Our primary objective is to improve our ability to apply remotely sensed measures of snow properties to improve in-season estimates of snow accumulation, snow melt and runoff in remote, snow-dominated mountain regions, which supply water to 1/5th of Earth’s population. The results enable decision makers to anticipate where supply is likely to be insufficient to meet demand, or where there may be enhanced risk of floods. The project team has developed such models for a variety of users and works to engage stakeholders.
Geographic Focus
Hindu Kush (Afghanistan)
Sierra Nevada (California)
Principal Investigator
Jeff Dozier, University of California, Santa Barbara
Project Team
Robert E. Davis and Carrie Vuyovich, US Army Engineer Research and Development Center (ERDC)
Edward H. Bair, University of California, Santa Barbara
Karl Rittger, National Snow and Ice Data Center
Collaborators & Stakeholders
US Army ERDC Cold Regions Research and Engineering Laboratory (CRREL)
US Army Corps of Engineers
US Embassy (Afghanistan, Pakistan)
California Department of Water Resources
Technical Overview
Accurate assessment of water supply in its historical context of drought and flood shape foreign policy and foreign aid. Since 2004, the US Army CRREL (Cold Regions Research and Engineering Laboratory), part of the Engineer Research and Development Center (ERDC), has produced operational snow cover assessments for the major river basins in the Middle East and South Asia. In distributing its products, ERDC-CRREL forms partnerships with a diverse spectrum of planning and decision making users: American and foreign embassies, regional water ministries, US Armed Forces and allies, USGS, USDA, USAID, Department of State, National Geospatial-Intelligence Agency and others.
To improve CRREL’s ability to analyze, synthesize, and present information, the project is developing methods to estimate the volume of seasonal snow, relative to historical trends, in snow-dominated areas that have emerging or enduring insecurity related to water resources, particularly drought. Such estimates enable the identification of crises events earlier than in current practice or through comparison to historical data. Remotely sensed data provide almost all information necessary to predict seasonal and paroxysmal runoff. The project approach synthesizes optical and passive microwave imagery, topographic information, modeled precipitation and snowmelt. The project is applying an independent method to assess the snow resource during winter: snow water equivalent generated from passive microwave imagery, along with measurements of snow-covered area from optical satellite data from MODIS. The project validates these calculations with retrospective reconstructions of snow water equivalent, which match measured streamflow better than any other method and which compare well in the Sierra Nevada with measurements from the Airborne Snow Observatory, a project independent of ours but which is also supported by NASA’s Applied Sciences Program.
To demonstrate the feasibility of the approach, the project focuses on two regions, a validation case and an operational case. For validation, the project is using the Sierra Nevada of California, a mountain range of extensive historical study, emerging scientific innovation, and conflicting priorities in managing water for agriculture, urban areas, hydropower, recreation, habitat, and flood control. For the austere regional focus, the project focuses on the mountains of Afghanistan in South Asia, where some of the most persistent drought in the world causes food insecurity and combines with political instability, and occasional flooding, to affect US national interests.
Using machine learning methods, we have found we can reliably estimate the snow water equivalent during the snow season from the microwave snow water equivalent estimates and the snow-covered areas. While we are still working to improve the methods using additional sources of data from seasonal atmospheric circulation, our method as it stands now is far more accurate the currently used forecasts, such as those from the US Air Force Weather Agency.
Related Research Areas