Hytlrologicitl Sciences-Journaï-des Sciences Ilydrologiques. 47(2) April 2002 203

Assessment of sedimentation in Bhakra Reservoir in the western Himalayan region using remotely sensed data

SANJAY K. JAIN, PRATAP SINGH National Institute of Hydrology, Jul Vigyan Bhawan, Roorkee 247667 India

saniavfainih.cruel.in

S. M. SETH Poomima College of Engineering, RIICO Complex. Sitapura Road, Jaipur 302022, India

Abstract Sediment particles originating from erosion processes in the catchment are propagated along with the river flow. When the flow of a river is stored in a reservoir, the sediment settles in the reservoir and reduces its capacity. Reduction in the storage capacity of a reservoir beyond a limit hampers the purpose for which it was designed. Thus assessment of sediment deposition becomes very important for the management and operation of such reservoirs. Some conventional methods, such as hydrographie survey and inflow-outflow approaches, are used for estimation of sediment deposition in a reservoir, but these methods are cumbersome, time consuming and expensive. There is a need for developing simple methods, which require less time and are cost effective. In this study, a remote-sensing approach has been attempted for assessment of sedimentation in Bhakra Reservoir, located on the Satluj River in the foothills of the Himalayas. Multi date remote sensing data (1RS-1B, L1SS II) provided the information on the water-spread area of the reservoir, which was used for computing the sedimentation rate. The revised capacity of the reservoir between maximum and minimum levels was computed using the trapezoidal formula. The loss in reservoir capacity due to deposition of sediments for a period of 32 years (1965-1997) was determined to be 807.35 Mm', which gives an average sedimentation rate of 25.23 Mm' year"1. The average rat£ of sedimentation using hydrographie survey data for the same period was 20.84 Mm' year"1. A comparison of the results shows that the rate of sedimentation assessed using the remote sensing based approach was close to the results obtained from the hydrographie survey.

Key words sedimentation; storage capacity; reservoir; remote sensing; hydrographie survey; erosion; Himalaya; Satluj; Bhakra

Estimation de la sédimentation dans le barrage Bhakra dans l'Ouest de l'Himalaya grâce à des données de télédétection Résumé Les particules sédimentaires produites par les processus d'érosion au sein du bassin versant sont transportées par l'eau des cours d'eau. Lorque l'écoulement d'un cours d'eau est stocké dans un barrage, les particules y sédimentent et en réduisent la capacité. La réduction de la capacité de stockage d'un barrage au delà d'un certain seuil nuit à sa vocation. L'estimation du dépôt sédimentaire est donc très importante pour la gestion d'un tel barrage. Quelques méthodes conventionnelles, comme la surveillance hydrographique et les calculs de bilan, sont disponibles pour estimer le dépôt sédimentaire dans un barrage, mais elles restent lourdes, lentes et onéreuses. 11 est nécessaire de développer des méthodes simples, moins consommatrice de temps et moins chères. Dans cette étude, nous essayons une méthode d'estimation de la sédimentation dans le barrage Bakhra, sur la rivière Satluj, dans le piemont himalayen, basée sur la télédétection. Des données multi-dates de télédétection (IRS-1B, L1SS H) ont permis de caractériser l'emprise du barrage, nécessaire au calcul du taux de sédimentation. La capacité corrigée du réservoir entre des niveaux maximum et minimum a été estimée à partir de la formule trapézoïdale. La diminution de la capacité de .stockage due à la sédimentation sur 32 ans (1965-1997) a été estimée à 807.35 Mm', ce qui correspond à un taux de sédimentation moyen de 25.23 Mm' par an. La surveillance hydrographique donne parallèlement la valeur de 20.84 Mm'' pour

Open for discussion until I October 2002

204 Sanjay K. Jain et al.

le même taux moyen, sur la même période. Une comparaison des résultats montre que les estimations du taux de sédimentation sont très proches avec la méthode basée sur la télédétection et avec la surveillance hydrographique. Mots clefs sédimentation; capacité de stockage; barrage; télédétection; surveillance hydrographique; érosion; Himalaya; Satluj; Bhakra

