IPB University Logo

SCIENTIFIC REPOSITORY

IPB University Scientific Repository collects, disseminates, and provides persistent and reliable access to the research and scholarship of faculty, staff, and students at IPB University

AI Repository
 
Building and Categories


      View Item 
      •   IPB Repository
      • Final Assignments
      • Undergraduate Final Assignments
      • UF - Faculty of Mathematics and Natural Sciences
      • UF - Geophysics and Meteorology
      • View Item
      •   IPB Repository
      • Final Assignments
      • Undergraduate Final Assignments
      • UF - Faculty of Mathematics and Natural Sciences
      • UF - Geophysics and Meteorology
      • View Item
      JavaScript is disabled for your browser. Some features of this site may not work without it.

      Analisis Probabilitas Kebakaran Hutan dan Lahan pada Musim Hujan dan Musim Kemarau Menggunakan Random Forest di Provinsi Riau

      Thumbnail
      View/Open
      Cover (562.4Kb)
      Fulltext (2.590Mb)
      Date
      2026
      Author
      Muhesa, Illona Annisa
      Boer, Rizaldi
      Metadata
      Show full item record
      Abstract
      Kebakaran hutan dan lahan (karhutla) merupakan permasalahan lingkungan tahunan di Provinsi Riau yang menimbulkan dampak ekologis, kesehatan, dan sosial-ekonomi yang signifikan. Penelitian ini bertujuan untuk menganalisis pengaruh variabel meteorologi, lingkungan, antropogenik, dan topografi terhadap probabilitas kejadian karhutla pada musim hujan dan musim kemarau, mengidentifikasi variabel prediktor yang paling berpengaruh pada setiap musim, serta memetakan variasi spasial tingkat kerawanan karhutla di Provinsi Riau. Data spasial berbasis grid beresolusi 1 km × 1 km periode 2015–2023 digunakan untuk membangun model, sedangkan data tahun 2024 digunakan untuk prediksi. Pemodelan dilakukan menggunakan Random Forest berbasis ensemble yang terdiri atas 10 model dengan teknik repeated Random Under-Sampling (RUS) untuk mengatasi ketimpangan kelas antara kejadian fire dan non-fire, serta dilengkapi kalibrasi probabilitas untuk mengoreksi bias akibat penyeimbangan kelas. Hasil evaluasi model pada data uji dengan proporsi kelas asli menunjukkan kinerja yang baik, dengan nilai AUC sebesar 0,905 pada musim hujan dan 0,938 pada musim kemarau, serta nilai recall masing-masing sebesar 87% dan 88%. Kelembapan udara, suhu udara, dan curah hujan merupakan prediktor yang paling berpengaruh pada musim hujan, sedangkan pada musim kemarau prediktor yang paling berpengaruh adalah suhu udara, tutupan lahan, dan kelembapan udara. Hasil penelitian ini diharapkan dapat menjadi referensi dalam mendukung upaya mitigasi dan pengendalian karhutla yang mempertimbangkan karakteristik musiman di Provinsi Riau.
       
      Forest and land fires are an annual environmental issue in Riau Province, Indonesia, causing significant ecological, public health, and socio-economic impacts. This study aimed to analyze the influence of meteorological, environmental, anthropogenic, and topographic variables on the probability of forest and land fires occurrence during the rainy and dry seasons, identify the most influential predictor variables for each season, and map the spatial variation of fire susceptibility across Riau Province. Spatial grid-based data with a resolution of 1 km × 1 km from 2015 to 2023 were used to develop the model, while data from 2024 were used for prediction. The modeling was conducted using an ensemble Random Forest approach consisting of 10 models with repeated Random Under-Sampling (RUS) to address the severe class imbalance between fire and non-fire occurrences. Probability calibration was subsequently applied to correct the bias introduced by class balancing. Model evaluation on the test dataset with the original class distribution demonstrated good performance, achieving an AUC of 0.905 for the rainy season and 0.938 for the dry season, with recall values of 87% and 88%, respectively. Air humidity, air temperature, and rainfall were identified as the most influential predictors during the rainy season, whereas air temperature, land cover, and air humidity were the most influential predictors during the dry season. The findings of this study are expected to serve as a reference for supporting fire mitigation and management strategies that consider seasonal characteristics in Riau Province.
       
      URI
      http://repository.ipb.ac.id/handle/123456789/177425
      Collections
      • UF - Geophysics and Meteorology [1818]

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository
        

       

      Browse

      All of IPB RepositoryCollectionsBy Issue DateAuthorsTitlesSubjectsThis CollectionBy Issue DateAuthorsTitlesSubjects

      My Account

      Login

      Application

      google store

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository