Hydrology and hydroclimatology
Climate change impacts
Spatial big data analytics
Projection of high-resolution climate information
Since precipitation, temperature and potential evapotranspiration (PET) are the most important components of the hydrological cycle, the high-resolution projections of future changes in precipitation, temperature and PET play a crucial role in examining the changes in hydrological characteristics caused by the human-induced climate change. In this project, we produce the high-resolution projections of future changes in climate variables such as precipitation, temperature and PET over Texas through dynamical downscaling using the Weather Research and Forecasting (WRF) model.
Wang et al. (2019), Clim. Dyn. 53, 1613–1636. [PDF]
Assessment of multivariate drought characteristics in a changing climate
Probabilistic projections of future drought characteristics play a crucial role in climate change adaptation and disaster risk reduction. In this project, we develop the copula‐based high‐resolution projections of future changes in multivariate drought characteristics, in which a probabilistic multivariate drought index is introduced to examine the joint effects of the soil moisture deficit (agricultural drought) and the runoff deficit (hydrological drought) across different temporal scales. The research findings are useful for strengthening resilience to the climate‐induced drought hazard and in facilitating sustainable agricultural development and water resources planning in a changing climate.
Development of a unified data assimilation framework for improving the robustness of hydrological ensemble prediction
Data assimilation using the ensemble Kalman filter (EnKF) has been increasingly recognized as a promising tool for probabilistic hydrological predictions. In this project, we develop a unified data assimilation framework for improving the robustness in ensemble streamflow predictions. Statistical pre-processing of assimilation experiments is conducted to identify the best EnKF settings with maximized predictive performance. When the posterior distributions of hydrological model parameters are estimated, statistical post-processing analysis is then performed to efficiently quantify predictive uncertainties. In addition, the Gaussian anamorphosis is performed to build a seamless bridge between data assimilation and uncertainty quantification through transforming posterior parameter distributions into normal distributions. Such a unified computational framework improves the robustness of sequential data assimilation by using statistical pre- and post-processing techniques, and strengthens our capability in providing probabilistic streamflow predictions.
Wang et al. (2018), Water Resour. Res. 54, 2129–2151. [PDF]