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 - School of Data Science, Mathematic and Informatics
      • UF - Computer Science
      • View Item
      •   IPB Repository
      • Final Assignments
      • Undergraduate Final Assignments
      • UF - School of Data Science, Mathematic and Informatics
      • UF - Computer Science
      • View Item
      JavaScript is disabled for your browser. Some features of this site may not work without it.

      Pemodelan Statistical Downscaling Data Sunshine Duration Menggunakan Random Forest di Hikmah Farm Pangalengan

      Thumbnail
      View/Open
      Cover (1.649Mb)
      Fulltext (1.678Mb)
      Date
      2026
      Jenis/Type
      Skripsi
      Subtype
      Undergraduate Theses
      Author
      Jocelyn, Tiffany Anastasia
      Agmalaro, Muhammad Asyhar
      Kustiyo, Aziz
      Metadata
      Show full item record
      Abstract
      Lama penyinaran matahari merupakan salah satu informasi agroklimat yang penting dalam mendukung pengelolaan budidaya dan pengambilan keputusan pertanian. Penelitian ini bertujuan membangun model klasifikasi kejadian sunshine duration lokal di Hikmah Farm Pangalengan menggunakan random forest dengan data atmosfer ERA5 sebagai variabel prediktor dan data sunshine duration Sinaubumi IPB sebagai variabel target biner. Pada penelitian ini, kelas 1 merepresentasikan kondisi terdapat lama penyinaran matahari pada jam yang diprediksi, sedangkan kelas 0 merepresentasikan kondisi tidak terdapat lama penyinaran matahari pada jam yang diprediksi. Data ERA5 terlebih dahulu diprosesmelalui interpolasi Inverse Distance Weighted (IDW) untuk meningkatkan resolusi spasial, kemudian dilakukan koreksi elevasi menggunakan data Shuttle Radar Topography Mission (SRTM) pada variabel suhu. Model dioptimasi menggunakan GridSearchCV dengan validasi silang 5-fold. Hasil evaluasi menunjukkan bahwa model menghasilkan accuracy sebesar 0,89, precision sebesar 0,89 untuk kedua kelas, recall sebesar 0,82 pada kelas 0 serta F1-score sebesar 0,85 pada kelas 0 dan 0,91 pada kelas 1. Model ini memberikan kinerja yang lebih baik pada rentang waktu yang lebih panjang, yaitu mulai 09 Desember 2025 hingga 08 Juni 2026. Analisis feature importance mengidentifikasi variabel surface short-wave radiation downwards (ssrd) sebagai prediktor yang paling berpengaruh dalam proses klasifikasi. Hasil penelitian ini menunjukkan bahwa kombinasi data reanalisis, statistical downscaling, dan random forest dapat digunakan untuk mengidentifikasi kondisi ada dan tidak adanya penyinaran matahari pada skala lokal.
       
      Sunshine duration is an important agroclimatic parameter that supports crop management and agricultural decision-making. This study aims to develop a local sunshine duration occurrence classification model for Hikmah Farm, Pangalengan, using the Random Forest algorithm, with ERA5 atmospheric data as predictor variables and sunshine duration data from Sinaubumi IPB as the binary target variable. In this study, class 1 represents the presence of sunshine duration during the predicted hour, whereas class 0 represents the absence of sunshine duration during the predicted hour. The ERA5 data were first processed using Inverse Distance Weighted (IDW) interpolation to improve spatial resolution, followed by elevation correction using Shuttle Radar Topography Mission (SRTM) data on the temperature variables. The model was optimized using GridSearchCV with 5-fold cross-validation. The evaluation results showed that the model achieved an accuracy of 0.89, a precision of 0.89 for both classes, a recall of 0.82 for class 0, and F1-scores of 0.85 for class 0 and 0.91 for class 1. The model demonstrated better performance over a longer observation period, spanning from December 9, 2025, to June 8, 2026. Feature importance analysis identified surface short-wave radiation downwards (SSRD) as the most influential predictor in the classification process. The results indicate that the combination of reanalysis data, statistical downscaling, and the Random Forest algorithm can be effectively used to identify the presence and absence of sunshine duration at the local scale.
       
      URI
      http://repository.ipb.ac.id/handle/123456789/175471
      Collections
      • UF - Computer Science [163]

      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