident energy is received per unit area of a surface, measured in
W/m2. Direct solar irradiation or beam irradiation is the radiation
received from the sun without being scattered by the atmosphere, while
diffused solar irradiation is the solar radiation received from the sun
after being scattered by the atmosphere. Solar global irradiation is the
total of the beam irradiation and diffused irradiation received on a
horizontal surface, commonly termed global insolation or global
horizontal irradiation. Direct normal irradiation is the radiation
received on a surface which is always perpendicular to the sun, whereas
global tilted irradiation is total radiation received on a tilted surface.
The commonly used terms for solar energy applications are Global
Horizontal Irradiation (GHI), Direct Normal Irradiation (DNI),
Diffused Horizontal Irradiation (DHI) and Global Tilted Irradiation
(GTI). The solar constant or extraterrestrial radiation is the rate of
energy received on the top of the Earth's atmosphere from the Sun on a
unit area of the surface perpendicular to the direction of propagation of
radiation. The global mean value of the solar constant adopted by the
WRDC is 32.81 kW h/m2 (1367 W/m2) [28].
Solar energy is captured and utilised by three different technolo-
gies; solar photovoltaic (PV), solar thermal and concentrated solar
power (CSP). These technologies use different components of the solar
radiation. The accurate and reliable data of solar energy resources is
the key element in the successful use of these technologies. For a
specific location, the historic trends with annual, seasonal, monthly,
daily, hourly and sub-hourly variation in solar radiation data are
required to predict the performance of the solar system [29]. The
global horizontal irradiation and global tilted irradiation are necessary
for Photovoltaic (PV) applications, whereas direct normal irradiation is
important for solar thermal applications, concentrating solar power
(CSP) and concentrated photovoltaic (CPV) applications [30,31]. In
this article, the terms solar radiation, solar insolation or solar irradia-
tion means global horizontal irradiation and diffused solar insolation or
irradiation means diffused horizontal irradiance, whereas other terms
used have their usual meanings.
The estimation of solar insolation from measured sunshine dura-
tion was first proposed by Ångstrom [32], as presented in equation (1)
in Ref. [33]; this relationship was based on total radiation received on a
day with clear sky. Black et al. [33,34] modified the relationship
presented as equation (2) in Ref. [33] and was based on total radiation
received if the atmosphere was perfectly transparent (Angot's value
2841
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
[35]). The relationship based on extraterrestrial radiation modified
later [36,37], given by Eq. (1).
H n
=a+b
Ho N (1)
where H is the monthly mean daily insolation on the horizontal surface,
Ho is the monthly mean daily extraterrestrial insolation on a horizontal
surface for the specific location using equation (1.10.3) of Ref. [28], N
is the monthly mean of maximum possible daily hours of bright
sunshine (sunshine duration) calculated from Cooper's formula [38],
and n is the monthly mean daily hours of bright sunshine. The
regression constants a and b are estimated by using measured values
of sunshine duration and monthly mean daily solar insolation. These
regression constants are used to estimate the global horizontal insola-
tion using measured data of sunshine duration for a potential location.
2.2. Solar radiation data and solar energy potential maps
This section presents the review of the articles on solar energy
potential maps and surface measured solar radiation data. The articles
are divided into three categories based on the outcome of the research;
first containing articles presenting solar potential maps, second global
horizontal insolation data and analysis, third diffused and direct
radiation data and analysis. Section 2.2.1 contains the review of articles
presenting monthly and annual maps of global horizontal insolation in
the form of iso-radiation lines. These maps were prepared by estimat-
ing GHI: from measured GHI and sunshine duration using by
Ångstrom type regression Eq. (1), from daily temperature, and using
ASHRAE formulation. Section 2.2.2 contains the articles presenting
solar radiation data measured by PMD and other sources, the articles
containing a statistical analysis of data are also presented in this
section. Section 2.2.3 contains the articles presenting data and maps of
diffused and direct (beam) insolation.
2.2.1. Solar energy potential maps
This section contains the articles presenting solar energy maps in
form of global horizontal insolation, these maps are either monthly and
annual. The global horizontal insolation in all articles (except two) was
estimated using Ångstrom equation from solar insolation and sunshine
duration data measured by PMD for different durations. The results of
these researchers were quite different due to the duration of data used
for estimation and discussed in details to see the interannual variability
of GHI for different long-term data.
The solar energy potential maps of the whole world was prepared by
Löf et al. in 1966 for the first time [39,40]. The solar energy potential
was presented as monthly mean daily values of global horizontal
isolation (total solar radiation incident on a horizontal surface) in the
form of iso-radiation lines at an interval of 0.58 kW h/m2 (50 cal/cm2).
The measured solar insolation data for 668 stations was used, and for
233 stations solar insolation data was estimated by using measured
data of sunshine duration (using Ångstrom type regression equation
described in Section 2.1) to develop maps. In 1973, Mani and Chacko
[41] presented the solar radiation maps and data of solar radiation of
thirteen meteorological stations of India for the duration of ten years.
The solar radiation maps in the form of iso-radiation lines were
presented for monthly mean daily values of solar global insolation,
solar diffused insolation, and the monthly mean hourly values for direct
normal insolation.
The monthly and annual maps of solar energy potential in the form
of iso-radiation lines (global horizontal insolation) for Pakistan pre-
sented in five articles [12,14,16,20,21], and annual maps of solar
energy potential were presented in three articles [19,22,27] using data
for various durations (from 1959 to 1989). The solar radiation data and
sunshine duration data used for the estimation of solar potential maps
was measured by PMD, the solar global horizontal insolation data was
measured with Epply and Kipp Pyranometers and sunshine duration
Z.R. Tahir, M. Asim
Table 1
Details of monthly and annual maps for global horizontal insolation developed for Pakistan.
Reference Maps type Duration of data used
for maps
Sukhera and Pasha [12] Annual, Up to 14 years
Monthly
Raja and Twidell [14] Annual, Up to 37 years
Monthly
Chaudhary [16] Annual, Up to 29 years
Monthly
Ghaffar [21] Annual, 15 years
Monthly
Muneer et al. [27] Annual Up to 15 years
Adnan et al. [22] Annual 30 years
were measured using Campbell-Stokes recorders [16].
The first attempt to develop solar radiation maps in the form of the
iso-radiation lines for Pakistan was done by Sukhera and Pasha [12] in
1987. Sukhera and Pasha [12], Raja and Twidell [13,14], Chaudhary
[16] and Ghaffar [21] developed monthly and annual maps of solar
potential from solar radiation and sunshine duration using Ångstrom
type regression Eq. (1) for various durations, the details are presented
Table 1. The iso-radiation lines for these maps were slightly different
due to interannual variability because of the different duration of long-
term data used, the annual mean daily global horizontal insolation
estimates are presented in Table 1. The global horizontal insolation was
estimated from sunshine duration using regression constant a and b,
the values regression constants used by these researchers to develop
maps were different which led to slight variation in the results. The
details of the results of these articles are presented in the following
paragraphs.
Sukhera and Pasha [12] estimated monthly maps using solar
radiation data of five stations and sunshine duration data for 35
stations. The measured data for sunshine duration were from 1969 to
1973 for Peshawar and 1969-82 for other stations. The measured data
for solar irradiation were from 1957 to 1982 for Multan and 1969-82
for other stations. The value of regression constants a and b for each
station was calculated using measured data, which vary from region to
region. The difference between the annual mean of measured and
predicted GHI values was less than 5%. The monthly mean daily GHI
was estimated by the Ångstrom type regression Eq. (1); the values of
the regression constants, a and b estimated from measured solar
insolation and sunshine duration of each station, were used, which
were presented in Ref. [11]. Thirty-five stations were divided into five
groups by geographical locations, including stations with measured
GHI in each group, and the regression constants calculated from
measured data were used in each group to estimate solar insolation.
They reported that the monthly mean daily GHI varies from 3.33
kW h/m2 to 7.22 kW h/m2 (12 MJ/m2 to 26 MJ/m2) for central
regions, from 2.78 kW h/m2 to 8.05 kW h/m2 (10 MJ/m2 to 29 MJ/
m2) for northern regions and from 4.44 kW h/m2 to 6.94 kW h/m2 (16
MJ/m2 to 25 MJ/m2) for southern regions of the country. The
minimum values of GHI were observed in December/January and
maximum values in June for all regions, the highest values were found
at stations with higher altitudes, towards the western part of the
country.
The monthly maps using 40 stations data up to 1987 for the
duration varying from 13 to 37 years, were developed by Raja and
Twidell [13,14]. The measured data of solar irradiation data from seven
stations (five from Pakistan and two neighbouring stations from India)
and sunshine duration data of 40 stations (37 from Pakistan and three
from India) were used to estimate the GHI. The measured data for GHI
– for Karachi, Multan and Quetta from 1957 to 1987, for Peshawar
from 1957 to 1981, for Lahore from 1957 to 1981 and two locations
from India (Srinagar and Jodhpur) – were used. The two set of values
of the regression constants a and b (calculated from measured GHI and
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
Stations used for solar Stations used for GHI estimated for Annual mean daily GHI
radiation sunshine duration stations (kW h/m2)
5 stations 35 stations 35 stations 4.80–5.71
7 stations 40 stations 40 stations 5.27
5 stations 37 stations 37 stations 4.65–6.28
5 stations 37 stations 37 stations 5.27
6 stations – – 5.06
– 37 stations 58 stations 5.11
sunshine hours for seven stations, five stations in Pakistan and two
stations from India), were used in Eq. (1) to estimate GHI. They
observed that the predicted GHI using regression constants for only
five stations of Pakistan gave better agreement with measured GHI
compared to GHI estimated by using regression constants for seven
stations. It was reported that solar distribution near the border of India
fits reasonably close when compared with [41] and is in agreement with
world solar distribution maps [39,42,43]. It was observed that the
lowest monthly mean daily GHI was in December/January, ranging
from 2.22 kW h/m2 (8.0 MJ/m2) in the extreme north to 4.31 kW h/m2
(15.5 MJ/m2) at the coastline of Balochistan and central parts of Sindh.
The annual mean daily GHI in the major regions of the country
received from 4.44 to 5.97 kW h/m2 16.0–21.5 MJ/m2) with a mean of
5.27 kW h/m2 (19.0 MJ/m2).
Chaudhary [16], developed global horizontal insolation maps using
measured solar isolation of five stations and sunshine duration for 37
stations for the duration of up to 29 years from 1958 to 1986. The
regression constants of Ångstrom regression equation were computed
for four seasons (winter, spring, summer and autumn). Thirty-seven
meteorological stations were divided into groups based on altitudes,
and monthly mean daily solar insolation was estimated from sunshine
duration using Ångstrom formula for using values of a and b for a
particular group. The solar insolation data for five stations using these
regression constants agree well with measured data with less than 10%
error. The comparison of measured and predicted monthly mean daily
solar insolation is presented in [16]. It was reported that the monthly
mean daily GHI received by Sindh and Balochistan provinces is more
than 5.11 kW h/m2 (440 cal/cm2), for Punjab and NWFP province it is
between 4.65 and 5.11 kW h/m2 (400–440 cal/cm2), the northern
areas and Kashmir are less than 4.65 kW h/m2 (400 cal/cm2). The
maximum solar radiation is received in June in Quetta and Zhob which
is 7.90 kW h/m2 (680 cal/cm2). The annual mean daily GHI for the
whole country is between 4.65 and 6.28 kW h/m2 (400–540 cal/cm2).
The monthly mean daily contour maps for global horizontal
insolation using measured data from 1975 to 1989 were presented by
Ghaffar [21]. The monthly mean daily GHI received by whole country
was 4.44–5.83 kW h/m2 (16–21 MJ/m2) with an annual mean of 5.27
kW h/m2 (19 MJ/m2) over most regions. The minimum annual mean
daily GHI (4.44 kW h/m2 or 16 MJ/m2) of Pakistan is higher than
world annual mean daily GHI (3.61 kW h/m2 or 13 MJ/m2) which
reflects the excellent solar energy potential of Pakistan [44]. It was
reported that the total available annual potential of solar energy over
the total area (796095 km2) of the country is 1453 GW h (5230 TJ)
based on annual mean daily GHI of 5.27 kW h/m2 (19 MJ/m2).
The solar potential maps in the form of annual mean daily global
horizontal isolation were developed by Raja and Abro [19], Muneer
et al. [27] and Adnan et al. [22]. The annual map using global
horizontal insolation data for 40 stations were developed by Raja and
Abro [19], it was reported that the maximum mean daily GHI, ranging
from 6.11 to 8.33 kW h/m2 (22.0–30.0 MJ/m2), is received in summer.
The annual mean daily GHI varies between 4.44 and 5.83 kW h/m2
2842
Z.R. Tahir, M. Asim
(16.0–21.0 MJ/m2), and most parts of the country receive insolation of
5.27 kW h/m2 (19.0 MJ/m2). Muneer et al. [27] presented the solar
potential map using global horizontal irradiation data for a duration up
to 15 years for six stations. The duration of the measured data was
1975–1989 for Karachi, Lahore, Multan and Quetta, 1975–1981 for
Peshawar and Islamabad it was 1985 only. It was reported that all
stations, except Quetta, had ceased solar insolation measurement. The
annual daily mean for these stations is 5.06 kW h/m2 (18.22 MJ/m2).
Adnan et al. [22] developed annual solar radiation map using data
from 58 meteorological stations (nine from the northern area and Azad
Kashmir, 12 from NWFP, fourteen from Punjab, ten from Sindh and,
13 from Balochistan) for the durations of 30 years from 1971 to 2000.
The GHI was estimated from the sunshine duration data by using
Angstrom formula [32] and regression constants (computed in Table 1
of [16]). The stations where sunshine duration data were not available,
the GHI was estimated using Hargreaves method [45] and de Wit
method [46] from the mean monthly maximum and minimum
temperatures. The monthly mean daily GHI for 58 stations is presented
as Fig. 1 to Fig. 5 in Ref. [22]. It was reported that monthly mean daily
GHI more than 4.80 kW h/m2 (200 W/m2) was observed from
February to October in Sindh, from March to October in Balochistan
and Punjab, and from April to September in NWFP, Northern Areas
and Azad Kashmir. In most parts of the country, monthly mean daily
GHI was 3.60–6.60 kW h/m2 (150–275 W/m2). The GHI values
presented in the article were based on 10 h a day, but daily mean
values are presented here; the annual mean daily GHI for 58 stations
was 5.11 kW h/m2 (213 W/m2).
Ahmad et al. [20] developed monthly maps of global horizontal
insolation in the form of iso-radiation lines for Pakistan. The GHI was
estimated using ASHRAE (American Society of Heating, Refrigeration
and Air Conditioning Engineers) formulation modified for Pakistan.
The estimated values were compared with measured data of hourly
solar radiation for some meteorological stations, the measured and
estimated data agreed well with 5% error. It was observed that annual
mean daily GHI is minimum in the northern region of the country
during December, which is 2.50 kW h/m2, and maximum in the south
of the country during June, which is 8.30 kW h/m2. Quetta receives
maximum GHI and the annual mean daily GHI for the whole country is
5.50 kW h/m2.
Solar energy resource data is a key element in designing, financing
and successful operation of any solar energy system (Solar PV, Solar
thermal and CSP systems) [47]. IEA task 46 was initiated in 2011 to
examine and compare various solar energy potential assessment
databases (ground measured and satellite) so that quality is ensured
in solar energy data for the potential location of the project [48,49].
The purpose of presenting this section was to collect and analyse the
ground measured solar insolation data for different locations and
conclude the reason for the variation in the values so that readers
and designers have a better understanding of this variation. The results
of above-discussed researchers were quite different due to the duration
of data used for estimation. As per recommendations of IEA Task 46,
the long-term ground measured data is more reliable and accurate, so
the solar radiation data for the researchers with 29–37 years record
(Table 1) should be used for solar energy system designing as the
quality of such data is reliable. The IEA Task 46 also demands the
supply of the corresponding uncertainty presented in the solar resource
assessment data [48]. The researchers discussed in the above section
had not provided much information about the uncertainty in the
measured data or the models used to assess the potential from
sunshine hours.
2.2.2. Global horizontal insolation data
This section presents the review of the articles that presented solar
radiation data from PMD and other sources. The global horizontal
insolation data measured by PMD from 1957 to 2000 was reported by
various researcher in the literature [10,11,16–18,27] in the form of
2843
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
monthly mean daily global horizontal insolation for different durations.
The global horizontal insolation data measured by the Pakistan Council
of Scientific and Industrial Research (PCSIR) Laboratories, Hyderabad,
Pakistan is presented in one article [50], and data measured by AEDB
at Baburband (near Karachi) were presented in one article [51]. The
solar radiation data presented in this section is compared with long-
term available data from WRDC in Section 2.4 to check inter-annual
variability.
In 1984 for the first time, Sukhera [10] presented monthly mean
daily solar global horizontal insolation data for Bahawalpur (Punjab)
for the utilisation of solar energy for the development of the Cholistan
Desert in Pakistan. The monthly mean daily solar insolation was from
3.22 to 7.39 kW h/m2 11.6–26.6 MJ/m2), the duration of the data
measurement was not mentioned in the article. The solar global
horizontal insolation data measured by PMD from 1957 to 1982 for
five metrological stations (Karachi, Lahore, Multan, Peshawar and
Quetta) was presented by Sukhera et al. [11] in 1986. The measured
data for solar irradiation were from 1957 to 1982 for Multan and 1969
to 1982 for other stations. The measured data for sunshine duration
were from 1969 to 1973 for Peshawar and 1969 to 1980 for other
stations. They reported that the measured annual total of mean daily
GHI ranged from 1750 to 2083 kW h/m2 6300–7500 MJ/m2) for five
stations. The measured monthly mean daily solar insolation of five
stations for 14 years from 1969 to 1982, except Peshawar, taken from
Table 3 of [11], are presented in Fig. 2. The solar isolation data of five
stations for the duration of up to 29 years from 1958 to 1986 was
presented by Chaudhary [16] in 1990. The measured data of GHI of
five stations is replotted in Fig. 3. The global horizontal insolation data
for four stations for the period of five years from 1980 to 84 was
presented by Nasir and Raza [17] in 1993. The annual mean daily GHI
for Quetta, Karachi, Peshawar and Lahore was 5.68 kW h/m2, 5.03
kW h/m2, 4.76 kW h/m2 and 4.38 kW h/m2, respectively. The monthly
mean daily GHI of four stations for a period of five years from 1980 to
1984, taken from Fig. 1 and Fig. 2 of [17], is presented in Fig. 4.
The global horizontal insolation data for six meteorological stations
for up to 15 years was presented by Raja and Twidell [18] in 1994. The
data were measured from 1975 to 1989 for Quetta, Karachi, Multan
and Lahore, from 1975 to 1981 for Peshawar, and, for Islamabad, for
1985, 1988 and 1989. The monthly mean daily GHI values vary from
2.78 to 9.72 kW h/m2 10–35 MJ/m2), the value is generally above 5.56
kW h/m2 (20 MJ/m2) during summer for all stations. The annual mean
daily GHI for Quetta, Karachi, Multan, Peshawar, Lahore and
Islamabad were 5.73 kW h/m2 (20.64 MJ/m2), 5.10 kW h/m2 (18.39
MJ/m2), 5.06 kW h/m2 (18.21 MJ/m2), 4.79 kW h/m2 (17.24 MJ/m2),
5.05 kW h/m2 (18.18 MJ/m2) and 4.75 kW h/m2 (17.10 MJ/m2),
respectively. The measured monthly mean daily GHI of six stations
taken from Table 2 of Ref.[18] is presented in Fig. 5. Muneer et al. [27]
presented global horizontal insolation data for six stations for the same
duration as described above except Islamabad which was 1985 only. It
8
1986 Sukhera et al.
7
GHI (kWh/m2)
6
5
4
3
KHI LHE MUL
PWR QUT
2
1 2 3 4 5 6 7 8 9 10 11 12
Month Number
Fig. 2. The measured GHI of five stations for 14 years 1969–1982) from Ref. [11].
Z.R. Tahir, M. Asim Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861

