The unprecedented snowfall during early February 2010 in the Baltimore/Washington area provided a unique
opportunity to map, monitor and measure snowfall, snow cover extent, snow water equivalent (SWE), and snow
melt using a suite of remote sensing instruments. Because snow cover in the Middle Atlantic area of the United States is
in most years patchy and a true multi-layered snow pack is rarely established, utilizing a remote sensing approach to
observe snow parameters is more challenging than in regions where falling snow and snow packs are more reliable. The
Advanced Microwave Scanning Radiometer for EOS (AMSR-E) was used to assess SWE and the onset of melt.
Although the passive microwave signatures illustrated in this study are clearly related to snow, it is not straightforward
whether or not the signatures are due to variations in SWE or to snowpack metamorphism or to a combination of both.
This study shows that the SWE algorithm was affected by the high variability of snowfall intensity and accumulation as
well as by the complex surface features in the Baltimore/Washington area. On the two days when intense snowfalls
occurred, February 6 and 10, 2010, retrievals of SWE were compromised. This was likely a result of thermal emission
from water droplets in low-level clouds within portions of the storm, which acted to increase AMSR-E Tbs, thereby
rendering minimal or zero values for SWE. The presence of such clouds strongly impacts the sensitivity of estimating
SWE using radiometric measurements near 19 and 37 GHz. Glaze or icy layers within and on the surface of the
snowpack served to increase scattering, thus lowering Tb and boosting the retrieved SWE values, resulting in an
overestimation of SWE, first in southern portions of the study area and then farther north as the month of February
progressed.
|