INTRODUCTION

The entire Himalayan region is afflicted with a serious problem of soil erosion and therefore the rivers flowing through this region transport a heavy load of sediment (Sharma et al, 1991). The important sources of the accelerated soil erosion and high concentrations of sediment into the Himalayan rivers are deforestation, road construction, mining, cultivation on steep slopes and seismic activities (Varsheney et ai, 1986). Approximately 30 000 km" have been severely eroded in the northeastern Himalayas due to shifting cultivation (Narayan & Ram Babu, 1983). The existence of a number of glaciers in the high altitude region of the Himalayas also contributes to the high rate of sediment load in the river (Singh & Singh, 2001). The silt transported by the rivers and their tributaries is deposited in the reservoirs reducing the reservoir capacity of the reservoirs and affecting their useful life. As a result, environmentalists and water resources planners are very much concerned with this impact on the fragile Himalayan ecosystem.

After arrival of sediment-laden flow into a reservoir, the coarser particles settle first in the upper reach of the reservoir due to the decrease in the flow velocity. Subsequently, the finer sediment material deposits further into and along the reservoir bed. Sediment deposition into reservoirs built for hydropower generation has several major detrimental effects which include loss of storage capacity, damage to or impairment of hydro equipment, bank erosion and instabilities, upstream aggradation, and effect on water quality. Assessment of reservoir sedimentation is part of the basic information needed for the operation of any reservoir. An up-to-date knowledge of the sedimentation process and deposition would help in ensuring remedial measures are taken well in advance so that the reservoir operation schedules can be planned for optimum utilization. For this reason, systematic capacity surveys of a reservoir are conducted periodically using conventional equipment, e.g. theodolites, plane table, sextant, range finders, sounding rods, echo-sounders and slow moving boats.

The two most common conventional techniques for quantifying sedimentation in a reservoir are (a) direct measurement of sediment deposition by hydrographie surveys, and (b) indirect measurement using the inflow-outflow records of a reservoir. Both of these methods are cumbersome, time consuming and expensive. Remote sensing techniques, offering data acquisition over a long time period and for a broad spectral range, are considered superior to the conventional methods for data acquisition. Spatial, spectral and temporal attributes of remote sensing data provide invaluable and timely synoptic infor­mation regarding (a) changes in the water-spread area of the reservoir after deposition of sedimentation and (b) sediment distribution patterns in the reservoir (Manavalan et al., 1990; Goei & Jain 1996). The remote sensing based approach can be cost effective, easy to use and requires less time in analysing the data, compared to the conventional methods discussed above. No study has been made to date for estimation of reservoir sedimenta­tion in the Himalayan region using remote sensing data. The present study deals with an

Assessment of sedimentation in Bhakra Reservoir using remotely sensed data 205

assessment of sediment deposition in the Bhakra Reservoir located on the Satluj River in the foothills of the Himalayas using a remote sensing approach.

