Characterizing Historical, Current, and Future Hydrological Variability Through The Development of Meteorological and Hydrological Ensemble-Based Datasets

Current hydrological datasets used for long-term planning are based primarily on the historical period of observations or statistics derived from those observations, particularly for flood magnitude and flood frequency estimation. Supplementing these datasets with synthetic hydrology typically uses statistical scaling applied to historical precipitation or streamflow events wherein prescribed volumes corresponding to lower frequency events are derived from statistical extrapolation of the historical record. This expands the representation of large events but does not change the representation of hydrological variability. Furthermore, often sequences of events affecting the hydrological and reservoir system states are more important to system vulnerability than the impacts of individual events, even the more extreme events. To represent a wide range of potential meteorological variability we apply output from CMIP6 earth system models, regional numerical weather modeling (ICAR), and hydrological modeling (SUMMA) to simulate a wide range of potential conditions for the period of 1950-2099 for the Pacific Northwest region. Using 10,000+ of years of continuous simulation in this ensemble we characterize potential variability and extremes throughout this 150-year period and separate the contributions of natural variability and climate change. This data development effort is supporting vulnerability assessments and resilience planning for water resource infrastructure of USACE and its partners in the Pacific Northwestern United States.