Thermal index effect in forest canopy density (FCD) methods based on remote sensing imagery

Abstract

The objective of this study is to identify variations in the level of accuracy of the Forest Canopy Density (FCD) method's utilization of thermal index (TI). This is significant because the FCD approach can be used to determine vegetation cover without relying on the TI indicator. The four primary indicators used in the initial development of the FCD approach by Rikimaru et al. were the vegetation index, shadow index, soil index, and thermal index. The Split Windows Algorithm (SWA), which is the most effective for Landsat 8 OLI/TIRS imagery with a combination of bands 10 and 11, is utilized as the thermal index calculation method. SWA is obtained by concentrating on variations in the vegetation index value used to calculate surface emissivity. Hence, two types of FCD-SWA FCD and non-SWA FCD-are developed. The results showed that accuracy is obtained using the error matrix: the non-SWA FCD is 42% and the SWA FCD is 53%. In addition, the 1 x 1 test plot further show that SWA FCD tends to overestimate, while non-SWA FCD tends to underestimate. The overall accuracy of the analysis conditions may be impacted by the availability of additional samples and the occurrence of the COVID-19 incident. Based on this, that FCD with four indicators may be more accurate than FCD without TI. Despite the need to deep attention to the high-FCD class of analysis, which has a propensity to overestimate