A data-driven approach provides simulation and forecast models based on historical observation, including statistical and Machine Learning(ML)/Artificial Intelligence(AI) approaches. Traditional statistical models have oriented toward mathematical regression based on data’s statistical relationships. In contrast, the AI models have begun to permeate the hydrologic community for solving nonlinear regression/classification problems with a blooming of different ML algorithms. In the current stage, the ML/AI models need to go further to reach the practically available service levels for supporting water resources management and decision-making. This session invites contributions enriching the statistical and ML/AI applications, including:
(1) Applying statistical and ML/AI models to forecasting hydrologic data (e.g., precipitation, streamflow, temperature, climate impacts).
(2) Introducing new ML/AI technologies that concurrently account for the spatial and temporal structure of data.
(3) Assessing persistent predictability of ML/AI models and limitations of conventional approaches.
(4) Uncertainty analysis and quantification.