Wheat yield prediction is critical for ensuring food security and efficient resource planning in agriculture. However, selecting and validating the most suitable crop simulation model under diverse agro-climatic conditions remains a significant challenge. Chhattisgarh, a state in central India, features varied soil types and climatic zones, making ideal region for testing crop models. The Agricultural Production Systems Simulator (APSIM) and Decision Support System for Agro-technology Transfer (DSSAT) crop models were used for wheat simulation using regional datasets, and compare their performance using key statistical metrics like R2 and RMSE. Intent of this study to determine the which model most effective for wheat yield simulation for Chhattisgarh state of India. Temperature and rainfall were used as primary input for model setup and 2009 to 2014 was used for calibration while 2015 to 2018 was used for validation. It showed lower RMSE values (74.36 kg/ha and 96.36 kg/ha) compared to APSIM (86.49 kg/ha and 103.97 kg/ha), indicating closer alignment with observed data. DSSAT also achieved better R2 (0.89 and 0.75), highlighting its ability to explain yield variability and model accuracy more effectively. The models can be extended to simulate yields under various climate scenarios.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Comparative Analysis of DSSAT and APSIM Crop Models for Wheat Yield Prediction in Chhattisgarh State of India

  • Shashi Bhushan Kumar,
  • Suyog Balasaheb Khose,
  • Rajul Upadhyay,
  • Ashok Mishra

摘要

Wheat yield prediction is critical for ensuring food security and efficient resource planning in agriculture. However, selecting and validating the most suitable crop simulation model under diverse agro-climatic conditions remains a significant challenge. Chhattisgarh, a state in central India, features varied soil types and climatic zones, making ideal region for testing crop models. The Agricultural Production Systems Simulator (APSIM) and Decision Support System for Agro-technology Transfer (DSSAT) crop models were used for wheat simulation using regional datasets, and compare their performance using key statistical metrics like R2 and RMSE. Intent of this study to determine the which model most effective for wheat yield simulation for Chhattisgarh state of India. Temperature and rainfall were used as primary input for model setup and 2009 to 2014 was used for calibration while 2015 to 2018 was used for validation. It showed lower RMSE values (74.36 kg/ha and 96.36 kg/ha) compared to APSIM (86.49 kg/ha and 103.97 kg/ha), indicating closer alignment with observed data. DSSAT also achieved better R2 (0.89 and 0.75), highlighting its ability to explain yield variability and model accuracy more effectively. The models can be extended to simulate yields under various climate scenarios.