The AI-soil map and Soil-Environmental application programming interface for visualization of fertilization effect of each agricultural field

Abstract

The data visualization of the fertilization effect of organic or inorganic fertilizer on a field-by-field basis can help cut fertilization costs, and early realization of a 30% reduction in chemical fertilizer use, as in the KPI of Strategy for Sustainable Food Systems, MIDORI. A high-resolution soil map is required to visualize fertilizer efficacy for each field. We developed a high-resolution soil map named “AI-soil map” covering Japanese agricultural land and a Soil-Environmental API (Application Programming Interface). This API estimates soil temperature and moisture for each field based on AI-soil map and weather data, predicts the fertilizer efficacy of slow-release fertilizers and organic materials daily based on the estimated values, and delivers the fertilizer efficacy prediction data. The Soil Environment API consists of a soil temperature and moisture estimation API, a slow-release fertilizer nutrient supply API, and an organic material fertilizer efficacy visualization API. In collaboration with public research institutions in 13 prefectures, tests were conducted on 22 field crops to confirm the possibility of reducing chemical fertilizers (target 30% reduction) by visualizing the fertilizer efficacy of organic materials using the “Soil-Environmental API”. It has been demonstrated that chemical fertilizers can be reduced by an average of 45% while maintaining crop yields. The “AI-soil map” and the “Soil-Environmental API” are opened by the agricultural data distribution platform, WAGRI.