A REVIEW ON RESERVOIR SEDIMENTATION STUDIES USING REMOTE SENSING DATA

The use of remote sensing techniques to estimate suspended sediment has been reported by several investigators (Solomonson, 1973; Bartoluci et al., 1977; Holeyer, 1978; Khoram, 1981). Smith et al. (1980) determined siltation in the Aswan High Dam Reservoir by comparing reflectance values in the green and red portions of the spectrum. The surface area of the entire reservoir was determined by totalling all pixels classified as water. Research findings indicate that siltation during the flood period was largely confined to the main river channel of the reservoir and large embayments. Areas of extensive siltation were identified and the amounts of deposition were determined through ground surveys. This information was used to predict the distribution of silt deposits in the reservoir. Rao et al. (1985) used a visual interpretation technique on large-scale imagery of Landsat-MSS to estimate the water-spread area at different levels of Sriramsagar Reservoir. They used these water-spread estimates to evaluate the capacity of the reservoir and concluded that the results are comparable with hydrographie survey observations. Water spread of Hirakud Reservoir from multi-date Landsat-MSS imagery was computed by Mohanty et al. (1986), who reported that the area capacity curves derived using remote sensing data were almost similar to the curves obtained from the conventional methods. Vibulsresth et al. (1988) employed digital techniques in which density slicing of Landsat-MSS near-infrared (IR) data was performed for extracting the water-spread area of Ubolratana Reservoir in Thailand. They correlated computed surface areas with the water levels and calculated the reservoir capacity based on the surface area obtained using cone formulae. For monitoring the use of water from a reservoir on a fortnightly/monthly basis, Jagadeesha & Palnitkar (1991) also used satellite data for determining water-spread area at various reservoir stages and average crop water requirements at different stages of growth were determined. The crop area information was also obtained from satellite data. They adopted the Borland and Miller (B&M) method for finding the pattern of distribution of deposited sediments between the various zones of the reservoir. Goel & Jain (1996) earned out a reservoir sedimentation study using the density-slicing approach for water-spread area extraction. They used IRS-1 A (LISS II) data to evaluate reservoir sedimentation in Dharoi Reservoir. The status of studies shows that no application of the remote sensing approach for sedimentation has been made for the Himalayan region, which has extensive areas that are prone to soil erosion.

THE STUDY RESERVOIR

For the present study, Bhakra Reservoir, which is also known as Gobindsagar, was chosen for assessment of sedimentation (Fig. 1). The Bhakra Dam, one of the oldest built in India, was commissioned in 1963; it has controlled devastating floods and the benefits to irrigation and power have brought prosperity to north India. Bhakra Dam

206 Sanjay K. Jain et al

Himachal Pradesh

Fig. 1 Location map of the Bhakra Reservoir.

has a designed dead storage of 2431.81 Mm"' and live storage of 7436.03 Mm1; i.e. a total storage capacity of 9867.84 Mm". The water-spread area of the Bhakra Reservoir extends over 168.35 km"" at full reservoir level (515.11 m) and its head touches a point about 12.87 km above Slapper village near Kasol (CBIP, 1990).

The Bhakra Reservoir is fed by the flows consisting of contribution from rain and snowmelt. Singh & Kumar (1997) have studied precipitation distribution for several Himalayan basins and found that the maximum contribution to annual rainfall (42-60%) is received during the monsoon season, whereas the minimum (5-10%) is received in the post-monsoon season. Consequently, the reservoir attains its maximum water level just after the monsoon season. The water level of the reservoir gradually reduces due to various types of use and reaches lower levels before the onset of the next monsoon. The Satluj River transports heavy amounts of sediment, which is detrimental to the life of the reservoir. The silt contribution in this basin is largely due to deforestation, over-grazing in the pasture lands, unscientific agricultural practices, farming at elevated terraces etc. (BBMB, 1997). This region is also very prone to landslides and slips which may be one of the major sources of sediment in this river. The natural factors that also attribute to high levels of sediment transport from the study region are steep topographic gradient, poor structural characteristics of soils; clay rich rocks such as Spiti shales and schists; and the widespread existence of limestone deposits (Sharma et ai, 1991).

DATA USED

To estimate the actual silt deposits in the Bhakra Reservoir and verify the project assumptions, hydrographie surveys have been carried out in the reservoir annually from 1963 to 1977 and every alternate year thereafter. The most recent hydrographie survey for this reservoir was carried out between October 1996 and March 1997. The surveys were

Assessment of sedimentation in Bhakra Reseivoir using remotely sensed data 207

carried out by observing the soundings by means of an echo sounder along predetermined cross-sections, approximately 610 m apart.. To work out the quantity of silt deposited at each cross-section and in the whole reservoir, these results were superimposed on the previous observations (WAPCOS, 1996; BBMB, 1997).

