Hurricanes bring heavy rain and induce catastrophic flooding. The damage and fatalities underscore the urgency for understanding and improving the hydrological forecasts. Here we build an integrated hydrological framework in support of decision making, specifically for heavy rainfall caused by tropical storms. We apply different ensemble approaches for short-lived tropical storms (e.g., Tropical Storm Imelda) and long-lasting and major hurricanes (e.g., Hurricane Harvey). To drive the WRF-Hydro/National Water Model (NWM), atmospheric inputs are derived from the dynamical ensemble prediction based on Hurricane Weather Research and Forecasting (HWRF) for Hurricane Harvey. For short-lived tropical storms, which do not have operational hurricane forecast from regional dynamical models, we manually generate an ensemble forecast from a deterministic weather forecast from the Global Forecast System (GFS) and perturb the precipitation intensity and location according to the new runs from the High-Resolution Rapid Refresh (HRRR). On top of the current operational forecast from NWM, both of our approaches generate more than 20 separate forecasts (ensemble members) to address uncertainties in atmospheric dynamics, specifically for tropical storms and hurricanes. We evaluate the storm track, precipitation, and streamflow over the hurricane-prone areas of Texas. By linking ensemble weather forecasts to hydrological forecasts, we seek to provide a more comprehensive understanding of the underlying models and support advanced research on flood resilience for critical infrastructures.

In the past four decades, Texas has experienced more than 80 hurricanes, including Harvey, which alone caused damages costing over $130B. Given this history and predictions of more frequent and/or more intense storms in the future, it is of paramount importance to make prudent investment decisions to enhance the resilience of the electric grid against such extreme weather events. In this work, we explore two storm-surge models and integration of these models with an inland flooding model to create representative future flood scenarios for the state of Texas. We then discuss how these flood scenarios can further be integrated with a synthetic power system model that accurately quantifies the loss of power in all contingencies for the same geographical region, using a stochastic optimization framework. Our proposed two-stage scenario-based stochastic optimization approach helps identify substations susceptible to flooding due to storm surge and inland flooding, and recommends optimal substation hardening solutions given a finite investment budget. The insights from our work can be used to decide substation hardening strategies to enhance the electric grid’s resilience against a multitude of future storm scenarios.

Our research team is involved in several projects that seek to integrate the science-based prediction models of flood-causing events such as hurricanes with the decision-making models for critical infrastructure resilience. To this end, we use the state-of-the-art hydrological models such as WRF-Hydro and ADCIRC to simulate potential realizations of inland and coastal flooding events caused by tropical storms. We use these simulations to generate statistically sound scenarios to populate the inputs of several resilience-based decision making models, all developed using the state-of-the-art scenario-based stochastic and robust optimization methodologies. We identify three time lines where these models can be used to improve the quality of decision making processes: (1) Short-term preemptive resource allocation (preparedness) just before impending tropical storms, (2) Mid-term hardening and resilience investment strategies (mitigation) within a multi-season horizon considering multitudes of potential storms, and (3) Long-term resilience investment and infrastructure design strategy development considering potentially increasing flooding risks due to climate change and sea level rise. We present the overall framework that our team developed relying on the team’s in-progress work, particularly for the short- and mid-term prediction-optimization models. We use two specific infrastructures as examples to instantiate our models: (1) Evacuation of patients from healthcare facilities (hospitals and nursing homes), and (2) Substation hardening and preparation for power grids. To create realistic, high-resolution case studies, we consider historical and synthetic storms that impact actual healthcare facilities and power grid for Texas.

Multi-physics ensemble simulations have emerged as a promising approach to ensemble hydrological simulations due to the advantages in process understanding and model development. As a multi-physics ensemble is constructed by perturbing the physics of multi-physics models, the ensemble members share a substantial portion of the same physics and hence are not independent of each other. It is unknown whether and to what extent the independence of the ensemble members affects the ensemble skill gain, especially compared with the multi-model ensemble approach. This study compares a multi-physics ensemble constructed from the Noah land surface model with multi-parameterization options (Noah-MP) with the North American Land Data Assimilation System (NLDAS) multi-model ensemble. The two ensembles are evaluated at 12 River Forecast Centers over the conterminous United States. The ensemble skill gain is measured by the difference between the performance of the ensemble mean and the average of the ensemble members’ performance, and the inter-member independence is measured by error correlations. The results show that the Noah-MP members outperform, on average, the NLDAS models, especially in the snow-dominated areas. In addition, the best-performing models among the two ensembles are mostly Noah-MP members. However, these two performance superiorities do not lead to the superiority of the ensemble mean. The Noah-MP multi-physics ensemble has a low ensemble skill gain, resulting from a high error correlation among the ensemble members. This study suggests that the methods of ensemble construction and optimization should be improved to also consider inter-member independence, especially for a multi-physics ensemble.