PENENTUAN ZONA KERENTANAN GERAKAN TANAH MENGGUNAKAN INTEGRASI METODE WEIGHT OF EVIDENCE DAN LOGISTIC REGRESSION DI KABUPATEN TRENGGALEK PROVINSI JAWA TIMUR

    Sifa Muyassarah, - and Nanang Dwi Ardi, - and Sumaryono, - (2025) PENENTUAN ZONA KERENTANAN GERAKAN TANAH MENGGUNAKAN INTEGRASI METODE WEIGHT OF EVIDENCE DAN LOGISTIC REGRESSION DI KABUPATEN TRENGGALEK PROVINSI JAWA TIMUR. S1 thesis, Universitas Pendidikan Indonesia.

    Abstract

    Kabupaten Trenggalek menjadi wilayah dengan tingkat kerawanan gerakan tanah tertinggi di Jawa Timur dalam lima tahun terakhir, dengan 200 kejadian dari total 362 kejadian di provinsi ini. Kondisi tersebut menegaskan pentingnya pengembangan model pemetaan kerentanan yang mampu merepresentasikan kompleksitas spasial dan geologis secara akurat guna mendukung efektivitas mitigasi. Penelitian ini bertujuan untuk mengidentifikasi parameter dominan gerakan tanah, memodelkan zona kerentanan secara spasial, serta mengevaluasi akurasi model menggunakan integrasi metode Weight of Evidence (WoE) dan Logistic Regression (LR). Dua belas parameter telah dianalisis, mencakup elevasi, kemiringan lereng, arah lereng, kelengkungan lereng, tata guna lahan, struktur, curah hujan, NDVI, litologi, jarak terhadap sungai, serta jarak dari jalan. Sebanyak 228 titik longsor digunakan dan dibagi menjadi 80% data pelatihan dan 20% data validasi. WoE digunakan untuk menghitung bobot spasial dan menyeleksi parameter berdasarkan nilai Area Under Curve (AUC) dari kurva Receiver Operating Curve (ROC), sedangkan LR digunakan untuk membangun model multivariat. Hasil menunjukkan lima parameter dominan dengan AUC ≥ 0,6 adalah elevasi, kemiringan lereng, arah lereng, tata guna lahan, dan litologi. Model WoE menghasilkan nilai AUC sebesar 0,869 untuk success rate dan 0,853 untuk prediction rate, sedangkan model WoE-LR mencapai AUC sebesar 0,873 untuk success rate dan 0,842 untuk prediction rate. Berdasarkan klasifikasi interpretasi ROC, nilai-nilai tersebut termasuk dalam kategori sangat baik. Parameter kemiringan lereng dan tata guna lahan memberikan kontribusi signifikan terhadap potensi gerakan tanah, sedangkan jarak dari jalan dan curah hujan berkontribusi paling rendah. Temuan ini menunjukkan bahwa integrasi WoE dan LR mampu meningkatkan akurasi pemodelan dan mendukung mitigasi gerakan tanah berbasis spasial. Trenggalek Regency has become the region with the highest level of landslide susceptibility in East Java over the past five years, with 200 events out of 362 in the province. This condition emphasizes the importance of developing a susceptibility mapping model that can accurately represent spatial and geological complexities to support effective mitigation. This study aims to identify dominant landslide-triggering parameters, spatially model susceptibility zones, and evaluate model accuracy using an integration of the Weight of Evidence (WoE) and Logistic Regression (LR) methods. Twelve parameters were analyzed, including elevation, slope, aspect, curvature, land use, structure, rainfall, NDVI, lithology, distance to rivers, and distance to roads. A total of 228 landslide points were used and divided into 80% training and 20% validation data. WoE was used to calculate spatial weights and select parameters based on Area Under Curve (AUC) values from the Receiver Operating Curve (ROC), while LR was used to build a multivariate model. Five dominant parameters with AUC ≥ 0.6 were elevation, slope, aspect, land use, and lithology. The WoE model produced AUC values of 0.869 for success rate and 0.853 for prediction rate, while the integrated WoE – LR model achieved AUC values of 0.873 and 0.842, respectively. Based on ROC interpretation, these values fall into the very good category. Slope and land use showed the highest contribution to landslide potential, while distance to roads and rainfall had the lowest. These findings indicate that integrating WoE and LR improves model accuracy and supports spatial-based landslide mitigation.

    [thumbnail of S_FIS_2100069_Title.pdf] Text
    S_FIS_2100069_Title.pdf

    Download (611kB)
    [thumbnail of S_FIS_2100069_Chapter1.pdf] Text
    S_FIS_2100069_Chapter1.pdf

    Download (200kB)
    [thumbnail of S_FIS_2100069_Chapter2.pdf] Text
    S_FIS_2100069_Chapter2.pdf
    Restricted to Staf Perpustakaan

    Download (877kB)
    [thumbnail of S_FIS_2100069_Chapter3.pdf] Text
    S_FIS_2100069_Chapter3.pdf

    Download (1MB)
    [thumbnail of S_FIS_2100069_Chapter4.pdf] Text
    S_FIS_2100069_Chapter4.pdf
    Restricted to Staf Perpustakaan

    Download (1MB)
    [thumbnail of S_FIS_2100069_Chapter5.pdf] Text
    S_FIS_2100069_Chapter5.pdf

    Download (118kB)
    [thumbnail of S_FIS_2100069_Appendix.pdf] Text
    S_FIS_2100069_Appendix.pdf
    Restricted to Staf Perpustakaan

    Download (2MB)
    Official URL: https://repository.upi.edu/
    Item Type: Thesis (S1)
    Additional Information: https://scholar.google.com/citations?hl=en&user=-H7rWLoAAAAJ ID SINTA Dosen Pembimbing: Nanang Dwi Ardi: 257201
    Uncontrolled Keywords: gerakan tanah, pemetaan kerentanan, weight of evidence, logistic regression, auc, trenggalek landslide, susceptibility mapping, weight of evidence, logistic regression, auc, trenggalek.
    Subjects: G Geography. Anthropology. Recreation > GB Physical geography
    Q Science > QC Physics
    Divisions: Fakultas Pendidikan Matematika dan Ilmu Pengetahuan Alam > Program Studi Fisika - S1 > Program Studi Fisika (non kependidikan)
    Depositing User: Sifa Muyassarah
    Date Deposited: 27 Aug 2025 02:57
    Last Modified: 27 Aug 2025 02:57
    URI: http://repository.upi.edu/id/eprint/136313

    Actions (login required)

    View Item View Item