Automation of Daisy and Crop Yield Optimization

Daisy is a soil-plant-atmosphere system model that is designed to simulate water balance, heat balance, solute balance and crop production in agro-ecosystems [1]. All these process are dictated by the various management strategies by the farmer and this is an important driving variable in model. Weather data, the other important driving variable, is provided by in-house IBM research project called Deep Thunder that models hyper-local, short-term forecasting [2]. In addition to these driving variables, the user is required to define the relevant soil data, vegetation and other data parameters that dictate the physical process of the agro-ecosystem.

Daisy requires few user inputs to simulate crop-production in agro-ecosystem. Daisy generates simulation results based on historical data of important variable and provide crop yield and influencing factor in production. Daisy is most compatible with windows operating system so one of the project objective was to automate daisy to make it platform independent. Along with daisy automation, Our objectives were to identify
i) Optimal crop management strategies,
ii) Optimizing external factors
iii) Understanding crop growth behaviour with driving variables
This project was designed to run for 6 weeks only to meet academic requirements, and so our initial focus was on automation and then move to optimization use cases.