See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271531649

DAM / RESERVOIR SITES SELECTION USING REMOTE SENSING & GIS

TECHNIQUES

Conference Paper · October 2010

DOI: 10.13140/2.1.4692.5128

CITATION READS

1 10,193

1 author:

Muhammad Ateeq Qureshi

Institute of Space Technology
4 PUBLICATIONS   7 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Muhammad Ateeq Qureshi on 31 January 2015.

The user has requested enhancement of the downloaded file.

DAM / RESERVOIR SITES SELECTION USING REMOTE SENSING &
GIS TECHNIQUES
Muhammad Ateeq Qureshi
RS&GIS Application Division
Pakistan Space and Upper Atmosphere Research Commission (SUPARCO)
Sector 28 Gulzar-e-Hijri, Off University Road, P.O.Box 8402, Karachi-75270, Pakistan
Tel: 92-0-21-34650765-79 Fax: 92-21-34694941 email: atqureshi_pk@yahoo.com

ABSTRACT
Water is one of the resources that cannot be generated but only be preserved. In future water scarcity will
be one of the biggest challenges in the World. The only way to conserve this resource is through
construction of dams/reservoirs. Dams have been built for at least 5000 years and their functions have
evolved with the developing needs of the society. Most likely, the earliest dams were built to store water
for domestic purpose and agriculture water supply. With the onset of industrial era, hydropower became a
major reason to built dams.
Choosing a suitable site is a crucial phase in reservoir construction. A well-selected site will not only give
the optimum benefits but its aesthetic value may also create a recreational area surrounding the reservoir.
Conversely, a poorly selected site could cause detrimental effects.
In this research, a criterion was developed and implemented to locate potential sites for Dams/reservoirs in
the study area. A raster based GIS was employed to implement the criteria using the Weighted Linear
Combination (WLC) methods. The WLC method produced potential sites. The outcomes indicated that the
developed criteria were sensitive to physical, environmental and economical settings on the study area.
Furthermore, GIS and remote sensing can be useful tools for generating, manipulating and handling
relevant data layers and ultimately providing management options for decision makers.

INTRODUCTION
Sites selection is not a new idea. The important is to utilize technology and techniques for site
selection, there are many traditional methods they are costly and time consumable. Nowadays advancement in
technologies, in present of space technology, we can utilize maximum resources at optimum cost.
Satellite Remote Sensing provides the synoptic view of the earth. We can use this imaging technology
for updating of our maps in possible time frame as well as acquire information about the landuse and landcover.
Geographical Information system is computer based systems that handle the attribute data as well as
spatial data where geographical information is an important characteristic. Use GIS in this study as a decision
support system regarding the site selection.

Map 1: Study Area

SELECTION CRITERIA FOR RESERVOIR LOCATION CONSTRAINTS
Factors influencing reservoir site selection includes topography (slope/aspect), hydrology (rainfall,
drainage network), geology (mineral), soil, land use/cover (agriculture, forestry), road network and development
plan (Adinarayana et al., 1995;Gismala et al., 1996). However, increasingly socio-economic and
environmental factors are also being considered (Murphy, 1977). It is more viable and humane to locate a
reservoir site near to demand points as well as minimizing settlement relocation or flooding areas with historical
archeological value. Based on these factors criteria for locating reservoir in the study area was developed
(Table-1). The constraint criteria indicate that reservoir sites should not be in the vicinity of densely populated
areas, and must have water accumulated areas, hence constraint criteria 1 and 2 . Factor 3 was included to avoid
any slope failure, and slopes up to 11 deg are recommended in the literatures. Constraint 4 was included to
account for the projected demand for water supply in the Karachi city. Total water demand was projected to be
653 mega gallon /day (Mgd) in 2010.
Table-1 Reservoir location constraints criteria

Criteria Consideration

The dam and reservoir site must:

