Study of the Effect of the Number of Check Points on the Evaluation of the Vertical Accuracy of the Free Digital Elevation Model SRTM-1. Case study: Latakia Governorate

Abstract

Global digital elevation models (DEMs) are considered essential inputs for many studies, but they still contain significant elevation errors that affect the evaluation process, including the number of test points used in the evaluation. This study evaluated the impact of the number of test points on the accuracy assessment of the SRTM-1 digital elevation model, which is one of the most commonly used global models. A new method was also proposed to detect outliers in elevation differences. The tests related to the number of check points were conducted in the Latakia governorate, which was divided into two areas: a low relief area with elevations ranging from 0 m to 50 m and a high relief area with elevations ranging from 50 m to 1550 m. As for the test points, they were extracted from a topographic map with a scale of 1/25000, and their number was 100 points in each area.

The results showed a convergence between the accuracy values resulting from the tested cases of the number of check points in the low ruggedness area, with confirmation that the best accuracy obtained was in the case of 80 test points, which was 2.993 m. The results also indicated that increasing the number of check points beyond 20 points did not increase accuracy in the cartographic sense. In fact, the contour period for the case of 20 check points was 10 m, and the scale was 1/25000, which was the same result for 80 check points. As for the high ruggedness area, we observed a clear improvement in the accuracy values resulting from the tested cases of the number of check points in the high ruggedness area when adopting 40 check points. In fact, the results showed that increasing the number of test points to 40 points increased the accuracy in the cartographic sense, where the contour period in this case was 50 m, and the scale was 1/250000.