8 8
1990 Chaudhry 2006 Muneer et al.
7 7

GHI (kWh/m2)
GHI (kWh/m2)

6 6

5 5

4 4

3 3
KHI LHE MUL ISB KHI LHE
PWR QUT MUL PWR QUT
2 2
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
Month Number Month Number
Fig. 3. The measured GHI of five stations for a period of up to 29 years 1958–1986) Fig. 6. The measured GHI of six stations for up to 15 years 1975–1989) from Ref. [27].
from Ref. [16].
Kalhoro and Abro [50] presented the global horizontal irradiation
8 data measured at the Pakistan Council of Scientific and Industrial
1993 Nasir and Raza
Research (PCSIR) Laboratories, Hyderabad Pakistan for a period of six
7
months from January to June 2003. The solar radiation data were
measured continuously by using EPLAB Pyranometer. The maximum
GHI (kWh/m2)

6
monthly mean daily GHI was recorded in June as 7.15 kW h/m2 and
the minimum in January as 4.11 kW h/m2. The total GHI received
5
during the six months was reported as 1802.38 kW h/m2 (this reported
value does not seem feasible, because it is much higher, even for the
4
maximum value of mean daily solar insolation, 7.15 kW h/m2)
3 calculated over a period of six months (181 days), which is approxi-
KHI LHE PWR QUT mately 1294.15 kW h/m2. The same measured data was presented as
2 monthly mean hourly GHI as Fig. 8 of Ref. [52]; the maximum average
1 2 3 4 5 6 7 8 9 10 11 12 value at noon ranges from 14.40 to 21.60 kW h/m2 (600–900 W/m2)
Month Number from January to June. It was reported that, for seven months (mid-
Fig. 4. The measured GHI of four stations for five years 1980–1984) from Ref. [17]. March to mid-October), the mean GHI at noon exceeded 19.20 kW h/
m2 (800 W/m2).
Khahro et al. [51] presented three years data (May 2008 to April
8 2011) of global horizontal insolation data for a prospective location
1994 Raja and Twidell (latitude 25.12 N and longitude 67.53 E) in the southern region of
7 Sindh, measured by AEDB. The monthly mean daily GHI is presented
in Table 5 and Table 6 of Ref. [51], the maximum value was reported as
10.41 kW h/m2 (37.49 MJ/m2) in May, whereas the minimum value
GHI (kWh/m2)

6
was 5.67 kW h/m2 (20.42 MJ/m2) in December and the annual mean
5 daily insolation was reported as 7.91 kW h/m2 (28.47 MJ/m2). The
annual total of mean daily GHI was 3.13 GW h/m2 (11.30 GJ/m2). The
4 values reported throughout the above article are not sensible and there
seems to be a systematic error in processing the raw data for the given
3 location.
ISB KHI LHE
MUL PWR QUT The annual mean daily global horizontal insolation of six stations
2 for various durations as discussed above is presented in Fig. 7 [11,16–
1 2 3 4 5 6 7 8 9 10 11 12 18,27]. The average of annual mean daily global horizontal insolation
Month Number
Fig. 5. The measured GHI of six stations for a period of up to 15 years 1975–1989) from
6.0
Ref. [18]. 1986 Sukhera 1990 Chaudhry 1993 Nasir
1994 Raja 2006 Muneer Mean of All
was reported that all stations, except Quetta, had ceased solar insola- 5.6
tion measurement. The annual mean daily GHIs for Quetta, Karachi,
GHI (kWh/m2)

Multan, Peshawar, Lahore and Islamabad were 5.73 kW h/m2 (20.64

5.2
MJ/m2), 5.38 kW h/m2 (19.38 MJ/m2), 5.06 kW h/m2 (18.22 MJ/m2),
4.84 kW h/m2 (17.43 MJ/m2), 4.79 kW h/m2 (17.25 MJ/m2) and 4.75
kW h/m2 (17.12 MJ/m2), respectively. The annual daily mean for these 4.8
stations is 5.06 kW h/m2 (18.22 MJ/m2). The measured monthly mean
daily GHI of six stations for a period of up to 15 years, taken from 4.4
Table 3 of [27] and Table 4 of [52], is presented in Fig. 6. The GHI data
presented in Figs. 5 and 6 are for the same duration but coming from
4.0
two different articles [18,27], there is slight difference in the monthly
ISB KHI LHE MUL PWR QUT
mean daily values so this data is a good source to compare with WRDC
data. Fig. 7. The annual mean daily GHI of six stations for various durations from Ref.
[11,16–18,27].

2844
Z.R. Tahir, M. Asim
of six stations for five durations is also presented. The annual mean
daily GHI and standard deviation (presented in parenthesis) for
Karachi is 5.14 kW h/m2 (0.10 kW h/m2), Lahore is 4.72 kW h/m2
(0.11 kW h/m2), Multan is 5.02 kW h/m2 (0.04 kW h/m2), Peshawar is
4.88 kW h/m2 (0.21 kW h/m2), Quetta is 5.73 kW h/m2 (0.04 kW h/
m2), and Islamabad is 4.75 kW h/m2 (0.00 kW h/m2). The standard
deviation in the annual GHI values is for Peshawar and Quetta is 0.04
kW h/m2, for Karachi and Lahore is about 0.10 kW h/m2. The
monsoon clouds affect the solar energy in the entire country during
July to September, except for western Balochistan, and lower the
sunshine duration [53]. The more deviation in the solar radiation is
observed in the regions which are affected by the monsoon, for same
duration the reduction in the solar insolation is different for different
stations which are presented in Fig. 9.
The monthly diurnal variation of the global horizontal insolation for
five meteorological stations from 1975 to 1981 is presented in [54].
The average hourly solar insolation for five stations is symmetrical
about the local noon with a small difference in values before noon and
afternoon. During a day, the best time interval for the utilisation of
solar energy is 10:00 a.m. to 3:00 p.m. local time (PST), and the mean
solar insolation during this period is 70% of the daily total value. The
average hourly solar insolation at midday lies between 0.47 kW h/m2
to 0.83 kW h/m2 (1.7 MJ/m2 to 3 MJ/m2). The solar insolation at
midday is greatest for the months of April to September, and the
highest value is in Quetta. Raja and Twidell [55] presented frequency
distribution of global horizontal insolation over Pakistan. The mea-
sured data of GHI for Lahore, Multan, Karachi and Quetta for the
twelve-year duration 1975–1986) and Peshawar for a seven-year
duration (1975–1981) were analysed. It was reported that all five
stations have mean daily GHI of more than 2.78 kW h/m2 (10 MJ/m2),
whereas Karachi and Quetta have more than 4.16 kW h/m2 (15 MJ/
m2), with at least 85% probability. The cumulative frequency distribu-
tions of measured global insolation data [18] reveals that Islamabad,
Karachi, Lahore, Multan, Peshawar and Quetta are expected to receive
mean daily GHI of more than 2.78 kW h/m2 (10 MJ/m2) on an average
for 324, 361, 338, 351, 331 and 349 days per year, respectively. The
monthly mean daily GHI values vary from 2.78 to 9.72 kW h/m2 10–35
MJ/m2), the value is generally above 5.56 kW h/m2 (20 MJ/m2) during
summer for all stations.
2.2.3. Diffused and direct (beam) solar isolation maps and data
Direct Normal Irradiance (DNI) is used for potential estimation and
operation of concentrated solar power (CSP) technologies. These
technologies are used for electricity generation using the solar thermal
system. Limited research work is carried out for the DNI estimation in
Pakistan. The details of these work are presented here.
The monthly diffused and direct (beam) insolation maps for
Pakistan in the form of iso-radiation lines were developed by Raja
and Twidell [15] using data from 40 stations. The measured diffused
solar insolation for only one meteorological station (Quetta) of
Pakistan was available for the three-year duration from 1960 to
1962. Three empirical relationships were used to predict diffused
insolation were reported: equation (1) of [15], developed by Liu and
Jordan [56,57], and also presented in [58]; equation (2) of [15],
developed by Modi and Sukhatme [59] for India; and equation (3) of
[15], developed by Page [36] and also presented in [58]. The measured
data of diffused insolation were compared with data predicted by the
three relationships. The empirical relationship developed by Modi and
Sukhatme [59] was found satisfactory and was used to estimate the
diffused insolation for 40 stations; 37 stations in Pakistan and three in
India (Bhavnagar, Jodhpur and Srinagar). The direct insolation for 40
stations was computed by the difference of global and diffuse insola-
tion. The diffused radiation maps were compared with the maps of the
neighbouring country, India [41], and world maps [60], and the results
were compatible. The monthly mean daily diffused solar insolation
received for the whole country during a year ranges from 1.94 to 2.50
2845
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
kW h/m2 7–9 MJ/m2), with maximum in July for all stations, ranging
from 2.78 to 3.05 kW h/m2 10–11 MJ/m2), and minimum in the
winter months, ranging from 1.11 to 1.94 kW h/m2 4–7 MJ/m2). The
maximum values of diffused insolation were in the extreme north and
south due to maximum cloud cover and the minimum values were in
Quetta. The annual diffused insolation was about 40–50% of the global
insolation. The monthly mean daily direct solar insolation received for
the whole country during a year ranges from 1.94 to 3.33 kW h/m2 7–
12 MJ/m2) with the maximum value in Quetta, and most of the country
receives an annual mean daily DHI of about 2.78 kW h/m2 (10 MJ/
m2), except the extreme northern region. It was reported that the direct
to the diffused insolation ratio was about 3:1 and 1:3 in winter and
summer, respectively.
Khahro et al. [51] estimated diffused, direct and total solar
insolation from measured global horizontal insolation for a prospective
location in the southern region of Sindh. Nineteen different models
from the previous literature and nine new models were developed to
estimate diffuse radiation on the horizontal surface. The optimum
model for diffused solar irradiation was used to estimate diffused,
direct and total solar irradiation on a tilted surface. The tilt angle of
10–30 degrees (annual optimum 23 degrees) gave the highest annual
total of solar insolation on the tilted surface as 3.69 MW h/m2 (13.3
GJ/m2). Abbasi and Qureshi [61] estimated global horizontal isolation
and diffuse solar insolation for Chhor city in Pakistan. The global
horizontal irradiation was estimated from extra-terrestrial solar radia-
tion using Angstrom formula [32], and the extra-terrestrial solar
radiation was estimated using equation (2) of Ref. [61] taken from
[62]. The diffused solar irradiation was estimated by using the three
relations presented in Ref. [61]; equation (3) from [56], equation (4)
from [63], and equation (5) from [36]. The estimated data for global
and diffused irradiation were compared with NASA satellite data, and it
was reported that the estimated GHI agreed well with NASA with a
value of the coefficient of determination of 98.7%. The diffused
irradiation estimated by Hawas and Muneer model [63] agreed well
with the NASA data. The maximum and minimum values of GHI were
observed in June and December respectively.
2.3. World Radiation Data Centre (WRDC)
The World Radiation Data Centre (WRDC) is the laboratory of
Voeikov Main Geophysics Observatory, Russian Federal Service for
Hydrometeorology and Environmental Monitoring located in St.
Petersburg. It is one of the globally recognised data centres, which
centrally collects and archives global solar radiation data and provides
open access to the international research community. The WRDC is
sponsored by the World Meteorological Organization (WMO) and
published data on solar radiation once a month since 1964 and four
times a year since 1993. The WRDC is the only reliable source to access
long-term measured solar radiation data for Pakistan; the other source
is PMD, which sells solar radiation data. The solar radiation data
available for Pakistan at WRDC were provided by the PMD to WMO.
The sunshine duration and global solar insolation as daily mean and
monthly mean daily are available at WRDC. The procedure to calculate
daily and monthly data for solar insolation and the procedure to
evaluate data quality is presented on the WRDC website.
The solar radiation data of seven meteorological stations in
Pakistan (Islamabad, Karachi, Khuzdar, Lahore, Multan, Peshawar
and Quetta) from 1964 are available at WRDC. The details of the seven
meteorological stations for Pakistan are presented in Table 2. The data
for the seven stations were obtained from the WRDC for the duration
presented in the Table; the data of some months were not included in
the WRDC database due to poor data quality. The monthly mean daily
solar insolation for available data was calculated for the whole duration
using all available values. The available data of Khuzdar were for five
years, but not a single year was complete and, therefore, it is not
presented here. The WRDC data of solar insolation were presented in
Z.R. Tahir, M. Asim Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861

