| 引用本文: | 宋 鹏,张羽翔,顾 涛,等.多源遥感数据时空融合在灌区用水管理中的应用与进展[J].灌溉排水学报,2026,45(4):119-124. |
| SONG Peng,ZHANG Yuxiang,GU Tao,et al.多源遥感数据时空融合在灌区用水管理中的应用与进展[J].灌溉排水学报,2026,45(4):119-124. |
|
| 摘要: |
| 灌区是保障国家粮食安全和农业可持续发展的核心区域，其用水管理直接关系到农业用水效率与区域水资源承载力。然而，由于卫星平台功能与运行模式差异，多源遥感数据在时间、空间分辨率、观测尺度及干扰敏感度等方面存在不一致性，限制了信息获取的精度与时效性。本研究系统梳理了国内外多源遥感时空融合技术的发展进展，从统计融合、像元重构、时空融合、机器学习、深度学习及物理-数据融合等方面总结典型方法及其原理特征，并分析不同模型在精度、稳定性与适用性上的差异。在此基础上，综述了多源融合技术在灌溉面积识别、土壤含水率监测、蒸散发估算、灌溉需水分析、用水效率评估及动态调度中的典型应用，归纳了配准一致性、模型泛化性、干扰适应性及不确定性量化等关键问题。最后，提出面向智慧灌区的未来发展方向，研究可为多源遥感融合在灌区智能化与高效用水管理中提供理论支持与技术参考。 |
| 关键词: 遥感数据融合；灌区智慧管理；蒸散发估算；灌溉面积识别；土壤含水率监测 |
| DOI：10.13522/j.cnki.ggps.2025279 |
| 分类号: |
| 基金项目: |
|
| Applications and advances in multi-source remote sensing spatiotemporal data fusion for irrigation district water management: A review |
|
SONG Peng, ZHANG Yuxiang, GU Tao, CHANG Mingqi, XU Lei, LEI Bo, LU Wenhong, HOU Peng
|
|
1. China Irrigation and Drainage Development Center, Beijing 100054, China; 2. Zhengzhou University, Zhengzhou 450001, China
|
| Abstract: |
| Irrigation districts play a crucial role in food security and sustaining agricultural development. In recent decades, remote sensing technologies have been increasingly used to improve water management in irrigation districts. However, differences in satellite platforms and their operational mechanisms often lead to inconsistencies in spatiotemporal resolution, observation scale, and sensitivity to environmental changes among datasets derived from different remote sensing sources, thereby limiting the accuracy, consistency and timeliness of information acquisition. To overcome these limitations and support more effective irrigation water management, this paper systematically reviews recent advances in multi-source remote sensing spatiotemporal fusion technologies and their applications in irrigation district management over the past decade. The review synthesizes representative methods and their underlying principles, including statistical fusion, pixel reconstruction, spatiotemporal fusion models, machine learning, deep learning, and physics-data hybrid modeling. The strengths and limitations of these approaches are comparatively analysed in terms of accuracy, stability, and applicability. Building on this synthesis, we further examine the applications of multi-source data fusion in irrigation district management, including irrigation area mapping, soil moisture monitoring, evapotranspiration estimation, irrigation water demand analysis, water use efficiency assessment, and dynamic irrigation scheduling. Through this comprehensive review, several key challenges are identified, including data registration consistency, limited model generalization across regions and sensors, sensitivity to environmental interference, and insufficient uncertainty quantification in remote sensing-based irrigation management. Finally, future research directions are proposed, emphasizing the integration of advanced spatiotemporal fusion algorithms, improved uncertainty assessment, and the development of smart irrigation districts supported by intelligent remote sensing systems. |
| Key words: remote sensing data fusion; smart irrigation management; evapotranspiration estimation; irrigation area identification; soil moisture monitoring |
