Scholarlink

<p>The increasingly scarce world water resources require proper planning and management. This study was designed to estimate total crop water requirements (CWR), gross and net irrigation water requirements (GIWR and NIWR), actual irrigation water requirements (AIWR) and the moisture deficit at harvest (MDH) for 27 irrigation scenarios under three cropping systems using CROPWAT 8.0 and surrogate machine learning (ML) algorithms as an alternative. CROPWAT software was employed to simulate CWR, GIWR, NIWR, AIWR and MDH for maize, soybeans, and sweet potatoes. Simulations were based on 27 irrigation scenarios; formed by combining different irrigation water application levels (FC)), allowed water depletion levels (Depletion), and irrigation efficiencies (Eff). Estimated CWRs were cross checked against potential evapotranspiration (PET) data from the Food and Agricultural Organization Water Productivity Open Access Portal ((FAO WAPOR). Five surrogate linear ML models were also used to estimate irrigation metrics from CROPWAT simulations across all scenarios. Estimated CWR values ranged from 378.7 to 396.2&#xa0;mm for soybeans, 508.8 to 528.3&#xa0;mm for maize, and 583.6 to 619.5&#xa0;mm for sweet potatoes, while NIWR varied between 93.5 and 654.2&#xa0;mm across cropping systems. Varieties of linear ML models captured the simple relationship between scenario parameters and IWRs or MDHs across cropping systems, with NSE ranging from 0.62 to 0.99. Findings from this study highlight the importance of efficient irrigation practices to optimise water use, reduce resource strain, and also support sustainable agriculture, especially in regions experiencing water shortages.</p>

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

Scenario-based Irrigation Water Requirement Modelling using CROPWAT Simulations and Surrogate Linear Machine Learning

  • 期刊论文

  • 发表日期：

    2026年5月8日

查看全文
  • Blessing Funmbi Sasanya,
  • Victory Chinaza Ikewuibe

摘要

The increasingly scarce world water resources require proper planning and management. This study was designed to estimate total crop water requirements (CWR), gross and net irrigation water requirements (GIWR and NIWR), actual irrigation water requirements (AIWR) and the moisture deficit at harvest (MDH) for 27 irrigation scenarios under three cropping systems using CROPWAT 8.0 and surrogate machine learning (ML) algorithms as an alternative. CROPWAT software was employed to simulate CWR, GIWR, NIWR, AIWR and MDH for maize, soybeans, and sweet potatoes. Simulations were based on 27 irrigation scenarios; formed by combining different irrigation water application levels (FC)), allowed water depletion levels (Depletion), and irrigation efficiencies (Eff). Estimated CWRs were cross checked against potential evapotranspiration (PET) data from the Food and Agricultural Organization Water Productivity Open Access Portal ((FAO WAPOR). Five surrogate linear ML models were also used to estimate irrigation metrics from CROPWAT simulations across all scenarios. Estimated CWR values ranged from 378.7 to 396.2 mm for soybeans, 508.8 to 528.3 mm for maize, and 583.6 to 619.5 mm for sweet potatoes, while NIWR varied between 93.5 and 654.2 mm across cropping systems. Varieties of linear ML models captured the simple relationship between scenario parameters and IWRs or MDHs across cropping systems, with NSE ranging from 0.62 to 0.99. Findings from this study highlight the importance of efficient irrigation practices to optimise water use, reduce resource strain, and also support sustainable agriculture, especially in regions experiencing water shortages.