Figure 2: Remotely sensed albedo evaluation for Athabasca Ice AWS.
Figure 3 shows both basins’ mid-summer remotely sensed snow and ice
albedo maps. These maps demonstrate the vast spatial variability of
intra-HRU and -basin snow and ice albedos. Such spatial variabilities
cannot be captured by modelling in scales smaller than its HRU units.
Although the intra-HRU albedo variability was averaged out before DA was
performed, the collective variations caused by wildfires and heatwaves
can generate different integrated responses. This heterogeneity suggests
that the remotely sensed albedos can contribute new information to a
cold regions DA framework. Intra-basin snow and ice albedo variability
is also observed in both basins. For example, in AGRB, there is a clear
distinction between lower glacier ice albedos and the higher albedos
from the Columbia Icefield. The transition from snow to firn albedos is
also evident in PGRB. This date also exhibits an area in which albedo
was not retrieved for the most terminal glacier HRU in PGRB. Masked
albedo values in Figure 3 can occur if clouds and shadows obstruct the
area or if the coarse resolution BRDF retrieval is unable to generate
enough observations for a particular landcover class. DA was not
performed when there were no snow and ice 20-m albedo pixels inside an
HRU, and the modelling continued with its old state variables. Note that
there were 68 and 33 dates with available remotely sensed albedo
estimates for assimilation in AGRB and PGRB, respectively.