Proceedings of the 2023 Mississippi Water Resources Conference

Year: 2023 Authors: Khanel P., Feng G., Huang Y., Han M.

Estimation of crop yield with approximate acreage is critical to crop production management decision support. This research aims to estimate the soybean yield and acreage during the wet, normal and dry years in Mississippi State using APEX (Agricultural Policy/Environmental eXtender) model which was developed to extend EPIC (Environmental Policy Integrated Climate model)'s capabilities of simulating management and land use impacts for whole farms and small watersheds (https://blackland.tamu.edu/models/apex). APEX model uses data from the past to establish trends such as rainfall average, precipitation, soil, wind etc. to determine how the climate trend, management practices, and soil can affect the future of agriculture. APEX model is capable of long-term simulation of the fields ranging from small-scale to cross-country boundaries. For the research, APEX model was implemented for all soybean fields county by county across the entire state. The results from the research provide valuable information on the total area of soybean planted and the total yield of specific soybean fields in every county in Mississippi. The APEX model was implemented by utilizing various databases, including data of soil, weather, crop, field management, and spatial locations. The database provided reference for APEX model to determine which data to run for each simulation. The Gridded Soil Survey Geographic Soil Map (gSSURGO-30) (USDA Natural Resources Conservation Service (NRCS)) and the Gridded Cropland data layer (CDL-30 m) (USDA National Agricultural Statistics Service (NASS)) were downloaded and then uploaded into ArcGIS to apply Spatial Analyst tools in ArcGIS (ESRI, Redlands, CA) to overlay and clip the two data layers to fit into the area of interest. We then conducted georeferencing to get the geographic coordinates of the individual fields in the area of interest. In ArcGIS, we then used the SSURGO's table to link the field with type of soil in field. We then created the individual fields based on the soil, field size, and geographic coordinates using parameters such as type of soil found on the field, geographic coordinate, crop to be run, management file and weather station near to the field. Accordingly, weather data, such as wind, precipitation , daily/monthly temperature, and soil data, were converted into the required format for APEX model with Python programming. As a result, soil, and weather data files for all soybean fields in the entire state were created and the area and location of soybean fields was determined. Those data files were used by APEX model to simulate soybean yield for individual fields and the total yield for all fields in each county of the entire state. The output from APEX model was compared with the total area of soyabean field planted and yield data reported by the USDA NASS. The study will contribute to improving accuracy of those NASS reports for stakeholders in commercial community and researchers in scientific community.