Table 2
Geographical location and duration of solar radiation data of six stations.

Station WMO index Latitude (°N) Longitude (°E) Altitude (m) Solar insolation (kW h/m2) Duration

Minimum Maximum Annual

Islamabad 41,571 33.62 73.10 547 2.23 5.42 3.95 1990–2003

Karachi 41,780 24.90 67.13 022 4.03 6.51 5.31 1966–1995
Khuzdar 41,744 27.83 66.63 488 – – – 1996–1998, 2002–2004
Lahore 41,640 31.55 74.33 214 2.61 5.93 4.36 1966–2003, 2006–2007
Multan 41,675 30.20 71.43 123 3.09 6.45 5.06 1964–1994
Peshawar 41,530 34.00 71.52 359 3.02 7.30 5.16 1964–1981
Quetta 41,661 30.18 66.95 1721 3.40 7.26 5.41 1964–1994, 1997–2008

8 daily solar insolation for Karachi from all data sources is 5.18 (0.18)
WRDC
kW h/m2; the standard deviation is mentioned in the parentheses. The
7 average of mean daily solar insolation for Lahore, Multan, Peshawar
and Quetta are 4.66 (0.28) kW h/m2, 5.03 (0.11) kW h/m2, 4.92 (0.26)
GHI (kWh/m2)

6 kW h/m2 and 5.68 (0.20 kW h/m2), respectively. The average of

5 0.07 kW h/m2 (maximum of 0.13 in July); for Lahore is 0.11 kW h/m2
4
3
ISB KHI LHE
MUL PWR QUT
2
1 2 3 4 5 6 7 8 9 10
Month Number
Fig. 8. Long-term measured monthly mean daily GHI of six stations taken from WRDC.
J/cm2, and for comparisons, the data is converted into kW h/m2.
The maximum and minimum values of monthly mean daily solar
global insolation for six stations for the long-term duration are
presented in Table 2. The minimum values are in the months of
December-January, whereas maximum values are in the months of
May-June. The annual mean daily solar insolation is also presented;
out of these six stations, the minimum solar insolation is for Islamabad
and the maximum is for Quetta. The maximum and minimum values of
monthly mean daily insolation and annual mean monthly insolation for
six stations for different durations are presented in Table 2 and shown
in Fig. 8.
2.4. Comparison of PMD measured data with WRDC data
This section presents comparison of the monthly mean daily global
horizontal insolation data measured by PMD [11,16–18,27] which was
used by researchers for different durations with long-term WRDC data.
The IEA SHC Task 46 [48] recommends that uncertainty in long-term
average data is a key factor in risk assessment of solar energy projects.
The data used by different researchers is compared with the WRDC
long term data and the uncertainty of each data is described. The solar
insolation data measured by PMD is presented by Sukhera et al. [11],
Chaudhary [16], Nasir and Raza [17], Raja and Twidell [18] and
Muneer et al. [27] presented solar insolation data for the duration of 14
years (1969–1982), 29 years (1957–1986), 5 years (1980–1984), 15
years (1975–1989) and 15 years (1975–1989), respectively. The
purpose of this comparison is to establish a link between published
data and WRDC data to check the suitability of WRDC data for further
research as readily assessable long-term data.
The average of monthly mean daily solar insolation data for six
meteorological stations measured by PMD was calculated (from various
available sources for different durations mentioned in Section 2.2). The
average of monthly mean solar insolation of data from WRDC and
PMD (Sukhera et al. [11], Chaudhary [16], Nasir and Raza [17], Raja
and Twidell [18] and Muneer et al. [27]) are presented in Fig. 9; the
standard deviation is presented as error bars. The average of mean
standard error for monthly mean daily solar insolation for Karachi is
(maximum of 0.18 in June and August); for Multan is 0.05 kW h/m2
(maximum of 0.10 in December); for Peshawar is 0.10 kW h/m2
(maximum of 0.17 in May); and for Quetta is 0.08 kW h/m2 (maximum
of 0.12 in June). The standard error for Islamabad, Karachi, Lahore,
Multan, Peshawar and Quetta are 0.19, 0.07, 0.11, 0.05, 0.10 and 0.08
11 12 kW h/m2, respectively. A comparison of monthly mean daily solar
insolation data for all data shows lower standard deviation, which
reflects lower variation in the data over the different time periods used
[11,16–18,27].
A correlation of monthly mean daily solar insolation from WRDC
and PMD is presented in Fig. 10 for five meteorological stations, the
coefficient of determination (R2) and the equation of the best-fit
regression line for each station are also presented. The mean percen-
tage error (MPE) and the mean absolute percentage error (MAPE) of
particular data were calculated by taking WRDC long-term data as a
reference. The mean bias error (MBE) was also calculated as the
difference of WRDC data and particular data. The positive error
represents that the particular data is lower than WRDC data, and vice
versa.
A comparison of data reported by Sukhera et al. [11] with WRDC
data is presented in Fig. 10. The MBE for Karachi, Lahore, Multan,
Peshawar and Quetta are +0.11, −0.40, +0.08, +0.17 and −0.27 kW h/
m2, respectively. The MAPE for Karachi, Lahore, Multan, Peshawar
and Quetta are 5.07%, 11.70%, 4.51%, 5.24% and 6.79%, respectively.
The data for Multan, Peshawar and Karachi agrees well with WRDC
with MPE of less than +4%, except for the month of July for Karachi
with an error of +16%. The data for Lahore and Quetta are higher
compared to WRDC, with MPE of −10.0% and −4.27%, respectively.
A comparison of data reported by Chaudhry [16] with WRDC data
is presented in Fig. 10. The MBE for Karachi, Lahore, Multan,
Peshawar and Quetta are +0.03, −0.17, +0.06, +0.51 and −0.38
kW h/m2, respectively. The MAPE for Karachi, Lahore, Multan,
Peshawar and Quetta are 1.12%, 4.15%, 1.76%, 10.10% and 6.78%,
respectively. The data for Karachi and Multan agree well with WRDC,
with MPE less than 2%. The data for Lahore and Quetta are higher
compared to WRDC, with MPE of −3.63% and −6.78%, respectively.
The data for Peshawar are lower compared to WRDC, with MPE of
+10.10%.
A comparison of data reported by Nasir and Raza [17] with WRDC
data is presented in Fig. 10. The MBE for Karachi, Lahore, Peshawar
and Quetta are +0.24, −0.38, +0.06, +0.51 and −0.30 kW h/m2,
respectively. The MAPE for Karachi, Lahore, Peshawar and Quetta
are 5.46%, 8.97%, 10.35% and 6.34, respectively. The data for Lahore
and Quetta are higher compared to WRDC, with MPE of −8.97 and
−5.79, respectively. The data for Karachi and Peshawar are lower

2846
Z.R. Tahir, M. Asim
Fig. 9. The average of monthly mean daily GHI (kW h/m2) with standard deviation of data from WRDC and PMD [11,16–18,27].
compared to WRDC, with MPE of +5.01% and +10.02%, respectively.
There is significant variation between the two data sets for all stations
because the duration of the data reported by Nasir and Raza [17] is five
years whereas the WRDC data is long-term.
A comparison of data reported by Raja and Twidell [18] with the
WRDC data is presented in Fig. 10. The MBE for Islamabad, Karachi,
Lahore, Multan, Peshawar and Quetta are −0.08, +0.20, −0.43, +0.00,
+0.11 and −0.32 kW h/m2, respectively. The MAPE for Islamabad,
Karachi, Lahore, Peshawar and Quetta are 19.29%, 3.88%, 9.68%,
1.57%, 2.41% and 5.57%, respectively. The data for Multan and
Peshawar agree well with WRDC, with MPE less than 3%, and the
data for Karachi also agree well, with MPE less than 4%, except for
September with an error of +12%. The data for Islamabad, Lahore and
Quetta are higher compared to the WRDC, with MPE of −18.79%,
−9.68% and −5.57%, respectively. The variation between the two data
sets for Islamabad is higher because the duration of the data is three
years compared with the long-term data of the WRDC. A comparison of
data reported by Muneer et al. [27] with WRDC data is presented in
Fig. 10, the statistical parameters considered for the data are similar to
Raja and Twidell [18].
The variations between datasets presented in this section for long-
term duration from various sources agree well with the WRDC data,
with MAPE of less than 10%, the slight variations are due to different
durations of the data measurements. The trend in the monthly mean
daily solar insolation for five stations during a year is similar to all
datasets. The solar radiation data shows good correlation with each
other, which reflects the reliability of the data; moreover, the WRDC
data are easily available compared to the PMD data. Thus, long-term
WRDC data can be used for further studies for the estimation of PV
power generation projects nearby these locations. The maximum and
minimum values of monthly mean daily, along with annual mean daily
solar insolation of five meteorological stations from the above-dis-
cussed studies and WRDC, are presented in Table 3.
2.5. Discussion on availability of surface measured data
The Pakistan Meteorological Department began measuring solar
global horizontal insolation data in 1957 for five meteorological
stations and stopped somewhere around 2000 in most of the stations,
presented in Table 2. The short-term data for two other stations,
Islamabad and Khuzdar, the duration is presented in Table 2. The
sunshine duration data for 37 stations was also recorded from 1957.
The major research work on the estimation of solar energy potential
based on the measured data (global horizontal insolation) was reported
from 1986 to 2000, and there was no significant work reported after
1996 due to discontinuation of measurement of the data. The annual
mean daily global solar insolation received in the major parts of the
country ranges from 4.44 to 5.97 kW h/m2, with a mean of 5.27 kW h/
m2 [13,14,19–21]. The majority of the articles (published before 1996)
reviewed in Section 2.2 reported the long-term data measured by PMD,
one article [50] reported six-month data measured by Pakistan Council
2847
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
of Scientific and Industrial Research (PCSIR) Laboratories. The solar
insolation data measured by AEDB were reported in [51]. Three masts
were installed in 2008 by the AEDB along with the UNDP under the
Global Environmental Facility (GEF) Wind Energy Program (WEP) to
measure wind speed [64]. The solar insolation data was also measured
on these masts for the duration of two of three years from 2008 to
2011. The diffused horizontal irradiation data for only one meteor-
ological station (Quetta) of Pakistan were available for the duration of
three years from 1960 to 1962, as reported in [15].
The long-term solar insolation data in Pakistan was measured by
PMD, as discussed in Section 2.2. This long-term data is available from
the PMD and can be purchased through their standard procedure,
while the same data is available from WRDC free of cost. The WRDC,
sponsored by the WMO, have published data on solar radiation since
1964 and is the only reliable source to access long-term measured solar
radiation data for Pakistan. The data from WRDC are for the sunshine
duration and global solar insolation as daily mean and monthly mean
daily, with the description of the procedure to calculate data and to
evaluate data quality. The short-term data for the solar insolation are
available from AEDB for three locations in Sindh province: Baburband
mast from July 2008 to May 2011, Keti Bandar from February 2009 to
October 2010 and for Hawkes Bay from April 2009 to May 2011. The
most recent measurement of solar radiation data for Pakistan was
started in 2015 at nine stations by the ESMAP of the World Bank in
Pakistan as a three-year project which will complete in 2017; the
details about ESMAP are described in Section 3.3.
3. Solar resource assessment based on satellite data
3.1. Introduction
This section presents the review of the solar energy resource
assessment of Pakistan based on satellite data. The research work
performed at the national level is very limited [23–25] and no attempt
has been made to estimate solar resource potential of the whole
country using long-term satellite data. The National Renewable
Energy Laboratory, USA (NREL) has developed solar resource maps
for Pakistan under the South Asia Regional Initiative for Energy
Cooperation (SARI/E) program of the U.S. Agency for International
Development (USAID); however, these maps were not validated by
surface measured data. The Energy Sector Management Assistance
Program (ESMAP) of the World Bank has developed maps for the solar
resource for Pakistan in the year 2015 under a three-year project which
will complete in 2017.
Malik et al. [23] develop solar insolation maps of Pakistan using
Geostationary Operational Environmental Satellite (GOES) cloud ima-
gery. The satellite data of three months were used; May 1986, August
1986 and October 1985. The solar insolation data were estimated from
GOES digital data obtained from INSAT satellite using a radiative transfer
model proposed by Dedieu et al. [65]. It was reported that GOES images
were available at a temporal resolution of three hours (0300, 0600, 0900
Z.R. Tahir, M. Asim Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861

