Please use this identifier to cite or link to this item:
http://repository.ipb.ac.id/handle/123456789/154747
Title: | Pengembangan Model Estimasi Produksi Padi Berbasis Data Sentinel-2A |
Other Titles: | |
Authors: | Impron Pramunita, Gian |
Issue Date: | 2024 |
Publisher: | IPB University |
Abstract: | Tanaman padi adalah salah satu jenis tanaman penghasil bahan pangan pokok masyarakat, bahkan menjadi komoditi pangan utama di Indonesia. Salah satu upaya agar tanaman padi menghasilkan beras dengan kualitas dan kuantitas yang baik dapat melalui cara pemupukan dan pemeliharaan yang baik. Dinamika pertumbuhan tanaman padi pada lahan yang luas, memerlukan metode pemantauan yang efisien terhadap waktu, melalui pendekatan penginderaan jauh dengan resolusi temporal menggunakan indeks vegetasi. Oleh karenanya penelitian ini bertujuan untuk (1) mengetahui respon nilai indeks vegetasi Normalized Difference Vegetation Index (NDVI), Soil-Adjusted Vegetation Index (SAVI), Optimization Soil-Adjusted Vegetation Index (OSAVI), Enhanced Vegetation Index-2 (EVI-2) dan Normalized Difference Red-Edge Index (NDRE) terhadap perlakuan pupuk dan jarak tanam dan (2) mengembangkan model estimasi produksi padi berdasarkan indeks vegetasi. Proses pengamatan lapang dan pengumpulan data dilakukan setiap 10 hari sekali selama satu siklus tanam di Desa Pasirkaliki, Karawang, Jawa Barat. Metode pengembangan model dilakukan berdasarkan analisis pada setiap akumulasi Thermal Heat Unit (THU) yang bersesuian dengan resolusi temporal Sentinel-2A, indeks vegetasi (NDVI, SAVI, OSAVI, EVI-2, dan NDRE) serta Random Forest Regression (RFR). Hasil penelitian menunjukkan bahwa pengaruh perlakuan dosis pupuk berpengaruh signifikan pada fase vegetatif dan fase generatif, ditandai dengan nilai R2 yang tinggi untuk nilai indeks vegetasi NDVI, SAVI, OSAVI, EVI-2, dan NDRE. Selain itu, hasil analisis RFR menunjukkan bahwa NDVI, OSAVI, dan NDRE memiliki performa yang baik untuk memprediksi berat kering gabah hasil panen. Performa model berbasis indeks NDRE menunjukkan akurasi yang lebih tinggi pada akumulasi THU 176°C hari (fase vegetatif) dengan angka persentase sebesar 94,58% dan RMSE 4,14. Performa model berbasis indeks OSAVI menunjukkan akurasi yang lebih tinggi pada akumulasi THU 481°C hari (fase genetatif) dengan angka persentase sebesar 95,05% dan RMSE 3,54. Saran untuk penelitian selanjutnya, menguji kehandalan model estimasi produksi gabah kering panen berbasis indeks NDRE digunakan pada fase vegetatif (akumulasi THU 176°C hari) dan indeks OSAVI digunakan pada fase generatif (akumulasi THU 481°C hari). Rice plants are one type of plant that produces staple food for the community, even becoming the main food commodity in Indonesia. One of the efforts for rice plants to produce rice with good quality and quantity can be through fertilization and goog maintenance. The dynamics of rice plant growth on a large area of land requires a time-efficient monitoring method, through a remote sensing approach with temporal resolution using vegetation indices. Therefore, this research aims to (1) determine the response of Normalized Difference Vegetation Index (NDVI), Soil- Adjusted Vegetation Index (SAVI), Optimization Soil-Adjusted Vegetation Index (OSAVI), Enhanced Vegetation Index-2 (EVI-2) dan Normalized Difference Red- Edge Index (NDRE) to fertilizer dosage treatment and planting technique and (2) develop a rice production estimation model based on vegetation index. The field observation and data collection were carried out every 10 days during one cropping cycle in Pasirkaliki Village, Karawang, West Java. The model development method is based on the analysis of each Thermal Heat Unit (THU) accumulation that corresponds to the temporal resolution of Sentinel-2A, vegetation index (NDVI, SAVI, OSAVI, EVI-2, and NDRE) and Random Forest Regression (RFR). The results showed that the effect of fertilizer dosage treatment had a significant effect on the vegetative and generative phases, indicated by high R2 values for the NDVI, SAVI, OSAVI, EVI-2, and NDRE. In addition, the RFR analysis results showed that NDVI, OSAVI, and NDRE had a good performance to predict the dry mass of harvested grain. The performance of the NDRE index-based model shows higher accuracy on accumulated THU 176°C days (vegetative phase) with a percentage figure of 94.58% and RMSE 4.14. The performance of the OSAVI index-based model shows higher accuracy on accumulated THU 481°C days (genetic phase) with a percentage figure of 95.05% and RMSE 3.54. Suggestions for further research, the reliability of the NDRE-based harvest dry grain production estimation model used in the vegetative phase (THU accumulation 176°C days) and OSAVI- based is used in the generative phase (THU accumulation 481°C days). |
URI: | http://repository.ipb.ac.id/handle/123456789/154747 |
Appears in Collections: | UT - Geophysics and Meteorology |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
cover_G24170042_9c24d416b1ea494c979ab62aa69eb3e4.pdf | Cover | 582.42 kB | Adobe PDF | View/Open |
fulltext_G24170042_be2648a047324abf8ad62d49b599bb81.pdf Restricted Access | Fulltext | 1.62 MB | Adobe PDF | View/Open |
lampiran_G24170042_f9c2add8b5ee4753b06142d5d002f56e.pdf Restricted Access | Lampiran | 447.25 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.