Impact of dynamic snow density on GlobSnow snow water equivalent retrieval accuracy
<p>Snow water equivalent (SWE) is an important variable in describing global seasonal snow cover. Traditionally, SWE has been measured manually at snow transects or using observations from weather stations. However, these measurements have a poor spatial coverage, and a good alternative to in...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2021-06-01
|
| Series: | The Cryosphere |
| Online Access: | https://tc.copernicus.org/articles/15/2969/2021/tc-15-2969-2021.pdf |
| Summary: | <p>Snow water equivalent (SWE) is an important variable in
describing global seasonal snow cover. Traditionally, SWE has been measured
manually at snow transects or using observations from weather stations.
However, these measurements have a poor spatial coverage, and a good
alternative to in situ measurements is to use spaceborne passive microwave
observations, which can provide global coverage at daily timescales. The
reliability and accuracy of SWE estimates made using spaceborne microwave
radiometer data can be improved by assimilating radiometer observations with
weather station snow depth observations as done in the GlobSnow SWE
retrieval methodology. However, one possible source of uncertainty in the
GlobSnow SWE retrieval approach is the constant snow density used in
modelling emission of snow. In this paper, three versions of spatially and
temporally varying snow density fields were implemented using snow transect
data from Eurasia and Canada and automated snow observations from the United States. Snow
density fields were used to post-process the baseline GlobSnow v.3.0 SWE
product. Decadal snow density information, i.e. fields where snow density
for each day of the year was taken as the mean calculated for the
corresponding day over 10 years, was found to produce the best results.
Overall, post-processing GlobSnow SWE retrieval with dynamic snow density
information improved overestimation of small SWE values and underestimation
of large SWE values, though underestimation of SWE values larger than 175 mm
was still significant.</p> |
|---|---|
| ISSN: | 1994-0416 1994-0424 |