Forest Degradation Assessment Using UAV Optical Photogrammetry and SAR Data

Abstract

The sequestration of carbon in the forest is very high after soil and ocean; this is one of the major factors which makes monitoring of forests very much essential. The traditional inventory techniques were very much exhaustive and needed encroachment into the inaccessible forest area, which create lot of difficulties to foresters and other researchers, that is why, most of the areas of intense forests are untouched for monitoring purposes. The degradation of the forest area is highly prominent in the hilly terrain; activities such as lopping and forest fire are very common in these areas. In this research, UAV optical photogrammetric technique is used for the extraction of canopy metrics as well as for the generation of DSM and DTM, which is further used for the validation of the results obtained with the Sentinel-2 and Sentinel-1A data. The Sentinel-1A of C-band data is used for the generation of backscatter image, texture images. Sentinel-2 is used for the calculation of various vegetation-based indices for the evaluation of the proficiency of the degradation in the forest area. On the basis of various vegetation indices and texture images calculated, the degradation of the forest can be quantified. The random forest classifier is used for the classification of the forest into degraded and non-degraded classes. The classification accuracy obtained for three classes, i.e. degraded, non-degraded, non-forest is 0.86 and precision is 0.75. The correlation is 0.72 for the classes, whereas error rate is found to be 0.13. Hence, the integration of UAV optical photogrammetry and SAR data can give much appreciated results; also, this technique is very much helpful in monitoring and management of inaccessible forest areas.