ASSESSMENT OF CLASSICAL, ASTER AND SRTM DEMs IN NAIROBI REGION, KENYA

Abstract

A digital elevation model (DEM) is a three-dimensional model of the terrain on Earth. To get
the most out of terrain elevation measurements collected by satellite, a global digital elevation
model (GDEM) has to have high vertical precision. Typical applications of digital elevation
models (DEMs) include reconnaissance surveys, hydrological analysis, biomass calculation,
geoid modeling, and others; this research evaluates the vertical accuracy of three DEMs:
Classical, ASTER (30 m), and SRTM (90 m). While ASTER and SRTM DEMs are derived
from satellite-based remote sensing missions, classical DEMs are derived from regional
topographical maps. For this evaluation, we used DEM-derived heights across Nairobi
County and the surrounding area to compare with orthometric heights obtained via accurate
leveling at 18 sites. The research identified that for conventional DEM, the mean and standard
deviation of the direct disparities between precisely leveled heights and DEM heights are 3.97
m and ±7.76 m, for ASTER DEM, they are 16.36 m and ±7.79 m, and for SRTM DEM, they
are -0.25 m and ±4.00 m, respectively. According to the findings, the traditional and ASTER
DEMs are the next most accurate, after SRTM DEM. After that, we used a second-order
surface polynomial at 12 sites to describe the discrepancies in heights between the DEM and
orthometric data, and we applied the same polynomial to 6 test points using cross-validation.
While the conventional and ASTER DEMs saw a decline in accuracy, the SRTM DEM saw
an improvement because to the polynomial findings.