8 8
WRDC vs 1986 Sukhera et al. WRDC vs 1994 Raja and Twidell
KHI ISB

1994 Raja and Twidell GHI (kWh/m2)

1986 Sukhera et al. GHI (kWh/m2) 7 7
LHE KHI
LHE
MUL
6 6 MUL
PWR PWR y = 1.404x - 0.794
QUT QUT R² = 0.955
5 5
y = 1.028x - 0.257 y = 1.012x - 0.264
R² = 0.849 R² = 0.940
4 4
y = 1.058x - 0.377 y = 0.983x + 0.471 y = 1.004x - 0.021 y = 1.124x - 0.112
R² = 0.967 R² = 0.882 R² = 0.991 R² = 0.979
3 3
y = 1.033x - 0.337 y = 1.150x - 0.545 y = 1.108x - 0.263 y = 0.987x - 0.046
R² = 0.982 R² = 0.970 R² = 0.993 R² = 0.995
2 2
2 3 4 5 6 7 8 2 3 4 5 6 7 8
WRDC GHI (kWh/m2) WRDC GHI (kWh/m2)
8 8
WRDC vs 1990 Chaudhry WRDC vs 2006 Muneer et al.
KHI ISB
7 7

2006 Muneer et al. GHI (kWh/m2)

LHE KHI
1990 Chaudhry GHI (kWh/m2)

LHE
MUL
6 6 MUL
PWR PWR y = 1.402x - 0.785
QUT QUT R² = 0.954
5 5
y = 0.960x + 0.184 y = 1.013x - 0.271
R² = 0.989 R² = 0.942
4 4
y = 0.968x + 0.103 y = 1.068x - 0.127 y = 0.999x + 0.009 y = 1.126x - 0.115
R² = 0.996 R² = 0.993 R² = 0.991 R² = 0.979
3 3
y = 0.919x - 0.091 y = 1.100x - 0.164 y = 1.108x - 0.265 y = 0.987x - 0.043
R² = 0.995 R² = 0.994 R² = 0.993 R² = 0.995
2 2
2 3 4 5 6 7 8 2 3 4 5 6 7 8
WRDC GHI (kWh/m2) WRDC GHI (kWh/m2)

8
WRDC vs 1993 Nasir and Raza

KHI
7
1993 Nasir and Raza GHI (kWh/m2)

LHE

6 PWR
y = 1.149x - 1.037
QUT R² = 0.969
5
y = 1.052x + 0.149
R² = 0.968
4
y = 0.900x + 0.003
R² = 0.962
3
y = 1.016x + 0.212
R² = 0.970
2
2 3 4 5 6 7 8
WRDC GHI (kWh/m2)

Fig. 10. A comparison of monthly mean daily GHI data (kW h/m2) from WRDC versus PMD, 1986 Sukhera et al. [11], 1990 Chaudhary [16], 1993 Nasir and Raza [17], 1994 Raja and
Twidell [18], and 2006 Muneer et al. [27].

and 1200 h GMT) on the 75th Meridian East corresponding to the local
time (0800, 1100, 1400 and 1700 h PST). The spatial resolution of
approximately 24 km. The monthly mean solar insolation at 11:00 h PST
are presented in Fig. 2 to Fig. 4 of Ref. [23]. It was reported that the
maximum solar insolation falls over Balochistan due to cloud free zone.
The suitable potential sites for solar energy applications are central
Makran, the sandy desert of north-western Balochistan, the Pothohar
plateau of central Punjab, the Thal desert of southern Punjab and upper
Sindh. The data presented in this study were on the hourly basis, unlike
other studies which are daily mean of 24 hours.
Malik et al. [25] used Geostationary Operational Environmental
Satellite (GOES) cloud imagery to develop solar radiation maps of
Pakistan. The daily satellite data at 11:00 PST (0600 GMT) for the
years 1985–1986 were used to evaluate monthly and annual mean of
solar radiation using a radiative transfer model proposed by Dedieu
et al. [65]. The monthly mean hourly solar insolation at 11:00 PST for
January, March, June, August and September are presented as Fig. 1 to
Fig. 5 of Ref. [25]. The maps show that solar radiation in the country
over the year varies from 14.4 to 15.6 kW h/m2 (600–650 W/m2) in
winter to 31.2–32.4 kW h/m2 (1300–1350 W/m2) in summer. The

2848
Z.R. Tahir, M. Asim Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861

Table 3
The maximum and minimum of monthly mean daily and annual mean daily solar insolation (kW h/m2).

Source Karachi Lahore Multan Peshawar Quetta

Min Max Mean Min Max Mean Min Max Mean Min Max Mean Min Max Mean

WRDC 4.03 6.51 5.31 2.61 5.93 4.36 3.09 6.45 5.06 3.02 7.30 5.16 3.40 7.26 5.41
1986 Sukhera et al. [11] 3.83 6.55 5.20 3.04 6.67 4.76 2.58 6.56 4.98 2.68 7.05 4.99 3.17 7.95 5.68
1990 Chaudhry [16] 4.08 6.46 5.28 2.60 6.12 4.53 3.20 6.51 5.00 2.64 6.63 4.65 3.65 7.97 5.79
1993 Nasir and Raza [17] 3.69 6.36 5.06 2.78 6.34 4.74 – – – 2.40 6.44 4.65 3.60 7.65 5.71
1994 Raja and Twidell [18] 3.91 6.39 5.11 2.82 6.54 4.79 3.07 6.57 5.06 2.92 7.37 5.05 3.42 7.93 5.73
2006 Muneer et al. [27] 3.92 6.39 5.11 2.83 6.56 4.79 3.08 6.56 5.06 2.92 7.36 5.05 3.42 7.94 5.73

solar insolation data presented as monthly mean hourly for a specific
time (11:00 PST) of the day in the study, this is different from the
commonly used solar insolation data which is monthly mean daily. The
reported values in summer are very high (1300–1350 W/m2) and are
near the solar constant (1367 W/m2); therefore, these values do not
seem feasible.
Fawz-ul-Haq and Siddiqui [24] carried out a study to identify
potential sites for solar power generation in Pakistan using satellite
automatic picture transmission (APT) images from November 1988 to
October 1990. These ATP images were received from National Oceanic
and Atmospheric Administration (NOAA) series satellites and provided
information about estimates of cloud cover, cloud type, the tempera-
ture of cloud top, and temperature of land and sea surface. The
potential sites for solar power generation were identified by maximum
number of sunny days and regions with least cloud cover. The satellite
data were compared with the ground measured data provided by the
PMD, and both data agreed well. It was observed that the most areas of
Punjab, Balochistan and Sindh remain almost cloud free. The most
cloud-free month was October, whereas the cloudiest month was
February in Pakistan for the duration of the study. The results showed
that highly sunny regions are Sukkur, Nokkundi and the Bela-triangle,
lie between latitude from 26 °N to 29 °N.
3.2. NREL's solar resource assessment
The U.S. Agency for International Development (USAID), South
Asia Regional Initiative for Energy Cooperation (SARI/E) program,
promotes regional integration of energy systems and enhances Cross-
Border Energy Trade in South Asia. The SARI/E program includes
eight countries: Pakistan, India, Bangladesh, Sri Lanka, Bhutan, Nepal,
Afghanistan and Maldives. The development of high resolution solar
and wind resource data is one of the activities of the program. The
USAID under the SARI/E program contracted the National Renewable
Energy Laboratory (NREL) to carry out renewable energy resource
assessment (solar and wind) and to develop resource maps for selected
countries. The resource assessment data were in geographical informa-
tion system (GIS) format and incorporated into a geospatial toolkit
(GsT). These products for Pakistan were released at the Renewable
Energy Symposium held on June 25–26, 2007 in Islamabad, Pakistan.
The solar resource assessment was conducted in two ways; high-
resolution maps (10 km) and low-resolution maps (40 km). The high-
resolution maps were developed using the State University of New York
at Albany (SUNY) model based on Meteosat 5 geostationary satellite
data. The low-resolution maps were developed using NREL's
Climatological Solar Radiation (CSR) model based on a global 40-km
gridded cloud cover data [66].
NREL has developed the CSR Model to estimate the daily solar
radiation at a spatial resolution of 40 km [67]. The inputs of the model
are satellite and surface observations of cloud cover, aerosol optical
depth, perceptible water vapours, surface albedo, atmospheric pressure
and total column ozone sampled at 40 km resolution [68]. The solar
radiation maps for Pakistan were developed using CSR model covering
the period of seven years from 1985 to 1991. The annual mean daily
global horizontal solar irradiation (GHI), direct normal solar irradia-
tion (DNI) and global solar irradiation on flat plate tilted at latitude
(GTI), maps for Pakistan are shown in Fig. 11(a), (b) and (c),
respectively.
The Atmospheric Sciences Research Centre at the State University
of New York (SUNY) has developed maps for Pakistan at a spatial
resolution of 10 km using input data from May 1, 2002, to September
8, 2005 [70]. The SUNY model [71,72] was initially based on a model
developed for Meteosat data, which was based on a Heliosat-1 method
[73] and evolved over the years with developments being directed to
modelling GOES data. The inputs of the model are hourly visible
satellite images, climatological means for perceptible water, strato-
spheric ozone, broadband atmospheric optical depth, daily snow cover
and terrain elevation data. The annual mean daily global horizontal
solar irradiation (GHI), direct normal solar irradiation (DNI) and
global solar irradiation on flat plate tilted at latitude (GTI), maps of
Pakistan are shown in Fig. 12(a), (b) and (c), respectively.
The high-resolution SUNY model results were compared against the
CSR model and NASA SSE data set [74]. The SUNY model gave lower
values of solar radiations compared to CSR model and consistently
higher values than the NASA SSE model [66,70]. The solar resource
maps using both CSR and SUNY models have not validated against
surface measured data due to non-availability of high-quality solar
radiation data [70]. The resource maps developed by NREL were the
only available source used to investigate the application of solar energy
systems in Pakistan before 2015, but the accuracy of these maps is not
yet evaluated. A comparison of NREL's CSR model and NREL's SUNY
model with surface measured data from WRDC for six meteorological
stations is presented in Section 3.4.
3.3. World Bank's Energy Sector Management Assistance Program
(ESMAP)
The Energy Sector Management Assistance Program (ESMAP),
established in 1983, is a global technical assistance trust fund admini-
strated by the Word Bank. This program provides advisory services to
low and middle-income countries to achieve sustainable energy solu-
tions for economic growth and poverty reduction. ESMAP is executing
a global initiative to support renewable energy resource assessment
and mapping for solar, wind, small hydro and biomass covering 12
countries. The Government of Pakistan has requested the support of
the World Bank and ESMAP's Renewable Energy Resource Mapping
Initiative to help to improve the country's knowledge and awareness of
renewable energy resources such as solar, wind and biomass [75].
The three-year project of the Renewable Energy Resource Mapping
Initiative for Pakistan has three phases of solar resource assessment
[76]. Phase I was the preparation of the solar resource estimate for
Pakistan based on solar radiation modelling using satellite and
reanalysed data. The initial maps of solar and wind potential for
Pakistan were presented in a workshop in Islamabad to the
Government of Pakistan on October 15, 2014. Phase II is a 24-month
period of ground measurement at nine selected sites throughout the
country, which is still in progress. Phase III will be the establishment of

2849
Z.R. Tahir, M. Asim Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861

Fig. 11. Annual mean daily solar radiation (kW h/m2) using NREL's CSR model [69] (a) GHI (b) DNI (c) GTI on flat plate tilted at latitude.

2850
Z.R. Tahir, M. Asim
Fig. 11. (continued)
a solar atlas based on satellite data with the validation from ground
measured data.
The main output of Phase I was an estimate solar atlas (solar
resource maps) containing the annual and monthly maps of global
horizontal irradiation (GHI), direct normal irradiation (DNI) and
diffused horizontal irradiation (DHI) with a spatial resolution of
2.5 km × 2.5 km. The high-resolution data from visible and infrared
channels of geostationary satellites from 2002 to 2012 (13 years) were
used for solar resource assessment. The HELIOSAT method is one of
the several computational methods developed over the past years to
estimate the downward solar radiation from satellite data observations
and has proven to be reliable in several European projects. Cano et al.
[73] initially proposed the HELIOSAT method, subsequently modified
by Beyer et al. [77] and Hammer [78]. The downward solar irradiance
was computed by using Heliosat-2-Method [79–81] using global
datasets of water vapour, cloud index, aerosol optical depth and ozone.
The solar atlas developed using satellite data is not validated yet with
the ground measurements and will be validated in Phase III of the
project.
The annual mean total of global horizontal irradiation (GHI) for 13
years is shown in Fig. 13(a) [30]. The estimate of the annual mean total
of the GHI for the whole country is 2071 kW h/m2. The highest values
of GHI is in the south-west of the country and gradually decreases to
north and north-west of the country; the maximum annual mean daily
total is 2300 kW h/m2 in the south-west of Balochistan. The estimate
shows that 75% of the land area has over 2000 kW h/m2 and 92.6% of
the land area has over 1500 kW h/m2. The estimate of the annual mean
total of direct normal irradiation (DNI) for 13 years is shown in
Fig. 13(b) [30]. The estimate shows that 83% of the land area has over
2000 kW h/m2; the highest values are for the Sindh and Balochistan
provinces, which exceed 2400 kW h/m2. The maximum annual mean
total of DNI is about 2700 kW h/m2 in north-west Balochistan. The
estimate of the annual mean total of diffused horizontal irradiation
(DHI) for 13 years is shown in Fig. 13(c) [30]. The annual mean total of
2851
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
DHI is 650–850 kW h/m2 for the major part of the country, the highest
values of DHI being on the south coast of the country and within
Punjab province.
3.4. Comparison of satellite data with surface measured data
This section presents a comparison between the surface measured
data, the satellite resource assessment by NREL and satellite data by
NASE SSE. The freely accessible solar radiation data source for
Pakistan is WRDC, and the ground measured solar isolation data is
only available for global horizontal irradiation for six stations. The
monthly mean daily GHI values for six stations were used to compare
the results of solar resource assessments. The solar resource assess-
ment by NREL is available on their website [69] in the form of GIS
data, whereas solar resource assessment by ESMAP is not yet available.
The NASA SSE [66,70] is easily assessable solar radiation data based
on satellite model at a spatial resolution of 1°×1° latitude-longitude
with a temporal resolution of one day.
The surface measured data from WRDC were compared with the
global horizontal irradiation (GHI) values from NREL's CSR model and
NREL's SUNY. The GHI data for six meteorological stations was
obtained from the GIS data provided by NREL. The duration of the
data NREL's CSR model is from 1985 to 1991 and, for NREL SUNY
model is May 2002 to September 2005. A comparison of monthly mean
daily GHI data from WRDC with NREL CSR model and NREL SUNY
model is presented in Figs. 14 and 15, respectively. The percentage
error, mean percentage error (MPE), mean absolute percentage error
(MAPE), root mean square error (RMSE) and mean bias error (MBE)
of NREL's data were calculated using WRDC data as a reference.
The MAPE of GHI for Islamabad, Karachi, Lahore, Multan,
Peshawar and Quetta were 23.43%, 4.22%, 15.73%, 4.34%, 7.55%
and 6.80% respectively for the NREL's SUNY model and 25.22%,
2.91%, 17.80%, 6.28%, 4.32% and 6.33% respectively for the NREL's
CSR model. The MPE of solar insolation for Islamabad, Karachi,
Z.R. Tahir, M. Asim Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861

Fig. 12. Annual mean daily solar radiation (kW h/m2) using NREL's SUNY model [69] (a) GHI (b) DNI (c) GTI on flat plate tilted at latitude.

