| 摘要: |
| 综述河道外需水量影响因素及预测方法，为水资源规划与管理提供支撑。采用Cite Space分析河道外需水量的影响因素和预测方法。人口、水价、降水量和气温是河道外需水量预测研究中普遍考虑的影响因素；机器学习在河道外需水量预测中的应用是当前的研究热点；机器学习和回归分析常考虑多个影响因素，定额指标法常用于实际工程管理，时间序列法多与各类方法组合修正以弥补自身的局限性，组合预测模型能够获得较高的预测精度。随着遥感大数据、人工智能和机器学习算法的发展，精细空间尺度的河道外需水预测成为可能，未来需进一步加强需水量预测结果合理性的论证和历史数据质量的控制。 |
| 关键词: 需水量预测；Cite Space；影响因素；机器学习 |
| DOI：10.13522/j.cnki.ggps.2024155 |
| 分类号: |
| 基金项目: |
|
| A review of the determinants and prediction methods for off-channel water demand |
|
HE Yanhu, XU Xiaodi
|
|
School of Ecology, Environment and Resources, Guangdong University of Technology, Guangzhou 510006, China
|
| Abstract: |
| Off-channel water demand and consumption are critical components of water resource management, influenced by various natural and anthropogenic factors. This study systematically analysed these influencing factors and the methods for predicting off-channel water demand, aiming to provide insights for effective water resource planning and management at the catchment level. The research employed Cite Space for bibliometric analysis to identify current research trends, key hotspots, and systematically categorised the factors that influence the off-channel water demand, as well as the methods used for its prediction. Key factors affecting off-channel water demand in most catchments include population, water pricing, precipitation, and air temperature. Machine learning algorithms have emerged as a prominent tool for predicting off-channel water demand, often used alongside regression analysis to assess the influence of multiple factors. The quota index method remains widely applied in practical water resource management. Additionally, hybrid approaches, combining time series analysis with other methods, address limitations in standalone models and enhance prediction accuracy. Advances in remote sensing, geospatial big data, artificial intelligence, and machine learning algorithms have significantly improved the accuracy of off-channel water demand predictions, particularly at smaller scales. Future research should focus on enhancing the validation of prediction models and ensuring the robust integration of historical data to improve modeling reliability for off-channel water demand and consumption. |
| Key words: water demand prediction; Cite Space; influencing factors; machine learning |
