Perbandingan Estimasi Evapotranspirasi Lahan Sawah Menggunakan Data Satelit Landsat dan Data Stasiun Klimatologi

Abstract

Estimasi evapotranspirasi yang akurat berperan penting dalam pengelolaan irigasi dan sumber daya air pertanian. Penelitian ini bertujuan membandingkan estimasi evapotranspirasi berbasis citra satelit Landsat dan data stasiun klimatologi pada lahan sawah di Sawah Baru Bogor, Plakaran Bantul, dan Kacangan Bojonegoro. Evapotranspirasi aktual (?????? ) diestimasi menggunakan algoritma SEBAL pada Google Earth Engine, sedangkan evapotranspirasi tanaman (??????) dihitung dari evapotranspirasi referensi metode FAO Penman–Monteith dengan pendekatan koefisien tanaman berbasis NDVI. Nilai ?????? rata-rata berturut-turut sebesar 2,80; 1,76; dan 1,39 mm hari-1 . Hasil perbandingan menunjukkan bahwa ?????? berbasis NDVI lebih mendekati ??????dibandingkan metode FAO konvensional dengan persentase galat sebesar 3,76%, 30,68%, dan 11,70%. Hasil penelitian menunjukkan bahwa pendekatan NDVI mampu merepresentasikan kondisi vegetasi aktual dengan lebih baik sehingga berpotensi mendukung estimasi evapotranspirasi dan pengelolaan irigasi berbasis penginderaan jauh.

Accurate evapotranspiration estimation plays an important role in irrigation management and agricultural water resource planning. This study aimed to compare evapotranspiration estimates derived from Landsat satellite imagery and climatological station data for paddy fields located in Sawah Baru (Bogor), Plakaran (Bantul), and Kacangan (Bojonegoro). Actual evapotranspiration (??????) was estimated using the Surface Energy Balance Algorithm for Land (SEBAL) implemented on the Google Earth Engine platform, while crop evapotranspiration (??????) was calculated from reference evapotranspiration using the FAO Penman– Monteith method combined with an NDVI-based crop coefficient approach. The average ??????values were 2,80, 1.76, and 1.39 mm day?¹ for Sawah Baru, Plakaran, and Kacangan, respectively. Comparison results indicated that NDVI-based ?????? produced estimates closer to ?????? than the conventional FAO approach, with percentage errors of 3,76%, 30,68%, and 11,70%, respectively. These findings demonstrate that the NDVI-based approach better represents actual vegetation conditions and has strong potential for supporting evapotranspiration estimation and precision irrigation management using remote sensing techniques.