# Calibration and Validation of the CSM-CROPGRO-Peanut Model Under Mulched Drip Irrigation Conditions in Xinjiang

## TL;DR

This study tested a peanut growth model under mulched drip irrigation in Xinjiang, finding that it works best with full irrigation and proper nitrogen application.

## Contribution

The study identifies the optimal parameter estimation protocol for the CSM-CROPGRO-Peanut model under mulched drip irrigation in Xinjiang.

## Key findings

- Parameters like FL-SH, FL-SD, and WTPSD showed high variability due to environment-management interactions.
- Scenario 3 provided the highest model accuracy with ARE and nRMSE of 9.1% and 10.1%, respectively.
- The model effectively simulated peanut growth and soil moisture under full irrigation conditions.

## Abstract

In order to explore the applicability of the peanut growth simulation model CSM-CROPGRO-Peanut under conditions of mulched drip irrigation in Xinjiang, and to determine the optimal scenario for parameter estimation and model validation, field experiments were conducted in 2022 and 2023 on the water and nitrogen regulation of peanut. Based on the water requirements during the stages of peanut growth, three irrigation levels (low, medium, and high) and two nitrogen application levels (100% N and 50% N) were set, resulting in six treatments. An additional control treatment (CK) with a medium irrigation level and no nitrogen application was also included. In this study, four different parameter estimation and validation protocols were designed, and different parameter estimation results were obtained using the DSSAT-GLUE parameter estimation module. The results showed that the FL-SH (time between first flower and first pod), FL-SD (time between first flower and first seed), SIZLF (time between first flower and first seed), XFRT (maximum size of full leaf), and WTPSD (maximum weight per seed) parameters exhibited strong variability, with coefficients of variation of 24.33%, 22.9%, 19.78%, 14.47%, and 23.82%, respectively, and were significantly affected by environment–management interactions. Other parameters showed weaker variability, with coefficients of variation that were all less than 10%. The model outputs varied significantly among different parameter estimation protocols. Scenario 3, which used data from the adequate irrigation and adequate fertilization treatment (W3N2) environment across both years for parameter estimation and data from other treatments for validation, showed the highest model calibration and validation accuracy. The average absolute relative error (ARE) and normalized root mean square error (nRMSE) for model calibration and validation were the lowest at 9.1% and 10.1%, respectively. The CSM-CROPGRO-Peanut model effectively simulated peanut growth and development as well as soil moisture dynamics under mulched drip irrigation conditions in Xinjiang, with the highest simulation accuracy observed under full irrigation conditions. The findings provide a basis for using the CSM-CROPGRO-Peanut model to develop suitable irrigation and nitrogen application regimes for peanuts under mulched drip irrigation in Xinjiang.