Vegetation index-based biomass model and Land Surface Temperature (LST) from urban green spaces in Bandung City derived from multispectral imageries

Abstract

Bandung City is the largest metropolitan city in West Java Province, with a population of more than 2.5 million people in 2019. With the vast number of population, the need for land that is inevitable. The need for land in this area continuously increased, but the balanced is not meet with the availability of land. Further, it will encourage the conversion of natural land such as green spaces into a built-up form such as settlements. A disturbed vegetation cover would have an impact on the carbon sequestration capabilities and, on the other hand, might increase the land surface temperature. The goals of this study are to determine the carbon sequestration capabilities of tree biomass in urban green space and to examine the relationship between biomass content and land surface temperature (LST). Some vegetation indices include NDVI, ARVI, and SAVI, derived from multispectral Sentinel 2B are being correlated with allometric-based biomass value to generate biomass estimation models. Spatial analysis is useful to provide information about the value of biomass in the tree canopy and spatial distribution of biomass in the City of Bandung. It results that,NDVI provides a better correlation than SAVI and ARVI. The amount of Biomass in the UGS vegetation in Bandung City is4098.504 ton.Spatial distribution of high-value biomass is located in the north of Bandung, while low-value biomass is spread spatially in the south of Bandung. Large-sized trees with high vegetation density are in areas with high biomass values.While LST values range from 0 to 34.73°C. If there is a correlation between biomass and LST, it produces an inverse correlation of 0.33 with a moderate relationship.