Development And Application Of A Spatial Large-scale Simulation Driver For Crop Models

With the progress of model development and application,growing demands for geospatial application of models have led to tool development for conducting simulations spatially.It's urgent to develop efficient methods and tools for conducting simulations spatially.A model-independent,open-source tool “Geospatial Simulation”(GeoSim)has been developed previously and applied at field scale.This study advances GeoSim application for national-scale and multi-year simulations by changing the way GeoSim pass data and optimizing the management of spatial information.The widely-applied Aqua Crop model was implemented by GeoSim to simulate rice yield and irrigation requirements on a daily step across China from 2009 to 2011.We then analyzed the potential of balancing of both irrigation water and grain production under a downscaled planetary water boundary by spatial optimization of rice production,with water footprint,water scarcity index,yield gaps and harvest gaps theory.The main results are as follows:(1)By changing the way GeoSim transmits spatial data from direct basic data to data filenames,the base map files(*.shp)and template files(*.gst)required in the simulation can be simplified.(2)By identifying unique spatial information units in QGIS,the original 89,334 spatial units are reduced to 7744 spatial units,which significantly improves simulation efficiency.By post-processing the output results,the high spatial resolution of the original file can be restored.(3)The Aquacrop driven by GeoSim only took 6 hours to simulate the rice yield and water consumption(daily)in China from 2009 to 2011,which was much lower than the singlepoint simulation with Aquacrop(estimated at more than 72 hours for a single computer).(4)Analysis of the spatial distribution of rice production found that to operate within the boundary of sustainable water resources utilization,national irrigation water use for rice cropping should reduce by 10% in water-scarce regions,implying a 10% loss in national rice production without further intervention.However,using scenario analysis,we found that the production losses can be reduced to approximately 7% by closing yield gaps,and fully compensated by closing harvest area gaps in water-rich regions.The closing of both the yield and harvest area allows an increase of 6.9 million tons of rice(3% of the national production).The spatial redistribution of rice production under these scenarios resulted in a reduction in the national water-scarcity footprint related to rice cropping of 52–55%.Results show that the new method developed in this study can make GeoSim efficient manage Aqua Crop in high resolution,national scale and continuous simulation.This makes Aquacrop more suitable for the study of the "crop-water" relationship at the regional level,and can provide a method reference for GeoSim to drive other crop models in broader fields.Optimizing the spatial distribution of rice production,especially fully exploiting the potential of double-cropping paddy region with abundant water resources in southern China,is of great significance to ensuring food security and sustainable utilization of water resources in China.