The reservoir level data for the Bhakra Reservoir were obtained from Bhakra Beas Management Board (BBMB), the agency responsible for collection of required data and operation of the reservoir. The reservoir level data are collected from the gauge installed on the dam. The hourly levels are recorded between 06:00 and 06:00 h and the mean daily value of the reservoir level is obtained by taking the average of 24 hourly values. During the study year 1996/97, the maximum reservoir level (512 m) was observed on 22 September 1996, which gradually reduced and reached the minimum level (449 m) on 5 June 1997. The IRS-1B satellite data (LISS II sensor, resolution of 36.25 m) were obtained from the National Remote Sensing Agency, Hyderabad, India. These were digital multi spectral data of four bands of wavelength region (0.45-0.52 jim, 0.52-0.59 urn, 0.62-0.68 urn and 0.77-0.86 jim). These wavelengths represent three visible and one near-infrared band. The reservoir water-spread area was covered in Bl quadrant of Path 29 and Row 45 of the satellite. After browsing the data of the study area, five cloud-free dates (6 October 1996, 7 November 1996, 21 December 1996, 12 January 1997 and 15 June 1997) were identified and used in this study.

METHODOLOGY

The methodology adopted for this study involves pre-processing of satellite data, identifying the water pixels and computation of the capacity of the reservoir. These are discussed in brief below.

To quantify the volume of sediments deposited in the reservoir, the basic information that needs to be extracted from the satellite data is the water-spread area of the reservoir at different water elevations. In this study the satellite data were processed and analysed using the ERDAS/IMAGINE 8.3.1 software for determining the water-spread area in the reservoir. Each scene of data consisted of 2500 rows, 2520 columns and the information of four bands. Initially, a false colour composite (FCC) of the satellite data was prepared and visualized. The FCCs of October 1996 and June 1997 depicting maximum and minimum water-spread areas are shown in Fig. 2(a) and (b). The pixels representing water-spread area of the reservoir were clearly distinguishable in the FCC. For processing of satellite data generally it is necessary to geo-reference the images of different time periods when using the temporal satellite data of the same area. In fact, determination of the water-spread area in a reservoir did not require the geo-referencing of the different scenes. However, using the geo-referenced imageries, it was possible to overlay the remote sensing data of different dates. Comparison of the change in the water-spread area and shrinkage in the water-spread area with time, particularly for the tail end of the reservoir, can also be made using geo-referencing. The image of October was considered as the master image and eight control points were selected for image-to-image registration. Based on the statistics, the points which generated big errors were deleted and replaced by other points to obtain satisfactory geo-referencing. All the available images were geo-referenced following this procedure. After completing this process, different images were displayed one over the other and the supeiposition was compared. It was noticed that the geo-referencing was very accurate.

208 Sanjay K. Jain et al.

(a)

(b)

Fig. 2 False colour composites of Bhakra Reservoir, located in the foothills of the Himalayas: (a) 16 October 1996, and (b) 15 June 1997.

For differentiation of water pixels from the other land-use features, a generalized algorithm based on the information of different bands was adopted (Goel & Jain, 1998). Each pixel has a numerical value called a digital number (DN), that records the intensity of electromagnetic energy measured for the ground resolution cell represented by that pixel. Using the spectral information, the algorithm matches the signatures of the pixel

Assessment of sedimentation in Bhakra Reservoir using remotely sensed data 209

with the standard signatures of water and identifies whether a pixel represents water or not. The spectral signature shows the reflectance/emittance pattern of any object at different wavelengths. The adopted algorithm states "If the DN value is of near-IR spectral region, the DN value of water pixels is appreciably less than the DN value of Band 2 and Band 3, then it must be classified as water, otherwise not". Since the absorption of electromagnetic radiation by water is maximum in the near-IR spectral region, the digital number (DN) of water pixels is considerably lower than that corresponding to other land uses. Even if the water depth is very shallow, the increased absorption in Band 4 will restrict the DN value to be less than Band 3 and Band 2. If the soil is exposed (possibly saturated) at the surface, the reflectance will be as per the signatures of the soil, which increases with wavelength in this spectral range. Thus following this algorithm, water pixels were clearly differentiated from the surrounding other pixels.

