DETEKSI PERKEMBANGAN LAHAN TERBANGUN KOTA GORONTALO BERDASARKAN CITRA LAST (LANDSAT, ASTER, & SENTINEL-2A) (Detection of the built-up area development in Gorontalo City Based on LAST (Landsat, ASTER, & Sentinel-2A) Imagery)

Abstract

Abstract – Built-up area is easily found in urban areas which is the most land use compared to other land use. The development of the built-up area has also increased with increasing population and increasing economic activity. Most of the population activities in the form of economy, services, trade, offices, education, health, and entertainment facilities that are centralized in urban areas have caused the availability of non-built-up area to shrink further. Detection of the built-up area can be assessed from remote sensing data using urban indices, multispectral classification (supervised and unsupervised classification), and spectral bands. This study aims to detect the built-up area based on multisensor and multitemporal imagery. Landsat 5 TM, Landsat 8, ASTER, and Sentinel-2B (LAST) images were used in this study. Digital image processing is performed on each image using the guided classification method support vector machine (SVM) algorithm. Four classes of land cover were taken, namely built-up area, vegetation, bare land, and water bodies. Samples of built-up area classes were taken as many as 31 random sampling points spread over the study area. Validation tests were carried out for each image based on the ground check. Results of the study showed that the development of the built-up area was directed to the north and the difference in the extent of information on the built-up area due to differences in spatial resolution.

Keywords: built-up area, landsat, aster, sentinel, supervised classification, gorontalo


Abstrak –Lahan terbangun mudah ditemukan di wilayah perkotaan yang merupakan penggunaan lahan paling banyak dibandingkan penggunaan lahan lainnya. Perkembangan lahan terbangun turut meningkat seiring pertambahan jumlah penduduk dan peningkatan aktivitas ekonomi. Sebagian besar aktivitas penduduk berupa ekonomi, jasa, perdagangan, perkantoran, pendidikan, kesehatan, dan sarana hiburan yang terpusat di wilayah perkotaan menyebabkan ketersediaan lahan non-terbangun kian menyusut pula. Deteksi lahan terbangun dapat dikaji dari data penginderaan jauh menggunakan indeks perkotaan (urban index), klasifikasi multispektral (supervised and unsupervised classification), dan saluran spektral (spectral bands). Penelitian ini bertujuan mendeteksi lahan terbangun berdasarkan citra multis-sensor dan multi-temporal. Citra landsat 5 TM, landsat 8, ASTER, dan sentinel-2B (LAST) digunakan dalam penelitian ini. Pengolahan citra digital dilakukan pada masing-masing citra yang menggunakan metode klasifikasi terbimbing algoritma support vector machine (SVM). Sebanyak empat kelas tutupan lahan diambil, yaitu lahan terbangun, vegetasi, lahan terbuka dan tubuh air. Sampel kelas lahan terbangun diambil sebanyak 31 titik secara random sampling yang tersebar di wilayah penelitian. Uji validasi dilakukan untuk masing-masing citra berdasarkan ground check. Hasil penelitian menunjukkan bahwa perkembangan lahan terbangun mengarah ke utara, dan perbedaan luasan informasi lahan terbangun yang disebabkan perbedaan resolusi spasial.

Kata kunci: lahan terbangun, landsat, aster, sentinel, klasifikasi terbimbing, gorontalo