Analisis Transformasi NDVI dan kaitannya dengan LST Menggunakan Platform Berbasis Cloud: Google Earth Engine

Abstract

This paper aims to convey the results of the study in the form of detecting the vegetation index and its relation to land surface temperature. Landsat data was taken between 2016 and 2021 in Semarang City, while the method used was spatio-temporal remote sensing with LST and NDVI algorithms, processed using the Google Earth Engine cloud-based platform with open source code. The results of the analysis in 2016 and 2021 in Semarang City, the largest NDVI transformation occurred in the low vegetation class, which increased by 26.80% and the decrease occurred in the high vegetation class by 19.65%. Meanwhile, the largest LST transformation was a decrease of 110.42% in temperature classes > 30ºC from 6196.68 Ha to 2944.98 Ha, and an increase in temperature class from 24 - 26ºC from 445.59 Ha to 2057.76 Ha. The results of the linear correlation test between NDVI and LST in 2016 obtained the equation y=-6.7124x+33.042 with R2 = 0.4758, while in 2021 it was y=-6.5081x+32.203 with R2 = 0.5316. This phenomenon requires great attention, because NDVI is strongly correlated with LST decline, so it is absolutely necessary to control it through spatial planning policies.