Determination of melt-onset and refreeze timing on southeast Alaskan icefields using SSM/I diurnal amplitude variations

Timing of snowmelt and freeze-up was estimated for glaciers in the Coast and St Elias Ranges of Alaska, U.S.A., and British Columbia, Canada, using twice-daily brightness temperatures (T-b) from the U.S. Defense Meteorological Satellite Program's Special Sensor Microwave/Imager (SSM/I). Melt and freeze-up were determined for a 37 GHz vertically polarized time series using changes in the average daily T-b and high-amplitude T-b diurnal amplitude variations (DAV). DAV are the running difference between the early-morning (usually minimum) and late-afternoon (usually maximum) T-b observations. Year-round temperatures taken at 2 hour intervals oil the Juneau Icefield (58degrees4'N, 134degrees15'W) validated the microwave response to melt. A bimodal distribution of T-b corresponding to frozen or melting snow helped estimate the Tb at which the transition from frozen to melting snow occurred on pixels without ground observations. Thresholds of T-b > 246 K) and DAV (> 10 K) were used to refine the selection of melt and refreeze timing for southeast Alaska. Melt timing correlates with stream discharge. In general, melt onset occurred progressively earlier and refreeze later in the season between 1988 and 1998. It is not known whether this is related to regional warming or to one of the shorter decadal-scale oscillations in the Gulf of Alaska.