ANALISIS AKURASI CITRA MODIS DAN LANDSAT 8 MENGGUNAKAN ALGORITMA NORMALIZED BURN RATIO UNTUK PEMETAAN AREA TERBAKAR (Studi Kasus: Provinsi Riau)

Abstract

Forest fires have become a serious threat to society globally in the last decade, it is closely related to the destruction of the environment and natural resources that exist in the world, especially in tropical countries like Indonesia have a lot of forest. Riau Province is one of eight provinces were declared disaster prone to forest fires. In this research will be processing of satellite imagery with different spatial resolution and uses the same method to get the same results, the burned area. The imagery used is Landsat 8 imagery with a spatial resolution of 30 meters and MODIS imagery with a spatial resolution of 250 meters using NBR (Burned Normalized Ratio). By comparing two different data, it can be specified accuracy assessment of both images that have different resolutions. Assessment of accuracy to be achieved is the extent of the burned area agreement, related comission error, omission related error, independent comission independent error and omission error. The results of this study showed burned area classification results Landsat 8 30 meters area of 125,923.790 hectares and MODIS image classification results of 250 meters area of 171,045.154 hectares. While the results of the analysis of the accuracy of the results obtained burned areas agreement 40200.882 hectares, related commission errors 94633.149 hectares, related omission errors 44744.897 hectares, independent comission error 23165.602 hectares, and independent ommission error 21766.780 hectares.