For computation of reservoir capacity between two consecutive reservoir elevations, usually three formulae, the prismoidal formula, the Simpson formula and the trapezoidal formula are used (Patra, 2001). Of these, the trapezoidal formula has been most widely used for computation of capacity (Jayapragasam et al, 1980; Manavalan et ai, 1990; Goel & Jain 1996). In this method, the cross-sectional areas of range lines are planimetred and these data, together with surface areas at full reservoir level between adjacent ranges, are used to computed the sediment volumes, as follows:

V = j(Al+A2+JÛ*li)) (1)

where V is the volume between two consecutive levels, A \ is the contour area at eleva­tion 1, A2 is the contour area at elevation 2 and H is the difference between elevations 1 and 2. The volume of sedimentation deposit between two reservoir levels is computed from the difference between previous capacity survey and satellite-derived information. The water-spread area of the reservoir was calculated from satellite data and the level corresponding to the date of pass was collected from the project authority.

From the elevation-area table, the original areas at the intermediate elevations (reservoir elevations on the dates of satellite pass) were obtained by linear interpolation. From the known values of original and estimated areas at different elevations, the corresponding original and revised capacities were determined as mentioned above. The overall reduction in capacity between the lowest and the highest observed water levels was obtained by adding the reduced capacity at all levels.

RESULTS AND DISCUSSIONS

The water-spread area of the reservoir was calculated using remotely sensed data. The difference in volume between two consecutive levels was calculated using the trapezoidal formula and is given in Table 1. In the present study, the cumulative revised capacity of the reservoir at the observed lowest level (450.44 m) was assumed to be the same as the original cumulative capacity (2392.95 Mnr ) at this elevation. Above the lowest observed level, the cumulative capacities between the consecutive levels were added up to arrive at the cumulative original and revised capacities at the maximum observed level. The difference between the original and estimated cumulative capacity represented the loss of

210 Sanjay K. Jain et al.

Table 1 Assessment of sediment deposition in Bhakra Reservoir using remote sensing (RS) for the year (1996/97).

Date of satellite pass

15/06/97 12/01/97 21/12/96 07/11/96 16/11/96

Reservoir elevation (m)

450.44 487.74 494.62 506.21 510.46

Original area (Mm2)

63.54 113.99 126.57 150.41 158.36

Revised area using RS data (Mm2)

48.81 99.87

118.80 138.13 139.69

Original volume (Mm3)

3264.89 828.02

1603.11 657.32

Revised volume using RS data (Mm3)

2716.47 752.16 1487.53 590.81

Original cumulative volume (Mm3)

2392.95 5657.39 6485.41 8088.57 8745.84

Revised cumulative volume using RS ^ data (Mm') 2392.95 5108.0 5860.15 7347.67 7938.48

Table 2 Results of hydrographie survey for the Bhakra Reservoir (1996/97) (BBMB, 1997).

Original designed capacity of reservoir in 1965: Reservoir capacity at the end of 1996/97: Dead storage Live storage Total Dead storage Live storage Total (Mm3) (Mm3) (Mm3) (Mm3) (Mm') (Mm3) 2431.81 7436.03 9867.84 1763.53 6769.96 8590.57

capacity due to sedimentation in the live zone of the reservoir. Table 1 presents the volume at different dates used to calculate the sediment deposition in the reservoir.

The capacity for the year 1996/97, estimated using remote sensing techniques (7938.48 Mm') was subtracted from the original capacity (8745.84 Mm') at the same level. The loss in capacity (807.36 Mm') was attributed to the sediment deposition in the zone of study, i.e. between 510.46 m and 450.44 m of the reservoir. Thus, the average rate of loss of capacity is computed to be 25.23 Mm' yeaf ' for the "live zone" using remote sensing data. A comparison of the cumulative original and revised capacities obtained using remote sensing technique for the year 1996/97 is shown in Fig. 3.

