Reducing Uncertainty of Continuous Streamflow Predictions in Ungauged Basins (PUB) Using Regional Constraints: Using Regional Constraints for PUB

Our environment is changing rapidly in many regions of the world, with often-significant implications for current and future water resources. Disturbances to the current environmental system include land cover change, and the increasing impacts of a changing climate. The potential of developing successful strategies to respond to these changes lies in our capacity to anticipate their impact. One important question in this context is the potential change to streamflow characteristics at the watershed scale. However, most watersheds in the US and elsewhere are ungauged, leading to large uncertainty in our hydrological predictions to establish benchmarks, which are often the basis for the development of water resources management strategies. How can we obtain a reliable and continuous spatio-temporal picture of regional environmental status, trends and long-term projections to identify information gaps — and thus monitoring needs — and likely hot spots and hot moments where for example critical ecosystem thresholds are exceeded or will be exceeded in the future? This paper presents a methodology to formalize regional information on streamflow characteristics using a Mixture of Gaussians approach, which provides constraints on continuous watershed simulations. US examples show how this approach leads to reduced uncertainty in continuous streamflow predictions for water management in ungauged basins and in terms of potential change impacts.