Enhancing Global Observation of Soil Freeze/Thaw State through Multi-scale Data Fusion



Journal Title

Journal ISSN

Volume Title



In cold regions, temperature acts as a hydrologic and ecological switch. As extended periods of sub-freezing temperatures can constrain various parts of the water cycle by preventing evaporation, infiltration, and runoff as well as defining the transition from periods of plant growth and ecologic activity to ecosystem dormancy. Hence, global knowledge of surface characteristics such as freeze/thaw state are exceedingly important for monitoring of biogeochemical cycles, such as the carbon cycle, and the implications of a changing climate. Current approaches leverage satellite observations in the microwave portion of the electro-magnetic spectrum to detect the transition between frozen and thawed states. However, investigations into the use of such observations have suggested freeze/thaw classifications remain uncertain and lack a clear physical meaning as to the type of freeze indicated (e.g., soil, surface snow and ice, frost). This project focuses on improving the characterization of freeze/thaw processes, in terms of both the observational techniques and the methods used to extract information from observations at multiple scales. To achieve this, we employ extensive analyses of existing satellite derived freeze/thaw classifications, the introduction of new predictive variables (i.e., land surface temperature and land cover), targeted field campaigns, and the integration of thousands of global soil temperature observing sites. The result, a new and robust decision tree-based classification algorithm which introduces a freeze/thaw probability metric, soil specific classifications, and improved resolution. This information is expected to provide better inputs to hydrologic models used to predict runoff and flooding, allow for more accurate estimates of the growing season, and enable assessments of the agricultural implications of changing melt and refreeze cycles. At its core, the implementation of an improved freeze/thaw classification can enhance our understanding of fundamental physical processes which occur on our planet.



Climate, Cryosphere, Freeze thaw, Global change, Remote sensing, Soil