Ahmad Rizaldi, A. Darmawan, Hari Kaskoyo, Agus Setiawan
{"title":"Pemanfaatan google earth engine untuk pemantauan lahan agroforestri dalam skema perhutanan sosial","authors":"Ahmad Rizaldi, A. Darmawan, Hari Kaskoyo, Agus Setiawan","doi":"10.22146/mgi.73923","DOIUrl":null,"url":null,"abstract":"Abstrak. Strategi pengelolaan hutan secara agroforestri dalam skema Perhutanan Sosial (PS) perlu dipantau menggunakan teknologi penginderaan jauh. Teknologi analisis citra penginderaan jauh dan teknologi informasi saat ini telah berkembang ke dalam penggunaan cloud computing dan Big Data seperti platform Google Earth Engine (GEE) yang membuat perolehan data turunan citra satelit seperti tutupan lahan menjadi sangat cepat. Makalah ini bertujuan untuk menganalisis citra satelit multiwaktu menggunakan platform GEE dengan algoritma Random Forest (RF) dan Classification and Regression Trees (CART) dalam konteks pemantauan program perhutanan sosial. Hasil uji penilaian akurasi klasifikasi menunjukkan bahwa algoritma RF memiliki hasil akurasi lebih baik dengan nilai overall accuracy sebesar 94,64% dan kappa accuracy sebesar 92,23% dibandingkan dengan algoritma CART yang mendapatkan nilai overall accuracy sebesar 89,77% dan nilai kappa accuracy sebesar 85,54%. Penggunaan platform google earth engine untuk pemantauan skema PS terbukti berhasil di beberapa daerah dalam penerapan mitigasi peningkatan deforestasi dan degradasi hutan. Abstract. Agroforestry forest management needs to be carried out and monitored using remote sensing technology. The latest development related to remote sensing technology today is using cloud computing and Big Data such as the Google Earth Engine (GEE), which makes the acquisition of derived data from satellite imagery such as land cover very quickly. This paper aims to analyze multi-time satellite imagery using GEE with Random Forest (RF) and Classification and Regression Trees (CART) algorithms. The results show that the RF algorithm has better classification accuracy with an overall accuracy value of 94.64% and kappa accuracy of 92.23% compared to the CART algorithm which gets an overall accuracy value of 89.77% and kappa accuracy value of 85.54%. The use of GEE platform for monitoring PS schemes has proven successful in several areas in implementing mitigation of increased deforestation and forest degradation. ","PeriodicalId":55710,"journal":{"name":"Majalah Geografi Indonesia","volume":" ","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2022-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Majalah Geografi Indonesia","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.22146/mgi.73923","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Abstrak. Strategi pengelolaan hutan secara agroforestri dalam skema Perhutanan Sosial (PS) perlu dipantau menggunakan teknologi penginderaan jauh. Teknologi analisis citra penginderaan jauh dan teknologi informasi saat ini telah berkembang ke dalam penggunaan cloud computing dan Big Data seperti platform Google Earth Engine (GEE) yang membuat perolehan data turunan citra satelit seperti tutupan lahan menjadi sangat cepat. Makalah ini bertujuan untuk menganalisis citra satelit multiwaktu menggunakan platform GEE dengan algoritma Random Forest (RF) dan Classification and Regression Trees (CART) dalam konteks pemantauan program perhutanan sosial. Hasil uji penilaian akurasi klasifikasi menunjukkan bahwa algoritma RF memiliki hasil akurasi lebih baik dengan nilai overall accuracy sebesar 94,64% dan kappa accuracy sebesar 92,23% dibandingkan dengan algoritma CART yang mendapatkan nilai overall accuracy sebesar 89,77% dan nilai kappa accuracy sebesar 85,54%. Penggunaan platform google earth engine untuk pemantauan skema PS terbukti berhasil di beberapa daerah dalam penerapan mitigasi peningkatan deforestasi dan degradasi hutan. Abstract. Agroforestry forest management needs to be carried out and monitored using remote sensing technology. The latest development related to remote sensing technology today is using cloud computing and Big Data such as the Google Earth Engine (GEE), which makes the acquisition of derived data from satellite imagery such as land cover very quickly. This paper aims to analyze multi-time satellite imagery using GEE with Random Forest (RF) and Classification and Regression Trees (CART) algorithms. The results show that the RF algorithm has better classification accuracy with an overall accuracy value of 94.64% and kappa accuracy of 92.23% compared to the CART algorithm which gets an overall accuracy value of 89.77% and kappa accuracy value of 85.54%. The use of GEE platform for monitoring PS schemes has proven successful in several areas in implementing mitigation of increased deforestation and forest degradation.