1 Not be located in or near settlement areas Safety

2 Be on water accumulated areas Water management

3 Be on a gentle slope Environmental/safety

4 Have a sufficient volume of rainfall Consumption/economic

Multi-criteria evaluation (MCE) approaches

Using GIS will require creating information layers corresponding to each constraint. The relevant
information obtained for the study area was converted into the required constraint layers as in Table-2 using
ArcGIS software spatial analyst. The necessary constraint layers for the study area were derived from various
sources. They were produced from satellite data of SPOT-5 and Shuttle Radar Topographic Mission (SRTM)
dated 2004, and rainfall data from 1989-2000. These images were used to extract information and construct
layers of information for constraints 1 to 3, i.e. to identify settlements areas, calculate slope and drainage
network respectively. Topography was derived from a SRTM. The data used to create each layer, the criterion it
corresponds to and data source are shown in Table-2. This process is usually followed by allocation of weights
to each constraint layer, multiplying each layer by its weight and then successively multiplying the result by
each of the constraints to combine all the layers. These operations are labeled as a Multi-Criteria Evaluation
(MCE) (Heywood et al., 1998). In this study, MCE approaches were considered for combining the information
layers, we used Weighted Linear Combination (WLC) methods.
Table-2 Data and data sources used to create various information layers for the criteria

Criteria Data Source

1 Settlements Satellite images (SPOT-5)

2 Elevation / Topography Satellite images (SRTM)

3 Rainfall Precipitation data

4 Drainage network Satellite images (SRTM/SPOT-5)

The pairwise comparison method was developed by Saaty (1980) in the context of the analytic hierarchy process
(AHP). This method involves pairwise comparison to create a ratio matrix. It takes as an input the pairwise
comparisons and produces the relative weights as output. Specifically, the weights are determined by
normalizing the eigenvector associated with the maximum eigenvalue of the (reciprocal) ratio matrix.
In case of my study, the problem involves evaluating a set of sites for development on the basis of three
criteria: Settlements (Se), Slope(S), and Rainfall(R). It requires assessing the relative importance of the three
criteria. This can be done by the pairwise comparison matrix, the criterion weight computation, and the
consistency ratio estimation.

Map 2: Slope of study Area

Map 3: Flow accumulation of study Area

Map 4: Drainage network & Settlements of study Area

Development of the pairwise comparison matrix
The method employs an underlying scale with values from 1 to 9 to rate the relative preferences for
two criteria (Table-3)
Table-3: Scale for Pairwise Comparison

Intensity of importance Definition

1 Equal importance

2 Equal to moderate importance

3 Moderate importance

4 Moderate to strong importance

5 Strong importance

6 Strong to very strong importance

7 Very strong importance

8 Very to extremely strong importance

9 Extreme importance

Source: Adapted from Saaty(1980)

Suppose that slope is moderately to strong prefer over the rainfall attribute; that is the comparison
result in a value of 4. Further, suppose that slope is very strongly preferred to settlement. This is a numerical
score of 7. Finally, consider the only other pairwise comparison, which is the rainfall attribute compared to
Settlements and suppose that the former is strongly preferred to the latter, a score of 5. These scores are places
in the upper right corner of the pairwise comparison matrix (Table-4). From this information I can determine the
remaining entries in such Table-4 in the following.

Table-4: Pairwise Comparison of the Evaluation Criteria

Criterion Slope Settlements Rainfall

Slope 1 3 7

Settlements 1/3 1 5

Rainfall 1/7 1/5 1

Total 1.4762 4.2000 13

Computation of the criterion weights

 This procedure involves the following operation (a) sum the values in each column of the pairwise
comparison matrix; (b) divide each element in the matrix by its column total (the resulting matrix is referred to
as the normalized pairwise comparison matrix); and (c) compute the average of the elements in each row of the
normalized matrix, that is, divide the sum of normalized scores for each row by 3 (the number of criteria). These
averages provide an estimate of the relative weights of the criteria being compared (Table-5&6). Using this
method, the weights are interpreted as the average of all possible ways of comparing the criteria. As I can see,
the criterion weights are 0.68, 0.25, and 0.07 for slope, rainfall, and settlements respectively. This means that
slope is the most important criterion, followed by rainfall and settlements.