2852
Z.R. Tahir, M. Asim
Fig. 12. (continued)
Lahore, Multan, Peshawar and Quetta were −23.43%, 2.74%, −15.73%,
−1.94%, 5.64% and −6.24% respectively for the SUNY model and
−25.22%, +0.18%, −17.80%, −6.23%, +3.02% and −5.76% respectively
for the CSR model. The MBE (in kW h/m2) of GHI for Islamabad,
Karachi, Lahore, Multan, Peshawar and Quetta were −1.00, +0.13,
−0.72, −0.14, +0.22 and −0.39 respectively for the SUNY model and
−1.03, 0.00, −0.79, −0.34, +0.12 and −0.35 respectively for the CSR
model. The RMSE of GHI for Islamabad, Karachi, Lahore, Multan,
Peshawar and Quetta were 1.18, 0.26, 0.80, 0.26, 0.40 and 0.50
respectively for the SUNY model and 1.14, 0.17, 0.84, 0.41, 0.23 and
0.46 respectively for the CSR model.
The monthly mean daily GHI values from the low-resolution CSR
model are higher compared to the high-resolution SUNY model, and
both are higher than WRDC data. Both SUNY and CSR models
overestimate the solar insolation compared to the ground measure-
ments of WRDC. The SUNY model overestimates the annual mean
values by 23.43%, whereas, in July the overestimate is 40.39% for
Islamabad. The CSR model overestimates the annual mean values by
25.22%, whereas, in January and May, the overestimates are 35.72%
and 35.05%, respectively for Islamabad. For Lahore, the SUNY model
overestimates the annual mean values by 15.73%, whereas, in July the
overestimate is 24.63%; the CSR model overestimates the annual mean
values by 17.80%, whereas, in January and September the over-
estimates are 23.28% and 22.03%, respectively. The SUNY model
underestimates the mean monthly mean solar insolation for Karachi
and Peshawar by a maximum of 7.81% and 16.05% respectively and
overestimates for Multan and Quetta by a maximum of 7.89% and
12.89% respectively. The CSR model underestimates the mean monthly
mean solar insolation for Karachi and Peshawar by a maximum of
4.66% and 14.00% respectively and overestimates the for Multan and
Quetta by a maximum of 12.62% and 13.94% respectively. The SUNY
model gave a good estimate (MAPE less than 5%) of the solar insolation
for Multan, whereas the CSR model gave a good estimate for Karachi
2853
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
and Multan. The maximum and minimum values of monthly mean
daily, along with the annual mean daily global horizontal irradiation of
six meteorological stations from WRDC, NREL's CSR model and
NREL's SUNY models for the duration as described in the above
sections, are presented in Table 4.
The National Aeronautics and Space Administration (NASA) have
developed the Solar Meteorology and Solar Energy (SSE) service, which
includes the long-term estimates of meteorological quantities and
surface solar energy fluxes. The NASA-SSE data sets include mean
daily solar radiation from July 1, 1983, to June 30, 2005. This data set
is available worldwide and has a spatial resolution of 1°×1° latitude-
longitude and is produced via the International Satellite Cloud
Climatology Project (ISCCP). The SSE Release 1 was made available
to the public in 1997 to provide easy access to parameters needed in
the renewable energy (solar and wind) industry. The Prediction of
Worldwide Energy Resource (POWER) project was started in 2003 to
improve previous releases of SSE and to create new datasets from new
satellite observations. The surface solar insolation in SSE data was
estimated using a modified method of Pinker and Laszlo [82], which
involves the use of a radiative transfer model. The inputs of the model
include visible and infrared radiances, cloud and surface properties,
temperature and moisture profile and column ozone amount [74]. The
SSE solar radiation data were validated against research quality
observation from the Baseline Surface Radiation Network (BSRN) [83].
The NASA-SSE data set is available 1° spatial resolution (approx-
imate 100 km) in the form of the grid with a temporal resolution of one
day. The NASA-SSE solar radiation data are available in three common
units used: daily mean of global horizontal irradiance in W/m2 and
daily mean global horizontal irradiation in Wh/m2 and J/cm2. The
mean daily data for solar insolation for the six stations were obtained
from NASA-SSE for the duration July 1, 1983, to June 30, 2005, for the
nearest location on the grid. The monthly mean daily solar insolation
(kW h/m2) was calculated for the long-term data (22 years) for
Z.R. Tahir, M. Asim Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861

8
NREL CSR vs WRDC

7
ISB

NREL CSR GHI (kWh/m2)

KHI
6
LHE
y = 1.221x - 0.176
MUL R² = 0.982
5
PWR
y = 1.155x - 0.444
QUT R² = 0.989
4
y = 1.410x - 0.587 y = 1.041x - 0.334
R² = 0.971 R² = 0.984
3
y = 1.056x - 0.299 y = 1.175x - 0.592
R² = 0.959 R² = 0.988
2
2 3 4 5 6 7 8
WRDC GHI (kWh/m2)
Fig. 14. A comparison of monthly mean daily GHI data (kW h/m2) from WRDC versus
NREL CSR model.

8
NREL SUNY vs WRDC

7
NREL SUNY GHI (kWh/m2) ISB
KHI
6
LHE
y = 1.136x - 0.553
MUL R² = 0.985
5
PWR
y = 1.265x - 0.436
QUT R² = 0.979
4
y = 1.540x - 1.129 y = 1.097x - 0.719
R² = 0.966 R² = 0.966
3
y = 1.067x - 0.492 y = 1.195x - 0.669
R² = 0.935 R² = 0.986
2
2 3 4 5 6 7 8
WRDC GHI (kWh/m2)
Fig. 15. A comparison of monthly mean daily GHI data (kW h/m2) from WRDC versus
NREL SUNY model.

Table 4
The maximum and minimum of monthly mean daily and annual mean daily GHI values
(kW h/m2).

Meteorological station WRDC NREL SUNY NREL CSR

Islamabad Min 2.23 2.56 2.63

Max 5.42 7.20 7.21
Mean 3.95 4.95 4.98
Karachi Min 4.03 3.82 3.96
Max 6.51 6.63 6.64
Mean 5.31 5.17 5.30
Lahore Min 2.61 2.78 2.91
Max 5.93 7.13 7.14
Mean 4.36 5.08 5.15
Multan Min 3.09 2.89 3.15
Max 6.45 6.95 7.19
Mean 5.06 5.20 5.40
Peshawar Min 3.02 2.53 2.59
Fig. 13. Annual total of daily solar radiation (kW h/m2) for 13 years duration from 2002 Max 7.30 7.21 7.30
to 2012 [30] (a) GHI (b) DNI (c) DHI. Mean 5.16 4.94 5.04
Quetta Min 3.40 3.35 3.28
Max 7.26 7.89 7.84
Mean 5.41 5.80 5.77

2854
Z.R. Tahir, M. Asim Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861

8 good agreement with the surface measured data for meteorological

NASA SSE vs WRDC
stations with less cloudy conditions. The NASA SSE data are easily
7 available for download from the website so that this data can use for
ISB the assessment of the site for the application of solar energy in
NASA SSE GHI (kWh/m2)

KHI
6
LHE
MUL
5
PWR
QUT
4 energy. Section 4.1 is a short review of the articles on renewable
y = 1.431x - 0.418
R² = 0.974
3 focus of these articles was the utilisation of renewable energy to
y = 1.049x - 0.125
R² = 0.956
2 in Section 4.2; however, solar energy potential has not been evaluated
2 3 4 5 6 7
WRDC GHI (kWh/m2)
Fig. 16. A comparison of monthly mean daily GHI data (kW h/m2) from WRDC versus
NASA SSE.
comparison with surface measured data. A comparison of the surface
measured data from WRDC and satellite data by NASA SSE for six
meteorological stations in presented in Fig. 16. The equation of the
lines of the best fit for each station and R2 values are also presented.
The MAPE of solar insolation for Islamabad, Karachi, Lahore, Multan,
Peshawar and Quetta were 32.01%, 3.35%, 21.99%, 3.65%, 4.02% and
4.42% respectively. The MPE of solar insolation for Islamabad,
Karachi, Lahore, Multan, Peshawar and Quetta were −32.01%,
−2.44%, −21.99%, 2.28%, −0.58% and −2.28% respectively. The
MBE of solar insolation for Islamabad, Karachi, Lahore, Multan,
Peshawar and Quetta were −1.28, −0.13, −0.95, +0.15, −0.05 and
−0.14 respectively. The RMSE of solar insolation for Islamabad,
Karachi, Lahore, Multan, Peshawar and Quetta were 1.38, 0.22, 0.99,
0.27, 0.25 and 0.28, respectively.
A comparison of surface measured data by WRDC with the resource
assessment by NREL using both CSR and SUNY models and satellite
data from NASE SSE shows similar results. The mean absolute
percentage error of GHI data from NREL's CSR, NREL's SUNY and
NASE SSE compared with WRDC is shown in Fig. 17. The satellite
models overestimate the solar radiation more than 23% for Islamabad
and more than 15% for Lahore, whereas, for the other four stations it is
less than 7%. One of the main reasons for the overestimation of satellite
data is that the radiative transfer model often predicts clear sky
conditions whereas actual conditions are cloudy; the opposite of this
observation is also possible, but that is less obvious when actual clear
sky conditions are predicted as cloudy [84]. This inappropriate
prediction of the occurrence of clear sky conditions leads to an
overestimation of satellite solar radiation data. The satellite data gave
35
NREL CSR NREL SUNY
30
25
MAPE (%)
Pakistan.
y = 1.214x + 0.018 4. Renewable energy potential and solar energy applications
R² = 0.988
This section presents the review of the literature published on the
y = 0.863x + 0.544
R² = 0.977 renewable energy potential of Pakistan, especially focus on solar
y = 1.058x - 0.244 energy (solar, wind, biomass and hydel energy) potential, the main
R² = 0.979
overcome the electricity shortage in the country. A review of articles
y = 1.082x - 0.299
R² = 0.979 describing the solar energy potential (mentioned in these) is presented
8
in any of these articles and solar energy potential described in Section
2.2 and Section 3.2 is only referred to in these articles. Section 4.3
contains the review of the status of solar energy applications (solar PV
and solar thermal), their feasibility and utilisation in the country.
Section 4.4 is about the analysis and critical discussion on the articles
presented in this Section 4. The importance and applications of the
surface measured data are presented in Section 4.5.
4.1. Renewable energy resources potential and utilisation
Energy is the basic necessity for life and a key factor to improve the
economy of a country. Pakistan despite being richly blessed with
renewable energy resources is dependent on foreign sources (for crude
oil, LNG and coal) and is currently facing severe energy deficiency in
the power sector. The electricity sector in the country is currently facing
the tough challenges of an insufficient installed capacity, a suboptimal
infrastructure, circular debt and revenue shortage. The renewable
energy resources (solar, wind, biomass, hydel) potential and it utilisa-
tion for Pakistan had been reported by many articles [85–95]. Sahir
and Qureshi [88] in their review article on the assessed renewable
energy resources potential and limitations of their use in the country
reported that; solar, wind, biomass and micro-hydel can play an
important role as supplementary resources but not as an alternative
to conventional resources. Sheikh [91] presented a review of conven-
tional energy and the renewable energy scenarios of Pakistan in
quantitative terms of supply, generation and exploitation of available
resources. An overview of the energy status and development of
Pakistan was presented by Rauf et al. [96], including information
about energy generation and distribution institutions, total installed
power capacity, total generation power capacity, energy reserves and
resources of the country, energy crises in the power sector and possible
solutions to overcome the energy crisis. A study to diagnose the energy
crisis and the potential and development of alternative energy projects
was reported by Shaikh et al. [97], it was recommended that the solar
power projects are quick remedies to reduce the shortfall of electricity
NASA SSE and may be developed within one to three years.
The availability of renewable energy resources (hydel, solar, wind
and biomass) and their use for meeting growing electricity demands
reported in [87] with the conclusion that substantial renewable energy

20 potential is available for growing demand for electricity. An assessment

of renewable energy potential for electricity generation in Pakistan
15 presented by Farooq and Kumar [94], the current and future potential
of renewable energy sources was estimated. The wind energy genera-
10
tion potential was estimated as the capacity installation of 13 GW for
5 the suitable available area, the biomass energy source potential could
be 15 GW in 2050. The solar energy potential for power generation
0 from solar photovoltaic and solar thermal technologies was estimated
Islamabad Karachi Lahore Multan Peshawar Quetta to be 169 GW in 2050. A survey of the availability of various renewable
Fig. 17. Mean absolute percentage error of GHI data from NREL's CSR model, NREL's energy sources (hydel, solar, wind and biomass) and their current and
SUNY model, and NASA SSE compared with WRDC. future prospects in the total energy mix had been carried out by

