COMPARISON OF NDVI, EVI, AND SAVI METHODS TO KNOW VEGETATION DENSITY WITH LANDSAT 8 OIL IMAGES, 2019

Case Study: Koto Tangah District, Padang City

Abstract

This study aims to determine: (1) The level of vegetation density in Koto Tangah District, Padang City in 2019 using the NDVI, EVI, and SAVI methods, (2) The vegetation index method has the highest accuracy in predicting vegetation density in Koto Tangah District, Padang City. The type of research conducted is quantitative research, with research data in the form of Landsat 8 imagery data to identify the vegetation index NDVI, EVI, and SAVI. These indexes utilize a combination of bands on Landsat imagery. The value of the vegetation index can be calculated using the existing formula. carried out ArcGIS by using the raster calculator tool by entering the band values and calculations. In taking the accuracy test on the sample used a simple random sampling technique and using the Fitzpatricklens formula for each vegetation index method. Data collection techniques used are literature study, observation, and documentation. Meanwhile, the data analysis technique uses vegetation density analysis by looking at the accuracy of the NDVI, EVI, and SAVI methods. The results in this study indicate that each vegetation index is vulnerable, namely NDVI -1 -0.3 Very rare, -0.03- 0.15 Rare, 0.15 – 0.25 Medium, 0.25 – 0.35 Meeting, 0.35 – 1 Very Meeting, SAVI -1- -0.26 Very Rare, -0.26 – 0.29 Rare, 0.29-0.66 Moderate, 0.66-0.99 Meeting, 0.99-1 Very Meeting; EVI -0.99-0.1 Very Rare, 0.1-0.17 Rarely, 0.24-037 Moderate, 0.37-0.47 Meeting, 0.47-1 Very Meeting. the value results obtained that the area of the sub-district of Koto Tangah, the city of Padang, is dominated by high. Based on the research results of the three indices, the most dominating class is very dense vegetation density. The accuracy test results for the NDVI method were 86.95%, for the EVI method it was 86.95%, and for the SAVI method, it was 91.30%.