Using radar images (Sentinel-1) in the study of vegetation (A case study on the regions of Prophet Matta and Sahl Akkar at Tartous Governorate)

Abstract

The visual images give an integrated view of the growth of crops and vegetation cover in general, but this process may be hindered by some circumstances and weather conditions (clouds, cirrus...). And since the topographic characteristics are considered as a basic pillar in the study of vegetation cover and its diversity, therefore, the aim of this research is to evaluate the accuracy of radar images (Sentinel-1) in particular in assessing the area of ​​​​vegetation cover and its diversity, studying and analyzing these characteristics (the height of the terrain, slope and aspect) and other natural factors that affect the vegetation cover in the studied areas.

The vegetation cover was classified based on the radar images (Sentinel-1) for the months (February, April, August, October) of the year 2017, and it was compared with the classification results based on the visual images (Sentinel-2) that contain spectral characteristics for the same months. Supervised classification was applied using the temporal tracking of the studied months. However, when classifying using the spectral fingerprint for each type of vegetation, it gave more accurate results than the supervised classification.

After testing the accuracy of the radar classification, it proved that it is good only in classifying field crops, but it is not suitable in classifying the rest of the vegetation cover varieties after comparison with the results of the field audit, because the (Sentinel-1) radar images are in the (C) band, which was found to be Not suitable for the classification of vegetation, especially at forests.

Maps, tables, and graphs have been developed showing the area of ​​plant varieties according to ranges of elevation, slope, and aspect. Pearson's correlation coefficient of vegetation cover areas with topographical characteristics was calculated.