Calibration of Crop Growth Models in Asia

DESCRIPTION: This project aims to calibrate crop growth models for rice and sugarcane in Asia, by which final information consumers will be able to carry out scenario simulations on fertilizer/risk assessment. ListenField (LF) has developed an advanced agro-intelligent platform. Sensor data are regularly collected and transferred to LF's SOS sensor web infrastructure (CloudSense). The CloudSense provides relevant data to LF's API system to generate crop model inputs for data assimilation and crop modeling. The simulation results are visualized through the LF's web service platform. Climate forecast information (e.g., from IRI or NOAA) is used to generate possible weather realizations by LF's weather generator for the coming or rest of the cropping season. The generated weather realizations will be used for crop yield prediction and risk assessment analysis. The predicted results are expected to lead to stable and high productivity for stakeholders, including farmers.

In this framework, the calibration of genetic coefficients of target cultivars is critical for reliable crop model performance. The calibration requires in-depth scientific knowledge in crop modeling. To come up with reliable and robust genetic coefficients of target cultivars, IRI will provide advice to LF and assist the LF's calibration work. LF will provide the field/local data required for the calibration to IRI. The relevant crop models embedded in Decision Support System for Agrotechnology Transfer (DSSAT) package will be used. The target crops include rice and sugarcane grown in Thailand. Other models such as SIMRIW, a rice simulation model developed in Japan, or machine learning models might be considered.