2855
Z.R. Tahir, M. Asim
Farooqui [95] in his review article, it was estimated that the country
has feasible potential of 30 GW and 50 GW of installed capacity of
hydel and the wind respectively by 2030. To ensure a sustainable
energy future, country need to diversify its energy mix recommended
by Asif in his review article [89], only 15% of available 42 GW of hydel
power potential have been exploited so far, a high level of available
solar energy can be utilised for both solar thermal and solar electrical
applications.
4.2. Solar energy potential
The solar energy potential in terms of annual mean daily and
annual total global horizontal insolation estimated from surface
measured data is reported in few articles. The daily mean global
horizontal insolation for the whole country is reported as 4.8–6.0
kW h/m2 (200–250 W/m2) in [90,94,98–100], 4.16–5.83 kW h/m2
(15–21 MJ/m2) in [86], 4.44–5.83 kW h/m2 (16–21 MJ/m2) in [87].
The annual mean daily GHI for the whole country is 5.28 kW h/m2 (19
MJ/m2) [87,99,101] and major area receive GHI between 5.27 and
5.56 kW h/m2 (19–20 MJ/m2) [86]. According to the Energy
Information Administration, U.S. Department of Energy [102], the
daily mean solar insolation for the country is 5.3 kW h/m2[98]. The
annual total of global solar insolation for the country is approximately
1900–2300 kW h/m2 (6840–8280 MJ/m2) [103] reported in
[90,98,99] whereas 1900–2200 kW h/m2 [104] reported in Asif
[89,100,105].
The solar energy potential in terms of annual mean daily global
horizontal insolation estimated from the solar resource assessment
[70] by National Renewable Energy Laboratory, USA is reported in five
articles [91,92,94,106,107]. The annual mean daily solar global
horizontal insolation for the whole country was reported as 5–7
kW h/m2 (18–25 MJ/m2) approximately in a review article by
Sheikh [106]. Average daily mean solar global insolation of 5–7
kW h/m2 exists over more than 95% area of the country, with a
persistence factor of over 85% [91,92,94,106].
The total available annual potential of solar energy over the total
area (796,095 km2) of the country is 1453 GW h (5230 TJ) based on
annual mean daily GHI of 5.27 kW h/m2 (19 MJ/m2) [21,87], this
estimation was based on monthly maps developed from the measured
surface data of 15 years duration [21]. The exploitable solar energy
resource potential of the country is 50 GW [96], the estimated solar
energy potential in the country was reported as 2900 GW [105,108]
and 2900 GW h [109] according to the Planning Commission of
Pakistan (the details of the source of this data is not mentioned).
The sunshine duration monthly counter maps using measured data
from 1975 to 1989 was presented by Ghaffar [21], the annual mean
sunshine duration is 8–10 h for major parts of the country except for
the northern areas; 7–8 h during winter (December to February),
whereas, 9–10 h in summer (May to August). The annual sunshine
duration for Bahawalpur (Punjab) was reported about 3000 h [10]. The
sunshine duration reported by Adnan et al. [22] ranges from 12 to 14 h
in summer and from 8 to 10 h in winter. The sunshine duration for the
whole country are 1500–3000 h in a year, particular Baluchistan
province has an annual mean sunshine duration of 8–8.5 h [98]. The
mean daily sunshine duration from 6 to 8 h and annual total sunshine
duration from 2300 to 2700 h estimated from NREL's resource
assessment was reported in [91,92,106]. The mean daily sunshine
duration is from 7 to 8 h in the south-western part of Balochistan
Province and north-eastern part of Sindh province [91], both regions
has excellent potential for solar energy. The average sunshine duration
of 7.6 h was reported in [101,107].
4.3. Solar energy applications
The availability of solar radiation is relatively consistent throughout
the country, which makes it highly favourable for solar energy
2856
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
applications like photovoltaics, solar water heating, solar desalination
and solar crop drying [89]. A review of the status of solar energy
application (photovoltaics, solar thermal, solar water heaters, solar
cooker and solar desalination) is presented by Mirza et al. [98], it was
reported that about seven solar stations with total capacity of 234 kW
was installed in mountain areas of Hindu Kush Himalayas region, more
than 2000 solar cookers has been used in the same region. A recent
review on the prospects of solar energy utilisation, potential applica-
tions of solar energy, and public sector organization working to
promote renewable energies in presented by Bhutto et al. [99]. The
utilisation of solar energy began in Pakistan in the 1980s. Pallet and
Brabben [9] reported the installation and operation of twenty solar
powered micro-irrigation units (solar water pumps) at research and
private agriculture farms throughout the country with the support of
the Pakistan Agriculture Research Council (PARC) and Agriculture
Development Bank of Pakistan (ADBP). The Directorate General of
Energy Resources, Ministry of Petroleum and Natural Resources
started a program of solar villages, and the first solar energy system
was installed in 1981; this solar village had a power capacity of 5 kW
and the electric power was utilised for domestic purposes and water
pumping [10]. The utilisation of solar energy for different applications
had been started by Pakistan Council for Scientific and Industrial
Research (PCSIR) and Directorate General of New and Renewable
Energy Resources (DGNRER) [19]. DGNRER had installed solar PV
systems at 18 locations all over the country with the total capacity of
434 kW, details in Table 2 of Ref. [86]. PCSIR had designed, developed
and installed five solar desalination plants, four having a capacity of
250 gpd (gallons per day) each in Therparker and one having a capacity
of 6000 gpd in Gwadar [86].
4.3.1. Solar photovoltaic based applications
Solar photovoltaic power is a commercially available technology
with a significant potential for long-term growth, there is rapid
development occurring both in the basic technology and the market
strategy and prospects for rapid growth of solar power. The high levels
of solar radiation in major parts of the country throughout the year
make favourable conditions for its development [99]. The minimum
level of annual mean daily solar horizontal insolation, 4.44 kW h/m2
(16 MJ/m2), in the country is higher than the World mean, 3.61 kW h/
m2 (13 MJ/m2) [87]. The estimated annual solar PV potential for
centralised and decentralised grid-connected systems is 3500 TW h
and 25 TW h respectively. [87]. The limited technological awareness of
the solar photovoltaic systems in the country [110] is one of the major
issues in the utilisation of its potential.
The utilisation of renewable energies especially solar PV as a viable
source has got more attention in the last decade due the shortage of
electricity in the country, the prospects of solar PV for meeting the
growing electricity demand is various articles [3,87,94,110,111]. A
recent review on the progress of photovoltaic-based energy generation
in Pakistan was presented by Sher et al. [3]. The Pakistan Engineering
Council has installed 178 kW capacity PV plants in both the Pakistan
Engineering Council and Planning Commission of Pakistan sites in
Islamabad, with the total power generation capacity of 356 kW.
According to the Punjab Energy Department, private entrepreneurs
are working on PV power plants in various regions, with a total capacity
of 90 MW.
The performance and economic analysis of Solar PV system were
reported in Refs [21,106,112]. The annual mean of the extractable
solar energy using PV cell array with 12% assumed efficiency for four
cities was estimated and reported in [21]; the annual total solar energy
for Lahore, Multan, Islamabad and Hyderabad was 163.89 kW h/m2
(590 MJ/m2), 166.67 kW h/m2 (600 MJ/m2), 161.11 kW h/m2 (580
MJ/m2) and 166.67 kW h/m2 (600 MJ/m2), respectively. The perfor-
mance analysis of the solar photovoltaic system of 10 MW capacity
based on NREL data had a capacity factor of 20% without storage for
Multan, the Levelized Cost of Electricity for the system was found as
Z.R. Tahir, M. Asim
0.192 USD/kW h [112]. The estimate of photovoltaic electric energy
generation over an area of 10×10 km2 (with 14% efficient PV panels
and 0.7 system performance ratio) is 8.33×103 GW h based on annual
sunshine hours of 2125 and average irradiance of 800 W/m2 (average
irradiance during sunshine hours) reported in [106]. The total esti-
mated demand of the country is 62 MTOE (Million Tonnes of Oil
Equivalent), whereas 100 km2 area of land covered with PV panels can
produce energy equivalent to 30 MTOE (this information is not
sensible, as 30 MTOE is equivalent to 349 TW h). Similar calculations
were reported by Kurokawa [113] for the Thar desert, an area of
20000 km2 which can generate 21.3×103 TW h of electricity annually.
A proposal for solar hydrogen energy system as an alternative
option for fossil fuel-based energy systems was presented in [114]
presented. It was proposed that hydrogen gas can be produced through
electrolysis using the photovoltaic system in the non-agriculture sunny
terrain of Balochistan. It is reported that major regions throughout the
country receive on average daily mean solar insolation of 5.26 kW h/
m2 (19.0 MJ/m2) [14]. Lutfi and Veziroǧlu [115] proposed solar-
hydrogen demonstration project for a potential location. The proposed
location of the project was Sonmiani (a coastal area in Balochistan,
85 km north-west of Karachi) where the annual total of daily mean
global solar insolation is 1921 kW h/m2 (7.0 GJ/m2).
4.3.2. Solar thermal applications
The application of solar energy for the solar thermal applications
has been reported in few articles [27,86,112,116]. The solar thermal
energy is abundant and economically feasible compared to conven-
tional energy resources, solar thermal energy is cost competitive due to
simpler design and fabrication [116]. The first ever solar water heater
was developed by DGNRER had a capacity of 50 gallons which gave the
temperature of 40–45 °C supplied at subsidised price [86]. The
experimental analysis of solar water heaters for the textile industry
was reported in [27]. A performance analysis of built-in storage water
heater (plain and finned type) show that finned type water heater has
55% efficiency compared to plane type heater which was 47%. The
monetary energy payback period for plain and finned type heaters are
6.7 and 6.1 years respectively. The performance analysis of parabolic
trough collector for Multan was reported by Sadati et al. [112]. The
typical meteorological year (TMY) data from NREL and TMY2 data
from Meteonorm were analysed; the estimated annual DNI values
based on NREL and Meteonorm were 1828 kW h/m2 and 1363 kW h/
m2, respectively. The performance analysis based on NREL data show
that parabolic trough collector system with and without storage having
capacity factor of 46% and 20% respectively. The levelised cost of
electricity for a parabolic trough collector system was estimated as
0.273 USD/kW h for a system without storage and 0.226 USD/kW h
with 7.5 hours of storage.
4.4. Critical discussion and analysis
The published research data is quite old now, as most of the data
used in from 50's till 90's. Currently, NREL [29] and IEA [49] and
WMO [117,118] suggested guidelines for collection, assessment and
use of solar energy resource data for improved accuracy and reliability.
Some information observed during review of the published literature
which is not feasible (due to errors in data processing and analysis) and
is misleading for the readers. The data presented in articles before the
90 s and NREL solar resource assessment were cited in recent
publications. No recent study has been carried out to investigate the
accuracy and validity of solar potential of Pakistan. The solar radiation
data in NREL assessment is overestimated which lead to lower capacity
factor of installed solar system increasing the operational cost and
lowering the confidence of the investors.
Ghaffar [21] reported the available solar energy potential of the
total geographical area of Pakistan was reported as 5.23 PJ/m2 per year
which is a calculation error (the correct value is 5.23 million PJ/m2),
2857
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
the same potential was reported by Uqaili et al. [87] without noticing
the calculation error. Sheikh [106] reported that the 10×10 km2 area of
land covered with PV panels (14% efficiency) can produce energy
equivalent to 30 MTOE (equivalent to 349×103 GW h). This informa-
tion is misleading, as the described area can produce 8.33×103 GW h,
which is equivalent to 0.7 MTOE only. Sheikh [91,106] reported that
the daily mean solar insolation varies from 5 to 7 kW h/m2 (18–25 MJ/
m2) without any reference, the upper value is much higher than the
measured values reported in Section 2.2. Amer and Daim [92] reported
that the daily mean solar insolation varies from 5 to 7 kW h/m2,
according to NREL's solar maps of Pakistan, However, there is no
information on the range of solar insolation provided by NREL's
published reports [66,70], while, in discussing solar PV applications,
the DNI map was referred to, which is, again, misleading the readers as
DNI is not applicable for solar PV systems.
Awan and Khan [105] reported the estimated solar potential of the
country as 2900 GW, without any calculations or valid reference. The
GHI map is shown in Fig. 4 of [105] and DNI map shown in Fig. 7 of
[105] are taken from SolarGIS [26] but are referred to as an NREL map
in the description, which is confusing and misleading for the reader.
Mahmood et al. [109] reported the estimated potential of solar energy
in Pakistan is 2900 GW h without any reference or calculations. Khalil
and Zaidi [111] reported that a 1 kW photovoltaic system could
produce an average of 0.17 kW per day for the entire country, but this
statement makes no sense; five cities were selected for analysing the
performance of a PV system, which is not representative of the country.
The population of Lahore city was reported as 90 million, which is
about half of the population of the country, electricity consumption in
different sectors was described as energy consumption, and the names
of the programs used for simulation were not correctly named. Rauf
et al. [96] reported an exploitable solar energy resource potential
greater than 50 GW without calculations or reference. Ghafoor et al.
[108] reported the solar energy potential of the country as 2.9 million
MW based on NREL without describing details, whereas NREL maps
show much higher potential.
Khahro et al. [51] estimated diffused, direct and total solar
insolation from measured global horizontal insolation for a prospective
location in the southern region of Sindh. The annual mean daily
insolation was reported as 7.91 kW h/m2 (28.47 MJ/m2). The values
reported throughout the article are not feasible, and there seems to be a
systematic error in processing the raw data for the given location; the
values reported for base calculation are 77% higher than the actual
measured values, which is misleading. The NASA SSE [74] annual
mean daily GHI for this location is 5.16 kW h/m2 with a maximum
monthly mean daily value of 6.69 kW h/m2 in June. The data was
measured at AEDB-UNDP Baburband Mast, the raw data for this
location was reanalysed for the present study, and an annual mean GHI
was found as 4.46 kW h/m2 (equivalent to daily mean insolation of
16.11 MJ/m2 or 186 W/m2).
4.5. Importance and application of surface measured data
Accurate and precise solar resource assessment of the potential site
is a prerequisite for the successful deployment of any solar energy
system. The accurate and reliable data of various components of solar
radiation are required for various applications. The global horizontal
irradiation (GHI) is required for photovoltaic systems, and direct
normal irradiation (DNI) is required for concentrated solar power
systems [49]. The solar resource assessment plays a major role in the
