Rice is one of the main food crops, not only in Indonesia but also in Asia and worldwide. Global rice production and consumption involve more than 250 million farmers and 3.3 billion consumers, respectively. The most common pest that frequently attacks rice is brown planthopper or BPH (Nilaparvata lugens Stal). The objective of this study was to estimate area infested by BPH by using satellite spectral data analysis. The methods consisted of five stages, i.e. data preparation, field checking, determination of planting dates, Vegetation Index (VI) analysis, and estimation of infested area. Data preparation included data downloading and projecting, image cropping and digitizing. Field checking was carried out to validate the data and to get historical data of BPH infestation. The planting dates were determined by investigating the annual pattern of VI and rice plant development. VI analysis was using NDVI (Normalized Difference Vegetation Index) and EVI (Enhanced Vegetation Index). Estimation of infested area consisted of 4 procedures i.e. data normalization, Vegetation Index Unit (VIU) calculation, infested area estimation, and comparison of NDVI-EVI value. This study indicated that satellite spectral data analysis could be used to estimate the area infested by BPH. It has been shown that the analysis could differentiate between healthy and infested area. The VI value and the peak of infested area were lower and earlier, respectively compared to the healthy plantation. NDVI analysis was more effective compared to NDVI analysis in estimating BPH infested area.
Accurate information of evapotranspiration (ET) will provide important information in irrigation systems. ET can be estimated using simple method using satellite data and field measurements. Field measurements was conducted in Air Buluh Village, Bangka Belitung Island on seven land cover (three different palm-oil ages, bare land, agricultural land, reclamation, shrub) by using the Penman-Monteith method. Enhanced Vegetation Index (EVI), Normalized Difference Vegetation Index (NDVI), Soil-Adjusted Vegetation Index (SAVI), spectral reflectance short-wave band (SWIR), were used from satellite data derived from Landsat 8-OLI/TIRS data. The analysis shows that the vegetation index has positive correlation with ET field measurements with a correlation only 12-15% and errors 1.3 mm/day, with SAVI has the lowest correlation, followed by EVI and NDVI. SWIR has negative correlation with 7% coefficient correlation. SWIR is a sensitive band with surface water content. SWIR increases with decrease of surface water content. Estimation of ET were built by using simple regression between ET and vegetation index and SWIR. Regression result shows that EVI has the highest correlation, 42% followed by NDVI and SAVI, 39%, and 30%. The equation formed by vegetation index and SWIR is: ET=EVI(2.68e6.14 SWIR), ET=NDVI(2.67e 4.57SWIR), ET=SAVI(4.45e4.07 SWIR) ,with P-value < 0.05 for all method with error 1-2.5 mm/day