Abstract

Digital elevation models (DEM) are useful for estimating phenomenon on the earth surface and understanding ecological processes. Satellite imageries have been extensively used for quantitative modeling of terrain forms of the earth surface but the global nature of the model produced numerous data holes due to cloud cover, lacking the required accuracy in capturing detailed regional and small area analyses. This study explored some spatial interpolation methods on elevation data collected for a small area of Yewa Division of Ogun State Nigeria using a handheld Global Positioning System (GPS). The GPS Elevation data were compared with Google Earth Pro Lands at 8 and Shuttle Radar Topographic Mission (SRTM) satellite sources to establish accuracy levels. There were no significant differences (F2,276 = 0.00056; p > 0.999) in the spot height data from the Ground Control Points (GCP), Google Earth Pro Landsat 8 (GOOGLE) and the Shuttle Radar Topographic Mission (SRTM). Among all the interpolation techniques that included Inverse distance weighting (IDW), Ordinary Kriging (ORK), Simple Kriging (SIK), Empirical Bayesian Kriging (EBK), Global Polynomial Interpolation (GPI), Local Polynomial Interpolation (LPI), Kernel Interpolation with Barrier (KIB) and Radial Basis Function (RBF), ordinary kriging appeared to be the most suitable spatial interpolation method (RMSE = 7.7944) that produced the most accurate digital elevation model for Yewa Division of Ogun State. It could be concluded that spot height data of GCPs of Yewa Division of Ogun State taken with a handheld GPS subjected to ordinary kriging interpolation produced reliable and suitable digital elevation model for quantitative analysis of the earth terrain forms and ecological processes of the area.