Pemodelan prediksi dan kesesuaian perubahan penggunaan lahan menggunakan Cellular Automata-Artificial Neural Network (CA-ANN)

Abstract

The increase in population and economic growth in Sleman Regency causes an increase in the need for land even though in theory the area of ??places does not increase. In order to meet the needs of the land, many land use conversions occur. This study aims to determine land use changes, predict the accuration of land use in Sleman Regency using Artificial Neural Network (ANN) and Cellular Automata (CA) models. The model builder on ArcMap is carried out for the preparation of spatial data before further analysis is carried out using QuantumGIS with CA-ANN analysis to predict land use changes and reclassify method in ArcMAp to test the accuracy of changes in 2019 with existing ones. The data used in the form of spatial data on land use in 2015 and 2017 as well as supporting data or variables in the form of road spatial data and the distribution of educational places. The results showed that there was a considerable increase in the amount of 287,342 Ha, while the use of land as paddy fields actually decreased by 291,93 Ha. Modeling with CA-ANN shows very strong accuracy results, namely the Kappa Index of .95621 and correction of 97.14082%. Results of the prediction conformity with the existing land use shows a suitable percentage of 93,52%. Peningkatan jumlah penduduk dan pertumbuhan perekonomian di Kabupaten Sleman menyebabkan meningkatnya kebutuhan lahan padahal secara teori menyatakan bahwa luas suatu daerah tidak mengalami pertambahan. Demi memenuhi kebutuhan lahan tersebut banyak terjadi konversi penggunaan lahan. Penelitian ini bertujuan untuk memprediksi dan melakukan akurasi kesesuaian perubahan penggunaan lahan di Kabupaten Sleman menggunakan model Artificial Neural Network (ANN) dan Cellular Automata (CA). Model builder pada ArcMap dilakukan untuk persiapan data spasial sebelum dilakukan analisis data lanjutan menggunakan QuantumGIS dengan analisis CA-ANN untuk memprediksikan perubahan penggunaan lahan dan reclassify dalam ArcMap untuk menguji akurasi perubahan di tahun 2019 dengan eksisting yang ada. Data yang digunakan berupa data spasial penggunaan lahan di tahun 2015 dan 2017 serta data atau variabel pendukung berupa data spasial jalan dan sebaran tempat pendidikan. Hasil penelitian menunjukkan adanya pertambahan penggunaan lahan sebagai pemukiman yang cukup banyak yakni sebesar 287,342 Ha sedangkan penggunaan lahan sebagai sawah justru mengalami penurunan luas sebesar 291,93 Ha. Pemodelan dengan CA-ANN menunjukkan hasil akurasi yang sangat kuat yakni pada Indeks Kappa sebesar 0,95621 dan koreksi sebesar 97,14082%. Hasil kesesuaian prediksi dengan eksisting penggunaan lahan menunjukkan persentase sesuai sebesar 93,52%.

Prediction of Peri-Urban Land Use in Sleman Regency and Its Suitability towards Spatial Planning

“…In a qualitative analysis, the kappa coefficient measures the level of agreement between two experts in assessing something (Priyadi et al, 2020). In land use prediction, the kappa coefficient is used to measure the accuracy of the resulting prediction results (Nabila, 2023). The formula for calculating the kappa coefficient is as follows (Roseana et al, 2019): The calculation of the kappa coefficient is done automatically using the IDRISI Terrset.…”

Integration of Geospatial Data and Information in Margoluwih Village, Yogyakarta and Its Utilization for Multiple Stakeholders

“…There is significant potential for integrating geospatial information between land-based data and regional spatial planning, given that both agencies utilize the same foundational maps (Meidodga et al, 2023). (Nabila, 2023;Setiawan, 2021). The installation of RDTR information can help curb the rate of land use conversion, particularly agricultural land.…”

Analysis of land cover changes case in the Batang Toru landscape

“…Land use classification is categorized based on the generalization of visual similarities of the studied objects. The land use class is divided into several types based on the level of detail of the spatial information to be conveyed [20]. The land cover category is obtained through an interpretation process using Geographic Information Systems (GIS) and Remote Sensing.…”