success of the solar energy project, from the initial site selection to
plant design and financing [29]. The solar radiations maps
[12,14,16,19–21] discussed in Section 2.2 are not enough reliable for
the solar resource assessment of the potential site, as those were
predicted on the basis of the theoretical formula without considering
the environmental conditions, like the cloud cover. The solar resource
assessment by NREL is the only assessment for three components of
Z.R. Tahir, M. Asim
solar radiation; GHI, DNI and GTI. This assessment is based on the
satellite data and is not validated against the surface measured data,
since its launch in 2007. The comparison of this assessment is done for
the first time in this study, and it is observed that the GHI values
estimated by NREL are much higher than the measured values for
some regions in the country. The comparison performed in this study is
a question mark on the reliability of the solar resource assessment by
NREL to use for commercial solar power projects.
The high-quality surface measured data for solar radiation are not
usually available on the potential site, to ensure proper solar assess-
ment and thus to secure the final acceptance of the project. The annual
variation of the solar radiation has a major effect on the prediction of
energy yield of the solar system [49]. The long-term time series of solar
radiation data (usually multi-decadal) are required at different stages
of the solar project [119]. The multi-decadal solar radiations time
series are seldom available for potential sites of solar power system, so
solar radiation can be estimated using satellite imaginary to derive the
effect of cloud cover on the solar radiation at the surface of Earth.
There are various methods to determine the solar radiation at the
surface of the earth and most commonly used are based on the
meteorological satellites data [79,80,120–123]. The global reanalysis
datasets of solar radiation were estimated by different radiative
transfer models and are available free of cost. The most common
global reanalysis datasets used for satellite solar resource assessment
are: NASA Modern-Era Retrospective Analysis for Research and
Application (MERRA) [124]; European Centre for Medium-Range
Weather Forecasts reanalysis (ERA-Interim) [125]; the joint product
of the National Centres for Environmental Predictions (NCEP) and the
National Centre for Atmospheric Research (NCAR) reanalysis (NCEP-
NCAR) [126]; Climate Forecast System Reanalysis and Reforecast
(CFSRR) [127]; Japanese 55-year Reanalysis Project (JRA-55) [128];
and Twentieth-Century Reanalysis (20CR) [129]. The time series from
these satellite datasets span over decades with spatial resolutions from
0.5° to 2.0° and temporal resolutions from one hour to six hours. The
time series of solar radiations from these reanalysis data have an
absolute bias for a specific region [130–132] and are corrected by bias
corrections method to compare with surface measured data.
The radiative transfer model used to estimate solar radiation global
in a particular reanalysis dataset is compared with high-quality surface
measured data and corrected for the specific region. In the case of
Pakistan, the reanalysis dataset can be compared with the ground
measured data available to find the bias for a specific region. The bias
correction approach may be applied to the time series for the potential
site to find solar resource potential. According to the Alternative
Energy Development Board [5], 28 solar PV power projects of the
total capacity of 956.8 MW are at various stages of development, and
by the year 2018, it will be 1556 MW. Out of 28 projects, two projects
of 100 MW capacity each are installed by Quaid-e-Azam Solar Power
Ltd, two projects of 100 MW capacity each are under construction and
24 projects of 556.8 MW are in the pipeline. Pakistan has huge solar
energy potential, so surface measured data available at WRDC can be
used to correct satellite reanalysis data for the different regions
throughout the country for the solar resource assessment. The satellite
reanalysis dataset can be compared with this surface measured data to
find possible bias and can be used for solar resource assessment of
solar energy projects in Pakistan.
5. Discussion
The Pakistan Meteorological Department (PMD) began measuring
data for global horizontal insolation for five stations from 1957 and
stopped somewhere around 2000. The World Radiation Data Centre
(WRDC) collects and archives global solar radiation data. The long-
term solar global insolation and sunshine duration data for Pakistan
can be obtained from PMD and WRDC; PMD charges for providing the
data, whereas WRDC provides data free of charge. The WRDC solar
2858
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
insolation data for the long-term duration were compared with the
PMD data in Section 2.4, and both datasets show good agreement,
except there is some variation due to the duration of data, which is
acceptable; therefore, WRDC data can be used for further solar
resource assessment. The WRDC data were not used or even referred
to in any of the articles presented in this review, which is unfortunate
and reflects that researchers in the field of solar energy are not aware of
WRDC or availability of long-term surface measured data. The majority
of the work for estimating solar potential was done before 2000, about
three times more articles were published after 2000, and only a few had
done any research work on solar potential, whereas the majority of
articles referred to previous work.
The first attempt to estimate the annual mean daily solar isolation
of whole country from surface measured data of solar insolation and
sunshine duration using Ångstrom type regression equation was made
by Raja and Twidell in 1990 [13,14], which was 4.44–5.97 kW h/m2
16.0–21.5 MJ/m2) with mean of 5.27 kW h/m2 (19.0 MJ/m2). Ghaffar
[21] estimated the annual mean daily solar insolation as 4.44–5.83
kW h/m2 (16–21 MJ/m2) with a mean of 5.27 kW h/m2 (19 MJ/m2).
Ahmad et al. [20] estimated the annual mean daily solar isolation using
ASHRAE formulation modified for Pakistan, which was 4.80–6.20
kW h/m2 with a mean of 5.50 kW h/m2. Adnan et al. [22] estimated
the annual daily mean solar insolation using sunshine duration data for
58 meteorological stations, which was 3.6–6.6 kW h/m2 (equivalent
irradiance of 150–275 W/m2) with a mean of 5.11 kW h/m2 (213 W/
m2). Mirza et al. [98] reported that the daily mean global horizontal
irradiation was 4.80–6.00 kW h/m2 (equivalent irradiance of 200–
250 W/m2) according to the World Energy Council [103], and the daily
mean solar insolation for Pakistan was 5.30 kW h/m2 (the annual total
of daily mean is 1930 kW h/m2) according to the Energy Information
Administration, U.S. Department of Energy [102].
The NREL published literature [66,70] did not include any in-
formation about the monthly mean daily solar insolation or even total
solar potential of Pakistan, along with the fact that the solar atlas was
not previously validated by ground surface measured data, and was
used by many researchers to estimate the solar potential. The mean
daily insolation estimated by NREL's SUNY model and CSR model are
significantly higher than then long-term surface measured data of
WRDC. The annual total of mean daily global horizontal irradiation
estimated by ESMAP's solar energy resource maps (based on 13 years
satellite data from 2000 to 2012) for the whole country was 2071
kW h/m2 and the monthly mean daily GHI was 5.67 kW h/m2 .
The annual mean daily solar insolation for the whole country,
estimated from measured data of solar insolation and sunshine
duration using empirical relationships, was from 5.11 to 5.50 kW h/
m2, which is slightly lower than the ESMAP's solar atlas, which is 5.67
kW h/m2. The solar insolation based on satellite data is overestimated
due to clear sky prediction while actual conditions were cloudy [84].
The annual mean daily solar insolation for all of Pakistan, based on the
PMD data [13,14,20,22,98], is estimated from 4.4 to 6.2 kW h/m2 with
a mean value of 5.30 kW h/m2 (19.0 MJ/m2 or 221 W/m2).
Ghaffar [21] reported that the total available annual potential of
solar energy over the total area (796095 km2) of the country is 5230 TJ
(1453 GW h). Awan and Khan [105] and Mahmood et al. [109]
reported the estimated solar potential of the country as 2900 GW
and 2900 GW h, respectively, according to the Planning Commission of
Pakistan. The same potential value of 2900 GW was recently reported
by the International Finance Corporation, the World Bank Group, the
World Bank Group [133] and in [108]. The reported values of solar
resource potential are not sensible in any respect; the 1453 GW h value
reported in Ref. [21] using the total land area of the country, which is
about one million times less than the calculated value, the 2900 GW
and 2900 GW h values reported in Ref. [105,108,109] have no
information about the area used to calculate potential and no calcula-
tions were found in the previously published literature.
The solar resource potential based on annual mean daily solar
Z.R. Tahir, M. Asim
insolation of 5.30 kW h/m2 for the total land cover area (796,095 km2)
of the country is estimated as 4219 TW h (terawatt hour or million
MW h) equivalent to 175,800 GW (Gigawatts). The utilisation of just
0.01% of this potential to generate electricity is approximately equal to
the current generation capacity, which is about 17 GW.
6. Conclusions
More than 70 articles, the majority of them published in peer-
reviewed international journals, related to renewable energy especially
solar were reviewed. The surface measured data for solar insolation for
five meteorological stations was reported from the 1980s. The major
research work to estimated solar resource potential from the surface
measured data of solar insolation and sunshine durations was done
before the year 2000. The research work based on the estimation of
solar potential after 2000 almost came to a halt because the measure-
ment of surface data ceased. The estimation of solar resource potential
based on long-term satellite data was not done on the national level.
About 30 articles after the year 2000, based on renewable potential,
were reviewed and no significant work was done for the estimation of
solar potential; almost all the authors referred to previously published
work.
The long-term measured data for global solar horizontal insolation
for six meteorological stations ((Islamabad, Karachi, Lahore, Multan,
Peshawar and Quetta) can be obtained from PMD or WRDC, the latter
free of charge. A comparison of the measured data of monthly mean
daily solar insolation presented in the Ref. [11,16–18,27] for different
durations with long-term WRDC data are in good agreement, except for
some variation due to different durations. The trend in the monthly
mean value of GHI is the same for each location so that WRDC can be
used for further assessment of the solar potential of Pakistan.
The annual mean daily solar insolation for all Pakistan, based on
the Refs. [13,14,20,22,98], is estimated, on average, from 4.4 to 6.2
kW h/m2 with a mean of 5.30 kW h/m2 (19.0 MJ/m2 or 221 W/m2).
The solar resource potential based on annual mean daily solar
insolation of 5.30 kW h/m2 for the total land cover area
(796,095 km2) of the country is estimated as 4219 TW h (terawatt
hour or million MW h), equivalent to 175,800 GW (Gigawatts).
Not a single attempt has been made to estimate solar resource
potential using long-term satellite data. NREL developed the solar
resource assessment maps for Pakistan under the SARI/E program of
USAID. These maps were not compared with the ground surface
measured data and a comparison of monthly mean daily GHI from
these maps with WRDC data show that these maps overestimate the
GHI values by up to 25%. The ESMAP of the World Bank has developed
solar resource maps for Pakistan in 2015 under a three-year project
which will complete in 2017. The solar atlas will be validated at the end
of 2017 with ground surface measured data. The annual mean daily
GHI estimated in ESMAP's solar atlas for the whole country is 5.67
kW h/m2, which is 7% higher than the value estimated from the surface
measured data.
Acknowledgements
The authors would like to express their gratitude to Faiz ul Rehman
Tahir of National Renewable Energy Services Pakistan for providing
the technical support. The authors thanks, Dr Shahid Imran of
Department of Mechanical Engineering, KSK Campus, University of
Engineering and Technology Lahore for their valuable suggestions.
References
[1] IFC. A solar developer’s guide to Pakistan. The international finance corporation.
World Bank Group; 2016.
[2] Pasha HA, Saleem W. The impact and cost of power load shedding to domestic
consumers. Pak Dev Rev 2013;52:355–73.
2859
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
[3] Sher HA, Murtaza AF, Addoweesh KE, Chiaberge M. Pakistan's progress in solar
PV based energy generation. Renew Sustain Energy Rev 2015;47:213–7.
[4] British Petroleum. Statistical review of world energy June 2014; 2014.
[5] Alternative Energy Development Board Pakistan. Solar current status; 2016.
[6] Alternative Energy Development Board Pakistan. Wind current status; 2016.
[7] Ministry of Finance Pakistan. Economic survey of Pakistan 2014-15. Islamabad
Pakistan; 2015.
[8] Ministry of Finance Pakistan. Economic survey of Pakistan 2013-14. Islamabad
Pakistan; 2014.
[9] Pallett R, Brabben T. Application and experience of photovoltaic pumps for
irrigation in Pakistan. Fourth EC Photovoltaic Solar Energy Conference: Springer;
1982. p. 125–9.
[10] Sukhera MB. Utilization of solar energy – a programme for the development of
Cholistan desert. Sol Energy 1984;33:233–5.
[11] Sukhera MB, Pasha MAR, Naveed MS. Solar radiation over Pakistan – comparison
of measured and predicted data. Sol Wind Technol 1986;3:219–21.
[12] Sukhera M, Pasha M. Solar radiation maps for Pakistan. Sol Wind Technol
1987;4:229–38.
[13] Raja IA, Twidell JW. Distribution of global insolation over Pakistan. Sol Energy
1989;43:355–7.
[14] Raja IA, Twidell JW. Distribution of global insolation over Pakistan. Sol Energy
1990;44:63–71.
[15] Raja IR, Twidell JW. Distribution of diffuse and direct insolation over Pakistan.
Sol Wind Technol 1990;7:277–92.
[16] Chaudhary QUZ. A procedure to obtain global solar radiation maps from sunshine
duration for Pakistan. Sol Wind Technol 1990;7:245–53.
[17] Nasir SM, Raza SM. Wind and solar energy in Pakistan. Energy 1993;18:397–9.
[18] Raja IA, Twidell JW. Statistical analysis of measured global insolation data for
Pakistan. Renew Energy 1994;4:199–216.
[19] Raja IA, Abro RS. Solar and wind energy potential and utilization in Pakistan.
Renew Energy 1994;5:583–6.
[20] Ahmad I, Akhter P, Zaidi Z. Solar radiation over Pakistan. Appl Sol Energy
1995;31:41–50.
[21] Ghaffar MA. The energy supply situation in the rural sector of Pakistan and the
potential of renewable energy technologies. Renew Energy 1995;6:941–76.
[22] Adnan S, Khan AH, Haider S, Mahmood R. Solar energy potential in Pakistan. J
Renew Sustain Energy 2012;4:032701.
[23] Malik AQ, Mufti A, Hiser HW, Veziroglu NT, Kazi L. Application of geostationary
satellite data for determining solar radiations over Pakistan. Renew Energy
1991;1:455–61.
[24] Fawz-ul-Haq KR, Siddiqui ZR. Identifying potential solar power generation sites
using satellite apt images. Adv Space Res 1994;14:243–6.
[25] Malik A, Mufti A, Hiser H, Veziroglu N. Solar mapping of Pakistan using visible
images from geostationary satellites. Renew Energy 1998;13:1–16.
[26] GeoModel Solar. Free solar radiation maps. Slovakia; 2016.
[27] Muneer T, Maubleu S, Asif M. Prospects of solar water heating for textile industry
in Pakistan. Renew Sustain Energy Rev 2006;10:1–23.
[28] Duffie JA, Beckman WA. Solar engineering of thermal processes. Wiley; 2013.
[29] Sengupta M, Habte A, Kurtz S, Dobos A, Wilbert S, Lorenz E, et al. Best practices
handbook for the collection and use of solar resource data for solar energy
applications. Golden, CO: National Renewable Energy Laboratory; 2015.