520

510

500

E 490 ,o

I 480 '5

5 470 in °i

oc 460

450

440

2000 3000 4000 5000 6000 7000 8000 9000 10000

Rese rvo i r capacityff i f l m3}

Fig. 3 Elevation capacity curves for Bhakra Reservoir, India (1996/97).

— Original Capacity

- . Estimated Capacity

Assessment of sedimentation in Bhakra Reservoir using remotely sensed data 211

The result of the sedimentation survey for dead and live loads is given separately (BBMB, 1997). The results of the recent survey taken from this report for the year 1996/97 are given in Table 2. It can be inferred from this table that, during a period of 32 years, the loss in the dead storage due to sedimentation is about 668.28 Mm1, whereas in live storage it is 666.85 Mm'. Thus, according to this survey, the average sedimentation rate is found to be 20.88 and 20.84 Mm3 year"1 for the dead and live zones, respectively. In other words, the sedimentation rate in both dead and live storages is almost the same. The sedimentation rate in the reservoir (25.23 Mm3 year"1) obtained using the remote-sensing approach is slightly higher than that obtained from the hydrographie survey (20.84 Mm1 year"1). This higher sedimentation rate can possibly be attributed to sensitivity in the determination of the water-spread area using remote sensing techniques. Moreover, the mixing of pixels having a large proportion of land and a smaller proportion of water, such as those around the periphery of the reservoir, may also affect the results.

CONCLUSIONS

The application of remote sensing techniques for estimating the sedimentation rate in the Bhakra Reservoir (located on Satluj River in the foothills of western Himalayas) shows that the average sedimentation rate for 32 years (1965-1997) is 25.23 Mnx' year"1, whereas ground observations through hydrographie survey provided a sedimentation rate of 20.84 Mnr for the same period. The higher sedimentation rate obtained using remote sensing data can be explained on the basis of accuracy in the determination of water-spread area and the mixing of water pixels with the land around the periphery of the reservoir. However, the use of better (spatial and temporal) resolution satellite data may be a remedy for these problems to some extent.

The use of remote sensing technique enables a fast and reasonably accurate estimation of live storage capacity loss due to sedimentation. Keeping in view the time and cost involved in hydrographie surveys, it is recommended that hydrographie surveys may be conducted at longer intervals and the remote sensing based sedimentation surveys may be carried out at shorter intervals, so that both surveys complement one another. However, there are some limitations in the remote sensing data collection method. For example, remote sensing techniques give the information on the capacities only in the water level fluctuation zone, which generally lies in the live zone of the reservoir. Below this zone, i.e. in the dead load zone, the information on the capacity could be taken from the most recently conducted hydrograhic survey.

Acknowledgements The authors express their sincere thanks to the Bhakra Beas Management Board (BBMB), Nangal, India for providing sediment data and water levels on different dates and other relevant information for the study. The help provided by Mr M. K. Goel (National Institute of Hydrology, Roorkee) in this study is thankfully acknowledged.

REFERENCES

Bartolucl, L. A., Robinson, B. F. & Siiva, XL XL (1977) Field measurement of spectral response of natural waters. Photogvamm. Engng Remote Sens. 43(5 ).

212 Sanjay K. Jain et al.

BBMB (Bhakra Beas Management Board) (1997) Sedimentation survey report. BBMB. Bhakra Dam Circle, Nangal, India.

CBIP (Central Board of Irrigation and Power) (1990) Sedimentation studies in reservoirs, vol. 1. Tech. report no. 20. Research Scheme Applied to River Valley Projects, Central Board of Irrigation and Power, New Delhi, India.

Goel, M. K. & Jain, S. K. (1996) Evaluation of reservoir sedimentation using multi-temporal IRS-1A LISS II data. Asian Pacific Remote Sens. & GISJ. 8(2), 39^13.

