ANALISIS PERUBAHAN TUTUPAN LAHAN MENGGUNAKAN GOOGLE EARTH ENGINE DAN PEMODELAN CELLULAR AUTOMATA ARTIFICIAL NEURAL NETWORK DI KABUPATEN WONOGIRI

ROFI NURHADI, 21303705 (2025) ANALISIS PERUBAHAN TUTUPAN LAHAN MENGGUNAKAN GOOGLE EARTH ENGINE DAN PEMODELAN CELLULAR AUTOMATA ARTIFICIAL NEURAL NETWORK DI KABUPATEN WONOGIRI. Diploma thesis, Sekolah Tinggi Pertanahan Nasional.

Text ROFI NURHADI_21303705_1.pdf Download (1MB)

Abstract

Land cover change reflects the dynamics of spatial utilization resulting from population growth, urbanization, and infrastructure development, which directly affect spatial planning and environmental sustainability—particularly in the rapidly developing Wonogiri Regency. Urbanization and strategic infrastructure development, such as the City Ring Road (Jalan Lingkar Kota) and the Southern Cross Road (Jalur Lintas Selatan), have accelerated land-use conversion, thereby necessitating comprehensive spatial analysis. This study aims to analyze land cover changes in 2019, 2022, and 2025, and to predict the condition in 2040 to support the evaluation of the implementation of Wonogiri Regency’s Spatial Plan (RTRW). This research employs a quantitative method with a spatial approach. Sentinel-2 satellite imagery was classified using the Random Forest (RF) algorithm on the Google Earth Engine (GEE) platform, while land cover prediction for 2040 was modeled using Cellular Automata–Artificial Neural Network (CA–ANN) through the MOLUSCE plugin in QGIS. The model considered spatial variables such as distance to roads, education centers, administrative centers, and slope. The results show that the "other natural/semi-natural vegetation" category dominated land cover but decreased from 61.87% in 2019 to 60.30% in 2025, followed by the "other cultivated vegetation" category, which remained relatively stable. Meanwhile, built-up/mixed residential areas increased from 5.40% to 6.62%. The spatial change patterns were clustered and concentrated in key sub-districts such as Eromoko, Wonogiri, and Baturetno. Validation of the 2025 prediction against the actual classification yielded a correctness of 78.99%, a Kappa Overall of 0.61392, a Kappa Histogram of 0.97535, and a Kappa Location of 0.62879. These results reinforce the ongoing issue of land cover change in Wonogiri Regency, indicating the urgency of monitoring spatial planning to prevent environmental impacts such as flooding and landslides caused by uncontrolled land-use conversion. Keywords: land cover, Google Earth Engine, CA–ANN, spatial prediction, Wonogiri

Actions (login required)

View Item