Authors: Martyn Clark – University of Calgary, Canada
Title: Probabilistic predictions across large geographical domains
Presentation Type: Lightning Talk
Abstract: Actionable hydrologic predictions have the capability to support decisions on both improving water sustainability and reducing the impacts of water-related hazards. On longer time scales of decades to centuries, predictions provide quantitative statements of change to support efforts in infrastructure design, environmental stewardship, and long-term resource allocation. These longer-term predictions are critical for applications such as transboundary water cooperation, hazard mapping, and planning nature-based solutions. On shorter time scales (e.g., lead times from seconds to seasons), hydrologic predictions provide advance warning of floods and threats to drinking water as well as information on drought and ecological stress. They also support decisions about planting crops, planning for wildfires, supplying drinking water, managing urban drainage networks, and optimizing hydropower production, as well as forecasting ecological risks such as algal blooms, industrial and urban wastewater overflows, and other critical water quality issues. These key prediction challenges can exceed the capability of entire modeling groups, entire agencies, and, in some cases, entire countries.
The presentation will summarize how recent advances in predicting the dominant hydrologic processes across large geographical domains. The presentation will focus on new capabilities to select models across diverse landscapes to improve simulations of dominant processes, the capabilities to quantify unc