Characterizing Spring Snowmelt in the Boreal Forest of Interior Alaska

Spring snowmelt is the dominant driver of watershed hydrology in the boreal forest, the world’s largest biome, and the characteristics of snowmelt timing can strongly impact a number of key ecosystem functions. This study uses an approach to model spring snowmelt response curves based on remote satellite imagery and define the characteristics of the melt timing across study sites located in the boreal forest region of Interior Alaska. Non-linear methods are used to characterize the melt timing and response at multiple sites. The non-linear curves are reconstructed using climate indices across two separate time periods – one for which remote satellite imagery is available (2000-2010) and an alternate 11-year period (1989-1999) prior to availability of the data. Validation is performed against observational data recorded at each of the 38 sites analyzed. Results indicate that there is a range of melt dates and maximum melt periods exhibited across the 11-year record, ranging across approximately 15 days – from April 4th to April 18th, with a median melt date at all sites of April 10th. The key climate drivers used to estimate the logistic parameters and reconstruct the non-linear curves were primarily winter and early spring temperature, freezing and thawing indices, and the date of spring radiation increase. Wind speeds, precipitation and relative humidity were considered secondary influences on the melt parameters used to estimate the snow melt trajectory. The alternate time period and the ‘forecast’ year 2011 were estimated with success, suggesting that this technique might be useful to address the lack of overlap between time periods and sensors, for forecasting applications and to extend the length of the relatively short term remote sensing products.