<![CDATA[Statistics Indonesia provides official statistics on rice harvested area for making government policies on food security. One of the problems encountered in collecting data on rice harvest areas is that it requires a high cost. This study aims to reduce ASF survey costs through a combined method of direct observation on odd months and indirect estimation through modeling on even months. We developed a multinomial logit model for panel data by adding the autoregressive effect and Sentinel-2 spectral indices. The research location was in Seruyan Regency, Kalimantan Tengah, with 144 observation points, and was observed for 37 months from December 2018 to December 2021. The results showed that the ASF observation estimation using the developed model was better than the model without Sentinel-2 satellite imagery and the ordinal pattern method. The average accuracy of the developed model was 84.14 percent. The highest level of sensitivity was to estimate the harvest category, fallow phase, and non-agricultural land. Statistics Indonesia can consider this research to reduce the cost of collecting ASF Survey data.]]>
Copyrights © 2023
