干旱区生态环境知识资源中心(Arid): Estimating distributed autumn irrigation water use in a large irrigation district by combining machine learning with water balance models

Arid

DOI	10.1016/j.compag.2024.109110
	Estimating distributed autumn irrigation water use in a large irrigation district by combining machine learning with water balance models
	Qian, Ximin; Qi, Hongwei; Shang, Songhao; Wan, Heyang; Wang, Ruiping
通讯作者	Shang, SH
来源期刊	COMPUTERS AND ELECTRONICS IN AGRICULTURE
ISSN	0168-1699
EISSN	1872-7107
出版年	2024
卷号	224
英文摘要	Autumn irrigation (AI) after crop harvesting is vital for agricultural production in cold arid regions, typically requiring more water than regular irrigation during crop growth phases. Understanding distributed AI water use (AIWU) is key to optimizing irrigation strategies and refining agro-hydrological simulations. However, field -scale monitoring of irrigation water use is challenging, particularly in large districts, and quantification methods for distributed AIWU have seldom been investigated. Given this, we proposed a remote sensing -based model to estimate AIWU from water balance analysis, with key components derived through machine learning algorithms. Specifically, to address the unique challenge of short water ponding duration during AI, we selected MODIS as the primary data source, supplemented with Sentinel -2 data to enrich field -level details. A Sentinel -2 empirical bathymetry model, informed by in -situ measurements, served as a foundation for constructing a MODIS bathymetry model using random forest. This MODIS bathymetry model was essential for determining water ponding depths, which allowed for accurate estimation of infiltration rates when applied to the MODIS-based water index time series. By integrating these estimates into the water balance framework, we achieved a more precise estimate of AIWU. The proposed model was applied to Hetao Irrigation District (HID), the largest irrigation district in the arid region of China. The estimated steady infiltration rate and canal efficiency align with similar studies. The results indicate that the average AIWU per unit AI area in the HID remains stable during 2010 - 2020 with substantial spatial variability at the pixel level, providing new insights into the local hydrological cycle. The AIWU of HID has been decreasing by 19 million m 3 per year during this period mainly due to the reduction in AI area, with a more prominent trend from 2010 to 2017. Our study deepens insights into the spatial - temporal dynamics of AIWU in arid regions, contributing to better irrigation management.
英文关键词	Autumn irrigation Irrigation water use Water balance model Bathymetry model MODIS time series Arid agricultural region
类型	Article
语种	英语
收录类别	SCI-E
WOS记录号	WOS:001253200900001
WOS关键词	SHALLOW-WATER ; GROUNDWATER ; HETAO ; INFILTRATION ; BATHYMETRY ; CHINA ; FIELD ; PERFORMANCE ; SENTINEL-2 ; SIMULATION
WOS类目	Agriculture, Multidisciplinary ; Computer Science, Interdisciplinary Applications
WOS研究方向	Agriculture ; Computer Science
资源类型	期刊论文
条目标识符	http://119.78.100.177/qdio/handle/2XILL650/403257
推荐引用方式 GB/T 7714	Qian, Ximin,Qi, Hongwei,Shang, Songhao,et al. Estimating distributed autumn irrigation water use in a large irrigation district by combining machine learning with water balance models[J],2024,224.
APA	Qian, Ximin,Qi, Hongwei,Shang, Songhao,Wan, Heyang,&Wang, Ruiping.(2024).Estimating distributed autumn irrigation water use in a large irrigation district by combining machine learning with water balance models.COMPUTERS AND ELECTRONICS IN AGRICULTURE,224.
MLA	Qian, Ximin,et al."Estimating distributed autumn irrigation water use in a large irrigation district by combining machine learning with water balance models".COMPUTERS AND ELECTRONICS IN AGRICULTURE 224(2024).