Table-5: Determining the relative Criterion Weight- Step-II

Slope Settlements Rainfall

0.677419 0.714286 0.538462

0.2258 0.238095 0.384615

0.096774 0.047619 0.076923

1.0000 1.00000 1.0000

Table-6: Determining the relative Criterion Weight- Step-III

Weight

(0.717949+0.769231+0.538462)/3 = 0.64

(0.1795+0.192308+0.384615)/3 = 0.28

(0.102564+0.038462+0.076923)/3 = 0.08

1.00

Estimation of the constituency ratio

In this step, I can determine if my comparison is consistent. It involves the following operations: (a)
determine the weighted sum vector by multiplying the weight for the first criterion(slope) times the first column
of the original pairwise comparison matrix, then multiply the second weights (rainfall) times the second column,
the third criterion times the third column of the original matrix, finally, sum these values over the rows; and (b)
determine the consistency vector by dividing the weighted sum vector by the criterion weights determined
previously (Table-7).
 Table-7: Determining the Consistency Ratio (CR)

Criterion Step-1 Step-2

Slope (0.64)(1)+(0.28)(3)+(0.07)(7)=2.008 2.008/0.64 =3.1214

Settlements (0.64)(0.33)+(0.28)(1)+(0.07)(5)=0.866 0.866/0.28 =3.0623

Rainfall (0.64)(0.143)+(0.28)(0.2)+(0.07)(1)=0.222 0.222/0.07 =3.0126

Now that I have calculated the consistency vector, I need to compute values for two more terms,
Lambda (λ) and the consistency index (CI). The value for lambda is simply the average value of the consistency
vector:

λ = 3.1214+3.0623+3.0126/3 = 3.0654 OR λ=3.0654

The calculation of CI (Consistency index) is based on the observation that lambda is always greater
than or equal to the number of criteria under consideration(n) for positive, reciprocal matrix, and λ=n if the
pairwise comparison matrix is a consistent matrix. Accordingly, λ-n can be considered as a measure of the
degree of inconsistency. This measure can be normalized as follows:

CI= λ-n/n-1 =3.0655-3/3-1=0.03275 OR CI=0.03275

The CI term, referred to as the consistency index, provides a measure of departure from consistency.
Further, I can calculate the consistency ratio (CR), which is defined as follows:

CR=CI/RI = 0.03275/0.58 =0.056 OR CR=0.056

Where RI is the random index, the consistency index of a randomly generated pairwise comparison
matrix. It can be shown that RI depends on the number of elements being compared (Table-8). The consistency
ratio (CR) is design in such a way that if CR<0.10, the ratio indicates a reasonable level of consistency in the
pairwise comparisons; if, however, CR ≥ 0.10, the values of the ratio are indicative of inconsistent judgments.

Table-8: Random Inconsistency Indices (RI)

n RI n RI n RI

1 0.00 6 1.24 11 1.51

2 0.00 7 1.32 12 1.48

3 0.58 8 1.41 13 1.56

4 0.90 9 1.45 14 1.57

5 1.12 10 1.49 15 1.59

Source: Adapted from Saaty(1980)

USING THE WEIGHTED LINER COMBINATION (WLC) METHOD
There are many MCE approaches are in practices; however, the two approaches are very common,
Weighted Linear Combination (WLC) and Boolean overlay methods. The weighted liner combination not only
reduced to simple Boolean Constraints but also standardized to a continuous scale of suitability from 0 (the least
suitable) to 255 (the most suitable). By rescaling the factors to a standard, continuous scale, this enables the
comparison and the combination of the factors, as in the Boolean case. This procedure not only allowed the
ability to retain the variability from the continuous factors, it also given the ability to have the factors to trade
off with each other. For example, a low suitability score in the slope factor for any given location can be
compensated for by a high suitability score in another factor. However, the Boolean overlay methodology has
little flexibility; the hard and arbitrary nature of Boolean standardization limits the flexibility and utility of any
approach using constraints. Boolean operations tend to produce extreme solutions, depending on the operator
used. For example, if a location does not meet a criteria (i.e., elevation) but meets all of the remaining criteria, it
is likely in reality, that the site may still be selected in the final review regarding the location of suitable site,
another limitation of the simple Boolean approach is that all factors had equal importance in the final suitability
map. However, in reality, this is not likely to be the case. Some criteria may not be very important in
determining the overall suitability for a given area while others may be of only marginal importance. By
performing another spatial analysis using WLC, it allows to overcome this limitation by weighting the factors
and aggregating them with a calculated weighted linear average.
The weighted linear combination (WLC) model has become popular in recent a year that’s why I am
using WLC methods over simple Boolean method. Site suitability is being calculated by the ARC/INFO GRID
module (ESRI 1992). The WLC is a mathematical model available for delineating and ranking suitable sites for
specific purposes (Hopkins 1977). This model has been used to identify and rank suitable sites for land
application of wastewater (Hendrix and Buckley 1992). The criteria are standardized to a continuous scale of
suitability from the least to the most suitable, thus giving flexibility in the site selection (Eastman, 1997).
The weighted linear combination equation is the following:

