Analisis Perubahan Tutupan Lahan Akibat Banjir Bandang Sumatera Tahun 2025 (Studi Kasus: Kabupaten Padang Pariaman, Provinsi Sumatera Barat)

Amadia, Afina Dahayu (2026) Analisis Perubahan Tutupan Lahan Akibat Banjir Bandang Sumatera Tahun 2025 (Studi Kasus: Kabupaten Padang Pariaman, Provinsi Sumatera Barat). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 03311940000088_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (5MB) | Request a copy

Abstract

Banjir bandang yang terjadi di Kabupaten Padang Pariaman pada Februari 2025 menyebabkan perubahan kondisi tutupan lahan di wilayah terdampak sehingga diperlukan pemantauan untuk mengetahui bentuk dan besarnya perubahan yang terjadi. Penelitian ini bertujuan menganalisis perubahan tutupan lahan pada wilayah terdampak banjir menggunakan citra satelit Sentinel-2 dengan metode klasifikasi Random Forest berbasis Google Earth Engine (GEE). Data yang digunakan meliputi citra Sentinel-2 sebelum banjir (1 Februari–30 Juni 2025) dan sesudah banjir (1 Februari–30 Juni 2026), data batas administrasi, serta data wilayah terdampak banjir dari Badan Informasi Geospasial (BIG). Proses penelitian meliputi cloud masking, pembuatan training sample, klasifikasi tutupan lahan, uji akurasi, dan analisis perubahan tutupan lahan melalui teknik overlay. Hasil penelitian menunjukkan bahwa luas wilayah terdampak banjir yang dianalisis mencapai 2.346,28 ha, dengan 1.188,32 ha (50,65%) mengalami perubahan tutupan lahan dan 1.157,96 ha (49,35%) tetap. Perubahan terbesar ditunjukkan oleh peningkatan kelas lahan terbuka seluas 606,88 ha, sedangkan vegetasi nonpertanian mengalami penurunan seluas 441,60 ha dan pertanian berkurang 179,20 ha. Hasil klasifikasi sebelum banjir memperoleh Overall Accuracy sebesar 93% dengan koefisien Kappa 0,90, sedangkan klasifikasi sesudah banjir memperoleh Overall Accuracy sebesar 89% dengan koefisien Kappa 0,85, sehingga metode Random Forest dinilai mampu menghasilkan klasifikasi tutupan lahan dengan tingkat ketelitian yang baik untuk analisis perubahan tutupan lahan pascabanjir.
======================================================================================================================================
The flash flood that hit Padang Pariaman Regency in February 2025 changed the land cover conditions in the affected areas, so keeping an eye on it was necessary to see how much and in what ways things changed. This study aims to look at land cover changes in the flood-hit areas using Sentinel-2 satellite images with a Random Forest classification method on Google Earth Engine (GEE). The data we used include Sentinel-2 images from before the flood (February 1–June 30, 2025) and after the flood (February 1–June 30, 2026), administrative boundary data, and flood-affected area data from the Geospatial Information Agency (BIG). The research steps include cloud masking, creating training samples, classifying land cover, checking accuracy, and analyzing land cover changes using overlay techniques. The study shows that the flood-affected area analyzed was 2,346.28 ha, with 1,188.32 ha (50.65%) experiencing land cover changes and 1,157.96 ha (49.35%) staying the same. The biggest change was an increase in open land by 606.88 ha, while non-agricultural vegetation dropped by 441.60 ha and agricultural land went down by 179.20 ha. The pre-flood classification got an Overall Accuracy of 93% with a Kappa coefficient of 0.90, and the post-flood classification got 89% Overall Accuracy with a Kappa of 0.85, showing that the Random Forest method can give accurate land cover classifications for analyzing post-flood land cover changes.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Banjir Bandang, Perubahan Tutupan Lahan, Sentinel-2, Random Forest, Google Earth Engine, Flash Floods, Land Cover Changes, Sentinel-2, Random Forest, Google Earth Engine
Subjects:	G Geography. Anthropology. Recreation > G Geography (General) > G70.5.I4 Remote sensing
Divisions:	Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Geomatics Engineering > 29202-(S1) Undergraduate Thesis
Depositing User:	Afina Dahayu Amadia
Date Deposited:	27 Jul 2026 07:00
Last Modified:	27 Jul 2026 07:00
URI:	http://repository.its.ac.id/id/eprint/137893

Actions (login required)

View Item