SSM/I AND AMSR-E PASSIVE MICROWAVE SATELLITE PREDICTIONS OF GLACIAL MELTWATER FLUXES AT MATANUSKA GLACIER, ALASKA

Passive microwave remote sensing observations in conjunction with detailed hydrological measurements can be used to predict timing of the annual glacial freshet (snow-melt flood). Satellite data draw from 20 years of multiple daily brightness temperature (Tb) observations of the Special Sensor Microwave Imager (SSM/I, 37-V GHz) and 4 years of Advanced Microwave Scanning Radiometer-EOS (AMSR-E, 36-V GHz). Hourly ice-proximal streamflow and suspended sediment concentrations at Matanuska Glacier are correlated to these satellite data.

The SSM/I threshold [Tb>246K, and daily difference of Tb > abs(10)K] reveals the date snow transitions from a frozen state to one of daily-melt and nightly-refreezing. AMSR-E thresholds differ by a few K which will be reported. Snow is ripe when refreezing ceases. At Matanuska the SSM/I algorithm yields a predictive relationship (R^2 0.84) between onset of ripe snow and the freshet. Typically melt-refreeze persists ~ 40-d followed by an 11-14 day lag separating ripe snow onset from the freshet. During 1995 and 2004, we observed low winter snow fall, shorter snow melt-refreeze intervals (25-d, 26-d) and longer lag times (22-d, 27-d). Lower snow volumes may account for the shorter snowmelt interval and longer lag time.

The style of glacial snow meltwater migration culminating in a freshet flood is of particular interest to this study. Two explanations exist for meltwater routing: a) supraglacial migration with little or no plumbing interaction, or b) meltwater routing to englacial and subglacial plumbing affecting plumbing connectivity. Workers at Haut Glacier d'Arolla and other glaciers favor plumbing system engagement calling this a ‘spring event.' At these glaciers, dye tracing, geophysics, subglacial drilling, ice velocity and suspended sediment responses support pluming engagement and reorganization during flood events. At Matanuska, responses of the two available years of ice velocity (1996, 1997) are consistent with ‘spring event' signatures. These data show statistically significant increases of ice velocity and subglacial sediment concentrations during the glacial freshet. We favor the interpretation that these data support a ‘spring event' and suggest that the passive microwave approach may be useful for predicting the timing of such ‘spring events.'