A Bayesian Hierarchical Approach for Simulating Multi-time Scale Hydrological Variability for Water Resource Planning

对未来干旱风险的评估依赖于随机模型提供的水文输入的模拟。目前的模式仅限于利用当地观测到的水文数据在单一时间尺度上模拟变率。这些数据只提供了有限的关于长期气候变率的信息，而长期气候变率是造成长期严重干旱的原因。提出的研究将开发一个新的贝叶斯框架来模拟水文数据的多时间尺度变化。这将使模拟长期变率的动态过程能够在不确定性框架中使用辅助信息进行识别。这将为水资源规划者提供对长期干旱风险更准确的评估。

Assessments of future drought risks are dependent on simulations of hydrological inputs provided by stochastic models. The current models are limited to simulating variability at a single time scale using only local observed hydrological data. This data has only limited information on the long-term climate variability which is the cause of long-term severe droughts. The proposed research will develop a new Bayesian framework for simulating multi-time scale variability in hydrological data. This will enable the dynamic processes which simulate long-term variability to be identified using auxiliary information in an uncertainty framework. This will provide water resource planners with more accurate assessments of long-term drought risks.