Goel, M. K. & Jain, S. K. (1998) Reservoir sedimentation study for Ukai dam using satellite data. Report no. UM-1/97-98, National Institute of Hydrology, Roorkee, India.

Holeyer, R. J. (1978) Suspended sediment algorithms. Remote Sens. Environ. 10(4), 323-338. Jagadeesha, C. J. & Palnitkar. V. G. (1991) Satellite data aids in monitoring reservoir water and irrigated agriculture.

Water Int. 16, 27-37. Jayapragasam, R. & Muthuswamy, K. (1980) Sedimentation studies in Vaigai Reservoir using grid system../. Irrig. Power,

CBIP 37( 1 ), 337-350. Central Board of Irrigation and Power, New Delhi, India. Khoram, S. (1981) Use of Ocean colour scanner data in water qualitv mapping Photogrumm. Engwg Remote Sens. 47(5),

667-676. Manavalan, P., Sathyanath, P., Sathyanarayn, M. & Raje Gowda, G. L. (1990) Capacity evaluation of the Malaprabha

reservoir using digital analysis of satellite data. Tech. Report no. RC: BO: WR: 001:90, Regional Remote Sensing Service Centre, Bangalore and Karnataka Engineering Research Station, Krishnarajsagar. India.

Mohanty, R. B., Mahapatra, G., Mishra, D. & Mahapatra, S. S. (1986) Report on application of remote sensing to sedimentation studies in Hirakud reservoir. Orissa Remote Sensing Application Centre, Bhubaneswar and Hirakud Research Station, Hirakud, India.

Narayan Dhurva, V. V. & Ram Babu (1983) Estimation of soil erosion in India. J. Irrig. Drain. Engng 109(4), 419-434. Patra, K. C. (2001 ) Hydrology and Water Resources Engineering. Narosa Publishing House, New Delhi, India. Rao, IT. G., Rameshwar Rao & Viswanatham, R. (1985) Project report on capacity evaluation of Sriramsagar reservoir using

remote sensing techniques. Andhra Pradesh Engineering Research Lab., Hyderabad, India. Sharma, P. D., Goel, A. K. & Minnas, R. S. (1991) Water and sediment yields into the Satluj River from the High

Himalaya. Mountain Res. Devel. 11(2), 87-100. Singh, P. & Kumar, N. (1997) Effect of orography on precipitation in the western Himalavan region. J. Hvdrol. 199, 183-

~ 206. Singh, P. & Singh, V. P. (2001) Snow and Glacier Hydrology. Vol. 37 in the Water Science and Technology Library,

Kluwer Academic Publishers, Dordrecht, The Netherlands. Smith, S. E., Mancy, K. H. & Latif, A. F. A. (1980) The application of remote sensing techniques towards the management

of the Aswan high dam reservoir. In: 14th Int. Svmp. on Remote Sensing of Environment (San Jose, Costa Rica, 23-30 April 1980), 1297-1307.

Solomonson, V. V. (1973) Remote sensing applications in water resources. In: Third Earth Resources Technokrey Svmp. ( Washington DC, USA, 10-14 December, 1973 ).

Vibulsresth, S„ Srisangthong, D., Thisayakorn, K„ Suwanwerakamtorn, R., Wongpam, S., Rodpram, C , Leelitharn, S. & Jiltanaon, W. (1988) The reservoir capacity of Ubolratana dam between 173 and 180 meters above mean sea level. Asian-Pacific Remote Sens. J. 1(1).

Varsheney, R. S„ Prakash, S. & Sharma, C. P. (I986) Variation of sediment rate of reservoirs in Himalayan region with catchment. Technical session IX paper, Proc. 53rd R&D Session (8-10 May 1986), 285-294. Central Board of Irrigation & Power, Bhubhaneshwar, India.

WAPCOS (Water and Power Consultancy Services) (1996) Operation of Bhakra and Pong reservoirs during Hood and other Hood mitigation measures. Final report. Water and Power Consultancy Services (India) Ltd, New Delhi, India. Received 18 January 2001; accepted 11 September 2001