Abstract:
Sediment's deposit in lakes and reservoirs is a natural process, and its regular monitoring is important for the efficient operation of reservoirs. This study focus on identification of sediment deposit patterns in Tarbela Dam using remote sensing data, and compare the results with physically collected data for validation. For the identification of sediments deposit patterns and the temporal changes, Landsat imagery of years 1982, 2000, 2003, 2010 and 2012 has been used. For the temporal analysis, the remote sensing data has been converted to Top of Atmosphere (ToA) reflectance. Analyzing the spatial and spectral profiling, it has been observed that the wavelengths in the optical region of the EM spectrum were more useful in identification of sediments. Different band combinations did help to identify the sediments' deposit patterns in the reservoir. Detailed statistical change detection analyses of the observed spectral classes also help identifying the temporal changes. Results of spectral and spatial profiling, pattern recognition and change detection are then used to map the sediment extents for different years. Two separate sediment's depth survey data (range lines and contours) from Survey and Hydrology Residency Tarbela Dam, for the years of 2000, 2003, 2010 and 2012 are used for the study. This physical data have been processed to find out the sediment deposit patterns. Since distance between the range lines varies from 600 m to 3000 m, the interpolated surface was not useful and therefore not used for further analysis. However, contour data of 10 feet interval were used to find the sediment deposit extents. The comparison of the sediment extents mapped from the satellite imagery and physically collected contour data showed the difference of about 27 %, 4 % and 1 % for the years 2003, 2010 and 2012 respectively. The use of remote sensing data can be useful for a quick and temporal monitoring of the sediment deposits when the reservoir is operating at minimum water level. However, remote sensing data alone cannot provide the complete solution, but its integrated use with the physically collected sediment survey data can definitely help detecting and monitoring sediments deposit patterns, and improving the accuracy of the results
