Simulating the Potential of Water Harvesting Technology in Rainfed Maize Using DSSAT.

Researchers using crop modeling tools often find themselves limited to simulate management practices that are already available in the software and face challenges to study the potential impact of new technologies that are not yet implemented. Advanced researchers with computer programming skills and good understanding of the codebase of models may modify the source code and implement such technologies, but this is often a rather risky and expensive proposition unless the model is simple and you receive direct support from the core model development team; seemingly small changes in the modeling code or parameters may cause unexpected changes or mistakes. Instead, we propose to use the model as-is but rather create an external driver program to use the entire model as a function to simulate the cropping system. As an example of this concept, water harvesting technology was implemented on DSSAT Cropping System Model. The driver program first runs DSSAT with no water stress to record phenology, runs again under rainfed and records the water balance including daily runoff water from the fields and decide when the additional water harvested from the runoff will be needed the most, and finally runs again under rainfed but with additional irrigation on dates when it’s needed with the amount supposedly collected from the water harvesting technology. This technique was implemented at global scale and used as one of the key technologies to address food security under scarce natural resources in a recently published integrated assessment study by IFPRI, estimating the regionally aggregated potential of increasing maize yield of up to 10% under future climate scenarios in 2050. This approach can allow researchers to study the potential of new technologies that are not yet implemented in the model and stimulate creative use of crop systems modeling tools beyond what they offer out of the box.