Where Si is the suitability value for each cell location fji.. Suit is the grid dot notation for classes (from Table 2);
and wj is the respective weight for factor fj
In this method the constraint layers need to be standardized to a continuous scale of suitability from 0
(the least suitable) to 255 (the most suitable) (Eastman, 1997). Therefore, all the constraints in the criteria
(Table-2) were standardized and re-scaled into 0 to 255 from the least suitable to the most suitable areas. The
procedure in the WLC requires that the principal eigenvector of the pair-wise comparison matrix be computed to
produce the best-fit set of weights. Subsequently, the best acceptable fit of the respective weights are used in the
Multi Criteria Evaluation function (MCE) to calculate the weighted linear combination (WLC) using the
constrains in the criteria. The principal eigenvector was calculated by taking the square reciprocal matrix of pair
wise comparisons between the criteria and these weights would sum to 1. This result could be achieved by
calculating the weights with each column and then averaging over all columns (Eastman et al., 1993). The
WEIGHT function in ArcGIS, which operates on this principle, was used to determine the best fit of weight
factors (Table-9). The slope factor is given more weight than other factors because areas with steep slopes have
a greater tendency to experience landslides and increase the erosion risk. An index of consistency, known as a
consistency ratio (CR), was also determined to be 0.056. The CR indicates the probability that the matrix ratings
are randomly generated and the value should be less than 0.10, otherwise the matrix rating should be re-
evaluated (Saaty, 1977). Subsequently, the acceptable best fit of the respective weights were used in the Multi
Criteria Evaluation function (MCE) to calculate the weighted linear combination (WLC) using the factors (slope,
height, land value, and rainfall) and constraints (Slope, Rainfall, and settlements) images. RECLASS functions
of ArcGIS were used. The outputs represent potential sites with the highest suitability for reservoirs shown in
(Map-5)
 Table-9: Best fit of weight factor

Factor Factor weight

Slope 0.64

Rainfall 0.28

Settlements 0.08

Map 5: Potential Sites for Dam / Reservior

DISCUSSION AND RESULT

The main considerations are being safety, economy and the environment. the criteria for locating
reservoir are sensitive to the physical, environmental and economical settings, GIS and remote sensing can be
useful tools for generating, manipulating and handling relevant data, leading eventually to identifying a number
of optimum sites for locating reservoir and ultimately providing options and, assisting with the planning process
and decision making. Remote sensing being demonstrated the effectiveness of using remotely sensed data in
providing the necessary spectral and spatial information for generating information layers for reservoir site
selection criteria. The GIS as a decision-making tool, being facilitated combining various information layers as
well as implementing the necessary analysis on the data.
The outcome of the potential sites has satisfied all the constraints in the criteria. The reservoir sites
selected were outside the settlement zone, on a gentle slope, have a water accumulation area, and have
maximum rainfall. However, in terms of numbers of suitable sites shown in Map-5
Although the GIS methodology makes the decision making process more objective, there is still an
element of subjectivity associated with the allocation of map weights and scaling. This also allows flexibility to
the planners to incorporate varying degree of importance to each criterion based on their experience. Thought
experience, a good and sound judgment in the evaluation of environmental, social and political constraints could
produce the best alternative decisions.
The result has been sent to relevant department for further site investigation and feasibility study.
CONCLUSION
This research has been focused on the application of Remote Sensing and Geographical Information
System (GIS) in dam/reservoir sites selection. The aim of the study was to explore the space technology for
maximum utilized resources at optimum cost. In third world countries the economic resources are already
limited they can’t afford the huge budget for resources surveying on traditional methods. My thesis is describe
the idea about site selection at initial level where the base line information’s are not available for resources
exploration and decision making is also a critical stage, the authority didn’t have an idea either it is
environmental friendly nor hazard. My thesis identifies new resources and maps the hydrology of these areas.
This research would be beneficial for planning and development department before going to be an actual project
we would be able to analyze the impact of the project from environmental damages to socio-economic changes
and optimum benefits to maximum utilization of resources. In this thesis we discover some new avenue to
explore resources in admire of new aspect.
In addition, I hope to bring down costs by helping to focus of new development in existing resources
areas. My thesis demonstrates that space technology can be used to analytical approach for site selection over
broad region.

