Geoscience Reference
In-Depth Information
expected to provide most accurate estimates of snow conditions (Pullen et al.
2011
; Essery
et al.
2009
; Drusch et al.
2004
; Brasnett
1999
).
2.2 Snow Observations
Snow data assimilation strongly relies on snow depth ground observations (Drusch et al.
2004
; Brasnett
1999
). A major source of snow depth measurements is that provided by
SYNOP stations (synoptic reports). These observations are available in near-real time
(NRT) on the Global Telecommunication System (GTS), so they are suitable for NWP
applications. In addition to SYNOP reports, most weather services maintain national snow
depth measurements networks. For example, the SNOTEL (SNOwpack TELemetry) net-
work provides snow depth measurements used in the NOAA (National Oceanic and
Atmospheric Administration) National Weather Service's National Operational Hydrologic
Remote Sensing Center (NOHRSC) SNOw Data Assimilation System (SNODAS). The
NOAA COoperative Observer Program (COOP) also provides snow depth measurements
over North America. However, data gathered from National Networks are not available on
the GTS, and therefore, they are not suitable to be used in NWP snow analysis systems. In
Europe, several countries are currently making available their snow depth measurements to
the NWP community. The Swedish Meteorological and Hydrological Institute was the first
to release its national network snow depth data on the GTS from December 2010. These
data have been assimilated at ECMWF since March 2011 (de Rosnay et al.
2011a
). Ground
measurements of snow depth provide a very accurate local information, however, because
of the variability of land surface and meteorological conditions, their representativeness can
be limited, particularly in heterogeneous and in mountainous areas. Besides, many areas are
sparsely observed (e.g., large areas in Siberia). Based on comparisons between pointwise
SYNOP snow depth data and snow survey data sets, Takala et al. (
2011
) estimated the
uncertainty of SYNOP snow depth data to be close to 0.12 m.
Satellite observations provide spatially integrated measurements with global coverage
which makes them of high interest to provide consistent snow information for climate and
NWP communities. SWE products based on passive microwave measurements, for
example, from AMSR-E (Advanced Microwave Scanning Radiometer for Earth Observing
System), product are available. However, retrieval algorithms are sensitive to many
parameters such as snow grain size distribution and snow liquid water content, which are
very difficult to estimate. Therefore, current satellite-based SWE products still have a
limited accuracy, particularly for deep snow conditions (Takala et al.
2011
). Future sen-
sors, such as the proposed ESA (European Space Agency) Earth Explorer CoReH2O
mission, are designed to accurately retrieve SWE, using dual polarisation measurements at
frequencies optimal to separate grain size and SWE effects on the microwave emission
(Rott et al.
2009
).
While there are still high uncertainties in SWE retrievals from space-borne sensors, it is
possible to estimate the Snow Cover Fraction with a good accuracy from Visible and Near
infrared measurements in cloud-free conditions (Brubaker et al.
2009
). The Moderate
Resolution Imaging Spectroradiometer (MODIS) instruments provide high-resolution
(0.05
°
) daily observations of snow cover. The MODIS snow cover product is used in the
NASA (National Aeronautics and Space Administration)/NOAA Global Land Data
Assimilation System (GLDAS, Rodell and Houser
2004
). The NOAA/NESDIS (National
Environmental Satellite, Data, and Information Service) Interactive Multi-sensor Snow and
Ice Mapping System (IMS) combines ground observations and satellite data from micro-
wave and visible sensors (using geostationary and polar orbiting satellites) to provide snow
