Prototype Development of SMART Agriculture Platform Integrating Remote Sensing Technology with AquaCrop Model

Abstract

This article aims to report the result of SMART Agriculture Platform prototype consists in AquaCrop processing system with satellite derived inputs i.e. FY rainfall and TERRA/AQUA land surface temperature (LST) and modification of canopy cover (CC) to predict rice yield using crop simulation model. The output is displayed via the mobile application called RiceSAP and web service, offering farmers and government officer with fully functions updated daily to monitor crop growth and conditions and predict rice yield. The web service displays rice yield prediction at sub-district level. The system can monitor spatial and temporal information of droughts and flood which updated every 15 days in order to support efficient planning on water management. The results show that yield predictions of White rice in Ayutthaya province shows 12.7% of mean absolute percentage error (MAPE), and 6.45% of MAPE for Jasmine rice in Roi Et province. For RiceSAP application, the platform operation was tested by 20 farmers in each province. It is founded that average of MAPE are 21.77% and 17.87% in Ayutthaya and Roi Et, respectively. As a result, the platform service can be extended for rice field monitoring at the country region without extra cost. Moreover, with the same approach, the platform usability can be applied for industrial crops such as maize and cassava. Therefore, this platform is useful for promoting smart agriculture sector in Thailand.