REFERENCES
1. CHAOJUN LIANG and D. SCOTT MACKAY. A general Model of Watershed extraction and
representation using globally optimal flow path and up-slope contributing areas [J]. International Journal
of Geographical Information Sciences, 2000, vol.14, No.4, pp 337-358.

2. JOAQUIN BUITRAGO MARTIN RADA HERNANDO HERNANDEZ ESPERANZA BUITRAGO, ,

A Single-use Site Selection Technique Using GIS for Aquaculture Planning: Choosing Locations for
Mangrove Oyster Raft Cultural in Margarita Island, Venezuela[J]. Environmental Management, 2005,
Vol.35, No.5, pp. 544-556.

3. P. A. BRIVIO, R. COLOMBO, M. MAGGI and R. TOMASONI. Integration of Remote Sensing data

and GIS for accurate mapping of flooded areas [J]. International Journal of Remote Sensing, 2002,
Vol.23, No.3, pp.429-441.

4. TAN CHENGXUAN, SUN YE, WANG REIJIANG, HU DAOGONG. Assessment and zonation of
regional crustal stability in and around the dam region of the Three Gorges Project on the Yangtze River
[J]. Environmental Geology, 1997, Vol.32, pp.285-295.

5. Y. WANG, M. LIAO, G. SUN, J. GONG. Analysis of the water volume, length, total area and inundated
area of the Three Gorges Reservoir, China using the SRTM DEM data [J]. International Journal of
Remote Sensing, 2005, Vol.26, No.18, 20 Sept. 2005, pp. 4001-4012.

6. I.J. BATEMAN, A.P. JONES, A.A. LOVETT, I.R. LAKE, B.H. DAY. Applying Geographical
Information System (GIS) to Environmental and Resource Economics [J]. Environmental and Resources
Economics, 2002, No.22, pp 219-269.

7. R. NAGARAJAN, M. SURESH. Performance assessment of water resources project using remote

sensing and a Geographical Information system [J]. International Journal of Remote Sensing, 2002, Vol.
23, No.20, pp. 4475-4485.

8. A. STASSOPOULOU, M. PETROU, J. KITTLER. Application of a Bayesian network in a GIS based

decision making system [J]. International Journal of Geographical Information Sciences, 1998, Vol. 12,
No. 1, pp. 23-45.

9. ALBERTO CARRARA, GABRIELE BITELLI, ROBERTO CARLA. Comparison of techniques for

generating digital terrain models from contour lines [J] International Journal of Geographical Information
Science, 1997, Vol.11, No.5, pp.451-473.
10. FRANCISCO NUNES CORREIA, FILIPE CASTRO REGO, MARIA DA GRAÇA SARAIVA,
ISABEL RAMOS, , Coupling GIS with Hydrologic and Hydraulic Flood Modeling [J]. Water Resources
Management, 1998, Vol.12, pp.229-249.

11. M. CHICA-OLMO, F. ABARCA, J. P. RIGOL. Development of a Decision support system based on

remote sensing and GIS techniques for gold-rich area identification in SE Spain [J]. International Journal
of Remote Sensing, 2002, Vol.23, No. 22, pp.4801-4814.

12. W. Z. SHI, Q. Q. LI, C. Q. ZHU. Estimating the propagation error of DEM from higher-order
interpolation algorithms [J], International Journal of Remote Sensing, 2005, Vol. 26, No.14, pp. 3069-
3084.

