Please use this identifier to cite or link to this item: http://repository.ipb.ac.id/handle/123456789/158990
Full metadata record
DC FieldValueLanguage
dc.contributor.advisorTaufik, Muh.-
dc.contributor.authorRomadhoni, Frydha Ayu-
dc.date.accessioned2024-10-04T08:48:26Z-
dc.date.available2024-10-04T08:48:26Z-
dc.date.issued2024-
dc.identifier.urihttp://repository.ipb.ac.id/handle/123456789/158990-
dc.description.abstractKelembapan tanah gambut merupakan parameter penting yang menentukan fungsi ekohidrologi dan biogeokimia lahan gambut. Kelembapan gambut diketahui dapat mengendalikan emisi gas rumah kaca, mendorong laju dekomposisi, menyerap karbon dan rentan terhadap kebakaran. Tujuan penelitian ini adalah mengestimasi kelembapan tanah lahan gambut di KHG Sungai Siak-Sungai Kampar menggunakan metode machine learning Random Forest dengan kombinasi parameter terbaik. Penelitian ini menggunakan data observasi kelembapan tanah lahan gambut sebagai data target dan curah hujan (CHIRPS), serta Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index (SAVI), Normalized Difference Moisture Index (NDMI), Normalized Difference Soil Index (NDSI), dan Normalized Difference Water Index (NDWI) yang diekstrak dari satelit Landsat 8 sebagai data prediktor. Hasil penelitian menunjukkan bahwa kombinasi mtry 5 dan ntree 400 menghasilkan model random forest dengan kinerja terbaik. Kombinasi parameter tersebut memperoleh nilai Root Mean Square Error (RMSE) 5,92%, Mean Absolute Percentage Error (MAPE) 10,71% dan R2 0,82 pada data training serta nilai RMSE 7,59%, MAPE 13,63% dan R2 0,69 pada data testing. Kelembapan tanah hasil estimasi pada data testing memiliki rentang 15,77%-76,78%, sedangkan pada data training berkisar antara 13,17%-77,62%. Implementasi model Random Forest untuk estimasi kelembapan tanah di KHG Sungai Siak-Sungai Kampar menunjukkan variasi signifikan terkait fase ENSO. Pada fase El Niño 2015, kelembapan tanah berada di rentang 22%-50% akibat curah hujan rendah, sementara pada fase La Niña 2020 sebaran kelembapan tanah sebagian besar di atas 50% serta fase normal 2017 menunjukkan kelembapan tanah lebih stabil dengan rentang 22%-61%.-
dc.description.abstractPeat soil moisture is an important parameter that determines the ecohydrological and biogeochemical functions of peatlands. Peat moisture is known to control greenhouse gas emissions, encourage decomposition rates, carbon sequestration and fire susceptibility. The aim of this research is to estimate peatland soil moisture in KHG Sungai Siak - Sungai Kampar using Random Forest with the best parameter combination. This research uses peatland soil moisture observation data as target data and rainfall (CHIRPS) and Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index (SAVI), Normalized Difference Moisture Index (NDMI), Normalized Difference Soil Index (NDSI), and Normalized Difference Water Index (NDWI )extracted from Landsat 8 satellite as predictor data. The results showed that the combination of mtry 5 and ntree 400 obtained the best performing random forest model. The parameter combination resulted in Root Mean Squared Error (RMSE) value of 5,92%, Mean Absolute Percentage Error (MAPE) 10,71% and R2 0,82 in the training data and RMSE value of 7,59%, MAPE 13,63% and R2 0,69 in the testing data. The estimated soil moisture in the testing data ranges from 15,77% to 76,78%, while in the training data it varies between 13,17% and 77,62%. The implementation of the Random Forest model for soil moisture estimation in the KHG Sungai Siak-Sungai Kampar shows significant variation related to ENSO phases. During the 2015 El Niño phase, soil moisture ranged from 22% to 50% due to low rainfall, while in the 2020 La Niña phase, most areas had soil moisture above 50%. The normal ENSO phase in 2017 showed more stable soil moisture with a range of 22% to 61%.-
dc.description.sponsorshipnull-
dc.language.isoid-
dc.publisherIPB Universityid
dc.titleEstimasi Kelembapan Tanah Gambut Berdasarkan Machine Learning pada KHG Sungai Siak–Sungai Kamparid
dc.title.alternativeEstimation of Peat Soil Moisture Based on Machine Learning in the PHU of Siak River-Kampar River-
dc.typeSkripsi-
dc.subject.keywordpeatlandid
dc.subject.keywordrandom forestid
dc.subject.keywordtuning parameterid
dc.subject.keywordLandsatid
dc.subject.keywordsoil moistureid
Appears in Collections:UT - Geophysics and Meteorology

Files in This Item:
File Description SizeFormat 
cover_G2401201037_d1ae5805231640fe906c53044e254cd8.pdfCover556.05 kBAdobe PDFView/Open
fulltext_G2401201037_beab5fc0b20140c990a9431b54cff263.pdf
  Restricted Access
Fulltext1.17 MBAdobe PDFView/Open
lampiran_G2401201037_6d73a19993f84f80a96e8e2d258615f0.pdf
  Restricted Access
Lampiran254.22 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.