[30] Stokler S, Schillings C. Solar resource mapping in Pakistan: solar modelling
report. Energy sector management assistance program (ESMAP). Washington,
DC: World Bank Group; 2015.
[31] GeoModel Solar. Solar and PV data. Slovakia; 2016.
[32] Angstrom A. Solar and terrestrial radiation. report to the international commis-
sion for solar research on actinometric investigations of solar and atmospheric
radiation. Q J R Meteorol Soc 1924;50:121–6.
[33] Black J, Bonython C, Prescott J. Solar radiation and the duration of sunshine. Q J
R Meteorol Soc 1954;80:231–5.
[34] Black JN. The distribution of solar radiation over the earth's surface. Arch für
Meteorol, Geophys Bioklimatol, Ser B 1956;7:165–89.
[35] Brunt D. Physical and dynamical meteorology. Cambridge University Press; 1939.
[36] Page JK. Estimation of monthly mean values of daily total short wave radiation on
vertical and inclined surfaces from sunshine records for latitudes 40°S-40°N. In:
Proceedings of the United Nations conference on new sources of energy: solar
energy, wind power and geothermal energy. Rome; 1961. p. 378–90.
[37] Black JN. Some aspects of the climatology of solar radiation. In: Proceedings of
the United Nations conference on new sources of energy: solar energy, wind power
and geothermal energy. Rome; 1961.
[38] Cooper PI. The absorption of radiation in solar stills. Sol Energy 1969;12:333–46.
[39] Löf GOG, Duffie JA, Smith CO. World distribution of solar radiation. Sol Energy
1966;10:27–37.
[40] Löf GOG, Duffie JA, Smith CD. World distribution of solar radiation engineering
experiment station report no. 21. The University of Wisconsin, College of
Engineering; 1966.
[41] Mani A, Chacko O. Solar radiation climate of India. Sol Energy 1973;14:139–56.
[42] Budyko MI. The heat balance of the earth's surface. U.S. Dept. of Commerce,
Weather Bureau; 1958.
[43] Trewartha GT, Robinson AH, Hammond EH. Elements of geography. McGraw-
Hill; 1967.
[44] Raja IA. Solar radiation assessment in Pakistan. United Kingdom: University of
Strathclyde, Glasgow; 1992.
[45] Hargreaves GH, Samani ZA. Reference crop evapotranspiration from temperature.
Appl Eng Agric 1985;1:96–9.
[46] Wit CTd. Photosynthesis of leaf canopies. Wageningen: Pudoc; 1965. p. 57.
Z.R. Tahir, M. Asim
[47] Almorox J, Bocco M, Willington E. Estimation of daily global solar radiation from
measured temperatures at Cañada de luque, Córdoba, Argentina. Renew Energy
2013;60:382–7.
[48] Polo J, Wilbert S, Ruiz-Arias J, Meyer R, Gueymard C, Suri M, et al. Integration of
ground measurements to model-derived data: a report of IEA SHC Task 46 solar
resource assessment and forecasting; 2015.
[49] Polo J, Wilbert S, Ruiz-Arias JA, Meyer R, Gueymard C, Súri M, et al. Preliminary
survey on site-adaptation techniques for satellite-derived and reanalysis solar
radiation datasets. Sol Energy 2016;132:25–37.
[50] Kalhoro A, Abro A. Global irradiation on horizontal surface at Hyderabad,
Pakistan. Mehran Univ Res J Eng Technol 2005;24:45–52.
[51] Khahro SF, Tabbassum K, Talpur S, Alvi MB, Liao X, Dong L. Evaluation of solar
energy resources by establishing empirical models for diffuse solar radiation on
tilted surface and analysis for optimum tilt angle for a prospective location in
southern region of Sindh, Pakistan. Int J Electr Power Energy Syst
2015;64:1073–80.
[52] Muneer T, Asif M. Prospects for secure and sustainable electricity supply for
Pakistan. Renew Sustain Energy Rev 2007;11:654–71.
[53] Shamshad KM. Solar insolation over Pakistan. J Jpn Sol Energy Soc 1998;24:6.
[54] Raja IA, Twidell JW. Diurnal variation of global insolation over five locations in
Pakistan. Sol Energy 1990;44:73–6.
[55] Raja IA, Twidell JW. Frequency distribution of global insolation over Pakistan.
Renew Energy 1991;1:499–505.
[56] Liu BYH, Jordan RC. The interrelationship and characteristic distribution of
direct, diffuse and total solar radiation. Sol Energy 1960;4:1–19.
[57] Liu B, Jordan R. Daily insolation on surfaces tilted towards equator. ASHRAE J
1961;10.
[58] Klein SA. Calculation of monthly average insolation on tilted surfaces. Sol Energy
1977;19:325–9.
[59] Modi V, Sukhatme SP. Estimation of daily total and diffuse insolation in India
from weather data. Sol Energy 1979;22:407–11.
[60] Ashbel D. World maps of diffuse (sky) radiation: third international geophysical
year (IGY)(1957–1958). Hebrew University, Department of Climatology and
Meteorology; 1961.
[61] Abbasi AA, Qureshi MS. Estimating global, diffuse solar radiation for Chhor and
validation with satellite-based data. Arab J Sci Eng 2014;39:175–9.
[62] Muneer T. Solar radiation and daylight models. Taylor & Francis; 2007.
[63] Hawas MM, Muneer T. Study of diffuse and global radiation characteristics in
India. Energy Convers Manag 1984;24:143–9.
[64] Memon N, Jafri AH, Yousuy I, Khan SA. Wind data analysis of AEDB-UNDP
(WEP) wind masts: installed in Gharo-Keti bandar wind corridor. Islamabad,
Pakistan: AEDB; 2010.
[65] Dedieu G, Deschamps P, Kerr Y. Satellite estimation of solar irradiance at the
surface of the earth and of surface albedo using a physical model applied to
Meteosat data. J Clim Appl Meteorol 1987;26:79–87.
[66] Renné DS, Kelly M, Elliott D, George R, Scott G, Haymes S, et al. Solar and wind
resource assessments for Afghanistan and Pakistan. In: Proceedings of ISES world
congress 2007 (Vol I–Vol V): Springer; 2007. p. 134–40.
[67] Maxwell EL, George RL, Wilcox SM. A climatological solar radiation model. In:
Campbell-Howe R, Cortez T, Wilkins-Crowder B, editors. The 1998 American
solar energy society annual conference. Albuquerque, NM; 1998. p. 505–10.
[68] George R, Maxwell E. High-resolution maps of solar collector performance using a
climatological solar radiation model. Golden, CO (US): National Renewable
Energy Lab.; 1999.
[69] National Renewable Energy Laboratory USA. Pakistan resource maps and toolkit;
2016.
[70] Perez R, Schlemmer J, Moore K, George R. Satellite-derived resource assessment
in Afghanistan & Pakistan in support of the USAID South Asia regional initiative.
USA: National Renewable Energy Laboratory; 2007.
[71] Perez R, Ineichen P, Moore K, Kmiecik M, Chain C, George R, et al. A new
operational model for satellite-derived irradiances: description and validation. Sol
Energy 2002;73:307–17.
[72] Perez R, Ineichen P, Kmiecik M, Moore K, Renne D, George R. Producing satellite-
derived irradiances in complex arid terrain. Sol Energy 2004;77:367–71.
[73] Cano D, Monget JM, Albuisson M, Guillard H, Regas N, Wald L. A method for the
determination of the global solar radiation from meteorological satellite data. Sol
Energy 1986;37:31–9.
[74] NASA. NASA Surface meteorology and solar energy; 2016.
[75] Energy Sector Management Assistance Program. Renewable energy resource
mapping in Pakistan; 2016.
[76] Stökler S, Schillings C, Kraas B. Solar resource assessment study for Pakistan.
Renew Sustain Energy Rev 2016;58:1184–8.
[77] Beyer HG, Costanzo C, Heinemann D. Modifications of the Heliosat procedure for
irradiance estimates from satellite images. Sol Energy 1996;56:207–12.
[78] Hammer A. Application-specific solar radiation information from Meteosat data.
Universität Oldenburg; 2001.
[79] Hammer A, Heinemann D, Hoyer C, Kuhlemann R, Lorenz E, Müller R, et al. Solar
energy assessment using remote sensing technologies. Remote Sens Environ
2003;86:423–32.
[80] Rigollier C, Lefèvre M, Wald L. The method Heliosat-2 for deriving shortwave
solar radiation from satellite images. Sol Energy 2004;77:159–69.
[81] Schillings C, Mannstein H, Meyer R. Operational method for deriving high
resolution direct normal irradiance from satellite data. Sol Energy
2004;76:475–84.
[82] Pinker R, Laszlo I. Modeling surface solar irradiance for satellite applications on a
global scale. J Appl Meteorol 1992;31:194–211.
2860
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
[83] Ohmura A, Gilgen H, Hegner H, Müller G, Wild M, Dutton EG, et al. Baseline
Surface Radiation Network (BSRN/WCRP): new precision radiometry for climate
research. Bull Am Meteorol Soc 1998;79:2115–36.
[84] Boilley A, Wald L. Comparison between meteorological re-analyses from ERA-
Interim and MERRA and measurements of daily solar irradiation at surface.
Renew Energy 2015;75:135–43.
[85] Hasnain S, Gibbs B. Prospects for harnessing renewable energy sources in
Pakistan. Sol Wind Technol 1990;7:321–5.
[86] Raja IA, Dougar M, Abro R. Solar energy applications in Pakistan. Renew Energy
1996;9:1128–31.
[87] Uqaili MA, Harijan K, Memon M. Prospects of renewable energy for meeting
growing electricity demand in Pakistan. Renewable energy for sustainable devel-
opment in the Asia Pacific Region. AIP Publishing; 2007. p. 53–61.
[88] Sahir MH, Qureshi AH. Assessment of new and renewable energy resources
potential and identification of barriers to their significant utilization in Pakistan.
Renew Sustain Energy Rev 2008;12:290–8.
[89] Asif M. Sustainable energy options for Pakistan. Renew Sustain Energy Rev
2009;13:903–9.
[90] Chaudhry MA, Raza R, Hayat S. Renewable energy technologies in Pakistan:
prospects and challenges. Renew Sustain Energy Rev 2009;13:1657–62.
[91] Sheikh MA. Energy and renewable energy scenario of Pakistan. Renew Sustain
Energy Rev 2010;14:354–63.
[92] Amer M, Daim TU. Selection of renewable energy technologies for a developing
country: a case of Pakistan. Energy Sustain Dev 2011;15:420–35.
[93] Malik SN, Sukhera OR. Management of natural gas resources and search for
alternative renewable energy resources: a case study of Pakistan. Renew Sustain
Energy Rev 2012;16:1282–90.
[94] Farooq MK, Kumar S. An assessment of renewable energy potential for electricity
generation in Pakistan. Renew Sustain Energy Rev 2013;20:240–54.
[95] Farooqui SZ. Prospects of renewables penetration in the energy mix of Pakistan.
Renew Sustain Energy Rev 2014;29:693–700.
[96] Rauf O, Wang S, Yuan P, Tan J. An overview of energy status and development in
Pakistan. Renew Sustain Energy Rev 2015;48:892–931.
[97] Shaikh F, Ji Q, Fan Y. The diagnosis of an electricity crisis and alternative energy
development in Pakistan. Renew Sustain Energy Rev 2015;52:1172–85.
[98] Mirza UK, Maroto-Valer MM, Ahmad N. Status and outlook of solar energy use in
Pakistan. Renew Sustain Energy Rev 2003;7:501–14.
[99] Bhutto AW, Bazmi AA, Zahedi G. Greener energy: issues and challenges for
Pakistan—solar energy prospective. Renew Sustain Energy Rev 2012;16:2762–80.
[100] Jaffery SHI, Khan M, Ali L, Khan HA, Mufti RA, Khan A, et al. The potential of
solar powered transportation and the case for solar powered railway in Pakistan.
Renew Sustain Energy Rev 2014;39:270–6.
[101] Shaikh PH, Shaikh F, Mirani M. Solar energy: Topographical asset for Pakistan.
Appl Sol Energy 2013;49:49–53.
[102] Energy USDo. Renewable energy annual 1996, Energy information administra-
tion: office of coal, nuclear, electric and alternative fuels. U.S. Department of
Energy Washington; 1996.
[103] World Renewable Energy Council. Renewable energy in South Asia, status and
prospectus; 2000. p. 51–5.
[104] McGrath C. UNESCO resource kit, science and technology education resources for
the 21st century. The Association for Science Education; 2000.
[105] Awan AB, Khan ZA. Recent progress in renewable energy – remedy of energy
crisis in Pakistan. Renew Sustain Energy Rev 2014;33:236–53.
[106] Sheikh MA. Renewable energy resource potential in Pakistan. Renew Sustain
Energy Rev 2009;13:2696–702.
[107] Khan MA, Latif N. Environmental friendly solar energy in Pakistan's scenario.
Renew Sustain Energy Rev 2010;14:2179–81.
[108] Ghafoor A, Tu Rehman, Munir A, Ahmad M, Iqbal M. Current status and overview
of renewable energy potential in Pakistan for continuous energy sustainability.
Renew Sustain Energy Rev 2016;60:1332–42.
[109] Mahmood A, Javaid N, Zafar A, Riaz RA, Ahmed S, Razzaq S. Pakistan's overall
energy potential assessment, comparison of LNG, TAPI and IPI gas projects.
Renew Sustain Energy Rev 2014;31:182–93.
[110] Khan HA, Pervaiz S. Technological review on solar PV in Pakistan: scope,
practices and recommendations for optimized system design. Renew Sustain
Energy Rev 2013;23:147–54.
[111] Khalil HB, Zaidi SJH. Energy crisis and potential of solar energy in Pakistan.
Renew Sustain Energy Rev 2014;31:194–201.
[112] Sadati SS, Qureshi FU, Baker D. Energetic and economic performance analyses of
photovoltaic, parabolic trough collector and wind energy systems for Multan,
Pakistan. Renew Sustain Energy Rev 2015;47:844–55.
[113] Kurokawa K. Areal evolution of PV systems. Sol Energy Mater Sol Cells
1997;47:27–36.
[114] Lutfi N, Veziroǧlu T. A clean and permanent energy infrastructure for Pakistan:
solar-hydrogen energy system. Int J Hydrog Energy 1991;16:169–200.
[115] Lutfi N, Veziroǧlu T. Solar-hydrogen demonstration project for Pakistan. Int J
Hydrog Energy 1992;17:339–44.
[116] Farooq M, Shakoor A. Severe energy crises and solar thermal energy as a viable
option for Pakistan. J Renew Sustain Energy 2013;5:013104.
[117] Gueymard CA, Myers DR. Solar radiation measurement: progress in radiometry
for improved modeling. Modeling solar radiation at the earth's surface. Springer;
2008. p. 1–27.
[118] Jarraud M. Guide to meteorological instruments and methods of observation
(WMO-No. 8). Geneva, Switzerland: World Meteorological Organisation; 2008.
[119] Meyer R, Lohmann S, Schillings C, Hoyer C. Climate statistics for planning and
siting of solar energy systems: long-term variability of solar radiation derived from
Z.R. Tahir, M. Asim
satellite data. Sol Energy Resour Manag Electr Gener Local Lev Glob Scale
2006:55–68.
[120] Mueller R, Dagestad K-F, Ineichen P, Schroedter-Homscheidt M, Cros S,
Dumortier D, et al. Rethinking satellite-based solar irradiance modelling: the
SOLIS clear-sky module. Remote Sens Environ 2004;91:160–74.
[121] Janjai S, Laksanaboonsong J, Nunez M, Thongsathitya A. Development of a
method for generating operational solar radiation maps from satellite data for a
tropical environment. Sol Energy 2005;78:739–51.
[122] Polo J, Zarzalejo LF, Ramírez L. Solar radiation derived from satellite images.
Modeling solar radiation at the earth's surface. Springer; 2008. p. 449–62.
[123] Perez R, Cebecauer T, Suri M. Semi-empirical satellite models. Solar energy
forecasting and resource assessment. Oxford: Academic Press; 2013. p. 21–48.
[124] Rienecker MM, Suarez MJ, Gelaro R, Todling R, Bacmeister Julio, Liu E, et al.
MERRA: NASA's modern-era retrospective analysis for research and applications.
J Clim 2011;24:3624–48.
[125] Dee DP, Uppala SM, Simmons AJ, Berrisford P, Poli P, Kobayashi S, et al. The
ERA-interim reanalysis: configuration and performance of the data assimilation
system. Q J R Meteorol Soc 2011;137:553–97.
[126] Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Gandin L, et al. The
2861
Renewable and Sustainable Energy Reviews 81 (2018) 2839–2861
NCEP/NCAR 40-year reanalysis project. Bull Am Meteorol Soc 1996;77:437–71.
[127] Saha S, Moorthi S, Pan H-L, Wu X, Wang J, Nadiga S, et al. The NCEP climate
forecast system reanalysis. Bull Am Meteorol Soc 2010;91:1015–57.
[128] Ebita A, Kobayashi S, Ota Y, Moriya M, Kumabe R, Onogi K, et al. The Japanese
55-year reanalysis “JRA-55”: an interim report. SOLA 2011;7:149–52.
[129] Compo GP, Whitaker JS, Sardeshmukh PD, Matsui N, Allan RJ, Yin X, et al. The
twentieth century Reanalysis Project. Q J R Meteorol Soc 2011;137:1–28.
[130] Decker M, Brunke MA, Wang Z, Sakaguchi K, Zeng X, Bosilovich MG. Evaluation
of the reanalysis products from GSFC, NCEP, and ECMWF using flux tower
observations. J Clim 2012;25:1916–44.
[131] Slater AG. Surface solar radiation in North America: a comparison of observa-
tions, reanalyses, satellite, and derived products. J Hydrometeorol
2016;17:401–20.
[132] Zhang X, Liang S, Wang G, Yao Y, Jiang B, Cheng J. Evaluation of the reanalysis
surface incident shortwave radiation products from NCEP, ECMWF, GSFC, and
JMA using satellite and surface observations. Remote Sens 2016;8:225.
[133] Mustafa R. New IFC guidebook supports solar investments in Pakistan. The
International Finance Corporation, The World Bank Group; 2016.