Compared here are storage volumes determined for Attabad Lake in Hunza valley from 30-m Advance Space borne Thermal Emission and Reflection Radiometer (ASTER) Global Digital Elevation Models (GDEM2) and void filled Shuttle Radar Topographic Mission (SRTM) Digital Elevation Models (DEMs) of 90-m and 30-m in resolution. Results indicate that either DEM yields acceptable storage estimates for lake but SRTM3_1arcsec and SRTM4.1 provide the more accurate estimate. In this study we investigate the suitability of the SRTM and the ASTER GDEM2 for the storage volume of reservoir. All the volumetric calculations of proposed landslide-dammed lake are carried out by using two methods; one of them is contour interpolation method that focuses on the creation of contours to represent the lake levels and the other one is pixel by pixel method that uses digitized shorelines while the statistical approaches to obtain mean pool elevations on specific dates. DEMs were applied in two different resolutions and the approximated values of volumes were compared to values derived from the NESPAK bathymetric data based on field survey which used as the reference. Storage estimates are not significantly different between the datasets for proposed Lake. Besides artifacts, also changes due to different acquisition dates and techniques (optical, radar) also have an impact on the volumetric calculations. In evaluating these three sets of data, the ASTER GDEM2 values are constantly higher than the SRTM4.1 and SRTM3_1arcsec values in both area and volumes for a certain lake surface elevation; this difference grows larger with increasing pool height and lake volumes. Moreover, we achieved a more precise estimation of area and volume of Lake Gojal during filling, overtopping, and partial draining from pixel by pixel methodology that based on rigorous delimitation of the lake shoreline elevation; using combination of satellite imagery and digital topographic dataset. We analyzed that for the given lake shoreline, the number of counted grid cells for GDEM2 are greater as compare to SRTM3 (~90m,~30m). A 3D model of these DEMs revealed that ASTER DEM projects terrain features better than SRTM4.1 but not as much better as void filled SRTM3_1arcsec presents. We obtained the vertical accuracy of these two elevation datasets by comparing it to the actual official aeronautic data that tells us actual vertical elevation of runways in the Attabad region and found out RMSE with a small vertical difference of ±3m for SRTM dataset while had a big absolute vertical difference of ±13m for GDEM2 overall runways. The results from both assessments showed their level of suitability for geometrics application for the concerned region. No matter whatever the method is used in calculating the volumes of impoundment but we concluded that higher accuracy is achieved with SRTM, recorded standard error of 0.1309 for SRTM3_1arcsec, 0.1966 for SRTM4.1, and 0.2961 for GDEM2. Our results indicate that a fusion of digital topographic data with high-resolution optical satellite imagery is very effective in producing key data on Attabad rockslide-dammed lake for geomorphic and hydrological analysis and engineering mitigation.
Index Terms - Digital Elevation Model (DEM), NESPAK, Attabad Lake, SRTM, ASTER, Storage capacity, RMSE
More..