13. BASSAM SALEH, BALQIES SADOUN. Design and Implementation of a GIS system for planning [J].
International Journal on Digital Libraries, 2006, pp. 210-218.

14. S. K. JAIN, R. D. SINGH, S. M. SETH. Design Flood estimation using GIS supported GIUH approach
[J]. Water Resources Management, 2000, Vol.14, pp. 369-376.

15. A. S. El-Hames. Determination of groundwater availability in shallow arid region aquifers utilizing GIS
technology: a case study in Hada Al-Sham Western Saudi Arabia [J]. Hydrogeological Journal, 2005,
Vol.13, pp. 640-648.

16. R. NAGARAJAN. Environmental impact analysis of Dudhganga Dam in India-a multi-temporal remote
sensing approach [J]. International Journal of Remote Sensing, 2000, Vol.21, No.3, pp. 483-497.

17. LUIS A. BOJORQUEZ-TAPIA, SALOMON DIAZ-MONDRAGON, EXEQUIEL EZCURRA. GIS-

based approach for participatory decision making and land suitability assessment [J]. International
Journal of Geographical Information Science, 2001, Vol.15, No.2, pp. 129-151.

18. JACEK MALCZEWSKI. GIS-based multicriteria decision analysis: a survey of the literature [J].
International Journal of Geographic Information Science, 2006, Vol.20, No.7, pp.703-726.

19. A. K. SARAF, P. R. CHOUDHURY, B. ROY, B. SARMA, S. VIJAY, S. CHOUDHURY. GIS based

surface hydrological modeling in identification of groundwater recharge zones [J]. International Journal
of Remote Sensing, 2004, Vol.25, No. 24, pp.5759-5770.

20. HENRI TARDIEU. Automatic DAM management and River regulation for irrigation purposes [J].
Irrigation and Drainage system, 1988, Vol.2, pp. 53-61.

21. SERWAN M. J. BABAN, KAMRUZAMANWAN-YUSOF. Modelling Optimum Site for Locating

Reservoirs in Tropical Environments [J]. Water Resources Management, 2003, Vol.17, pp. 1-17.

22. BADRI B. BASNET, ARMANDO A. APAN, STEVEN R. RAINE. Selecting Suitable Sites for Animal
Waste Application Using a Raster GIS [J]. Environmental Management, 2001, Vol.28, No.4, pp.519-531.

23. ARONOFF, STAN. GIS-A Management Prospective [B]. Ottawa, Canada: WDL publication, 1989

24. CURRAN PAUL J. Principles of Remote Sensing [B]. New York: Longaman Pub Inc., 1985

25. MARSH, M. W. Environmental Analysis for Land Use and Site Planning [B]. New York: Mcgraw-Hill
Book Company, 1978

26. HARRIS, RAY. Satellite Remote Sensing an Introduction [B]. London & New York: Routledge &
Kegan Paul, 1987

27. KHAN, F. K. A Geography of Pakistan [B]. Karachi, Pakistan: First Edition, Oxford University Press,
1991.
 28. PSIHP. Project Direction and Review Report[R], Pakistan Snow and Ice Hydrology Project, H&R
Directorate, WAPDA, Lahore, Pakistan. 1991

29. KHAN, A. R. An Analysis of Surface Water Resources and Water Delivery Systems in the Indus
Basin[R], International Water Management Institute (IWMI), Lahore, Pakistan, 1999.

30. McKinney, D. C. C.A. Scott, Modeling Water Resources Management at the Basin Level[R]: Review
and Future Directions, SWIM Paper 6, IWMI-Colombo, Srilanka, 1999.

31. MUNIR S., I. M. KALWIJ, M. BROUWER. Assessment of Water Distribution at watercourse and
Minor Level of Bahadurwah Minor [R]: IIMI Report No. R-91, Lahore, Pakistan. 1999

32. RAY, K. and KAKBEEKE, W.J. Community participation in urban water and sanitation: The missing
link in public- private partnership. Presented at the International Conference on Water and
Sustainable Development [C], Paris, 12 p. 1998 March 19-21

View publication stats