13.02.2013 Views

Download the entire proceedings as an Adobe PDF - Eastern Snow ...

Download the entire proceedings as an Adobe PDF - Eastern Snow ...

Download the entire proceedings as an Adobe PDF - Eastern Snow ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Evaluation of <strong>the</strong> global distribution of retrieved SWE<br />

Figure 7 depicts example of retrievals of AMSU SWE (left) <strong>an</strong>d L<strong>an</strong>d Surface Temperature<br />

(LST) (right) in March 2002 <strong>an</strong>d 2003. Note <strong>the</strong> predomin<strong>an</strong>ce of high SWE values (depicted in<br />

red) over North-central <strong>an</strong>d E<strong>as</strong>tern Siberia, over Nor<strong>the</strong>rn C<strong>an</strong>ada <strong>an</strong>d Al<strong>as</strong>ka. Interestingly, <strong>the</strong>se<br />

patterns are similar to <strong>the</strong> ones retrieved in 2006 (Figure 4). Although global SWE data are not<br />

available to qu<strong>an</strong>titatively evaluate retrieved SWE, some general observations c<strong>an</strong> be made. For<br />

example, <strong>the</strong> incre<strong>as</strong>e in SWE from Western to E<strong>as</strong>tern Siberia may be unrealistic given <strong>the</strong> dry,<br />

low-temperature climate predomin<strong>an</strong>t in E<strong>as</strong>tern Siberia. These trends appear to be strikingly<br />

similar in 2002, 2003 <strong>an</strong>d 2006. Despite large-scale similarity, a closer visual inspection revealed<br />

some inter-<strong>an</strong>nual differences over specific regions. For inst<strong>an</strong>ce, retrieved SWE over some<br />

Western US regions, e.g., Colorado, Wyoming <strong>an</strong>d Idaho, <strong>an</strong>d over C<strong>an</strong>ada, e.g, Alberta, were<br />

higher in 2002 th<strong>an</strong> in 2003. Ano<strong>the</strong>r observed difference w<strong>as</strong> <strong>the</strong> larger extent of high SWE over<br />

Siberia in 2003 th<strong>an</strong> in 2002. This inter-<strong>an</strong>nual variability in SWE appears to follow that of LST:<br />

The lower <strong>the</strong> LST <strong>the</strong> higher <strong>the</strong> retrieved SWE. For inst<strong>an</strong>ce, lower LST in 2002 th<strong>an</strong> in 2003<br />

over Western US w<strong>as</strong> <strong>as</strong>sociated with retrieved SWE higher in 2002 th<strong>an</strong> in 2003. LST also<br />

appears to influence to a large extent <strong>the</strong> global distribution of retrieved SWE. A possible<br />

expl<strong>an</strong>ation of this temperature bi<strong>as</strong> could be <strong>the</strong> influence of temperature on <strong>the</strong> evolution of<br />

snow cover properties that impact <strong>the</strong> microwave response. As explained earlier, <strong>the</strong> AMSU SWE<br />

algorithm coefficients are static. Foster et al., 2005 report a new dynamical approach where<br />

algorithm coefficients are adjusted b<strong>as</strong>ed on a geo-referenced snow cl<strong>as</strong>sification system (Sturm et<br />

al., 1995). The authors report improved perform<strong>an</strong>ce compared to static retrievals.<br />

Figure 7. AMSU retrieval of SWE (top) <strong>an</strong>d L<strong>an</strong>d Surface Temperature (bottom) in March 2002 <strong>an</strong>d<br />

2003.<br />

99

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!