ESTD PERPUSTAKAAN UNTAD
ESTD PERPUSTAKAAN UNTAD
| JudulPREDIKSI PERUBAHAN EKOSISTEM HUTAN MANGROVE DI KECAMATAN PARIGI SELATAN KABUPATEN PARIGI MOUTONG MENGGUNAKAN MODEL CELLULAR AUTOMATA MARKOV CHAIN |
| Nama: KLAUDITA DEBORAH PANSIANG |
| Tahun: 2025 |
| Abstrak Mangrove forest is a coastal ecosystem that must be maintained as a provider of natural resources and as a life support system. The study of mangrove forest ecosystems provides a lesson that this ecosystem is absolutely necessary and must be guaranteed its survival. Satellite imagery can be used to identify mangrove forest ecosystems because the geographical location of mangrove vegetation is at the transition between land and sea so that mangrove vegetation will provide a distinctive recording effect when compared to other land vegetation. This study aims to analyze land cover changes in mangrove ecosystems and make predictive modeling in 2030 with a ten-year time span using Landsat 7 and 8 images in 2000, 2010, and 2020. This research was conducted for three months from January 2024 to March 2024, which took place in South Parigi District, Parigi Moutong Regency. This research used a visual image analysis method with Cellular Automata (CA) - Markov Chain modeling. This analysis was carried out by creating a spatial model based on several driving factors such as land elevation roads and points of interest (facility points) with the CA-Markov method. The results of visual interpretation of land cover for the period 2000, 2010, and 2020 resulted in 8 land cover classes, namely forests, mangroves, ponds, rice fields, shrubs, settlements, water bodies, and agriculture. The results of this study resulted in image interpretation accuracy in 2020 of 94.4% which was then used to carry out scenarios and validation of land cover for the 2030 prediction year period. Validation of 2020 land cover and actual land cover in 2020 resulted in a Kstandard accuracy value of 87.45%, which means that the level of accuracy is high because the results are >75%. The accuracy test results show a value that is good enough to be used as a prediction model design that has been confirmed and then continued to model land cover in 2030. |