IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, VOL. 8, NO. 8, AUGUST 2015 3833
Snow Density and Ground Permittivity Retrieved from L-Band Radiometry: A Synthetic Analysis
Mike Schwank, Christian Mätzler, Andreas Wiesmann, Senior Member, IEEE, Urs Wegmüller, Jouni Pulliainen,
Juha Lemmetyinen, Kimmo Rautiainen, Chris Derksen, Peter Toose, and Matthias Drusch
Abstract—A synthetic study was performed to determine the potential to retrieve dry-snow density and ground permittivity from multiangular L-band brightness temperatures. The thereto employed emission model was developed from parts of the “microwave emission model of layered snowpacks” (MEMLS) coupled with components adopted from the “L-band microwave emission of the biosphere” (L-MEB) model. The restriction to Lband made it possible to avoid scattering and absorption in the snow volume, leading to a rather simple formulation of our emission model. Parametric model studies revealed L-band signatures related to the mass density of the bottom layer of a dry snowpack.
This gave rise to the presented analysis of corresponding retrieval performances based on measurements synthesized with the developed emission model. The question regarding the extent to which random noise translates into retrieval uncertainties was investigated. It was found that several classes of snow densities could be distinguished by retrievals based on L-band brightness temperatures with soil moisture and ocean salinity (SMOS)-typical data quality. Further synthetic retrievals demonstrated that propagation effects must be taken into account in dry snow even at L-band when retrieving permittivity of the underlying ground surface.
Accordingly, current SMOS-based retrievals seam to underestimate actual ground permittivity by typically 30% as dry snow is wrongly considered as “invisible.” Although experimental validation has not yet been performed, the proposed retrieval approach is seen as a promising step toward the full exploitation of L-band brightness temperatures available from current and future satellite Earth observation missions, especially over the cold regions of the Northern Hemisphere.
Index Terms—Ground permittivity, microwave radiometry, retrieval, soil moisture and ocean salinity (SMOS), snow density.
R ADIATION, heat, and mass fluxes through the surfaceof terrestrial land areas are major drivers of climate.
The quantities involved in the associated exchange rates are
Manuscript received November 28, 2014; revised February 16, 2015; accepted April 13, 2015. Date of publication May 12, 2015; date of current version September 12, 2015.
M. Schwank is with the Swiss Federal Research Institute WSL, CH-8903
Birmensdorf, Switzerland (e-mail: email@example.com).
C. Mätzler, A. Wiesmann, and U. Wegmüller are with the Gamma Remote
Sensing AG, CH-3073 Gümligen, Switzerland.
J. Pulliainen, J. Lemmetyinen, and K. Rautiainen are with the Finnish
Meteorological Institute, FI-00101 Helsinki, Finland.
C. Derksen and P. Toose are with the Environment Canada, North York, ON
M3H 5T4, Canada.
M. Drusch is with the European Space Agency, ESTEC, 2200 AG
Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/JSTARS.2015.2422998 fundamentally linked to soil moisture, the freeze/thaw state of the ground, as well as the states of vegetation and snow cover.
Accordingly, the provision of appropriate data products at the global scale has been defined as the objective of the second Earth Explorer Opportunity mission “soil moisture and ocean salinity” (SMOS)  launched on November 2, 2009.
Since May 2010, the L-band radiometer MIRAS  on board the SMOS satellite measures (m) multiangular brightness temperature T pB,m(θk) at horizontal (p = H) and vertical (p = V) polarization at observation angles θk. With spatial resolution of ≈(40× 40) km2 and with a repeat-time of approximately 3 days at the equator, SMOS T pB,m(θk) are measured globally within the protected 1.400–1.427 GHz range of the microwave
L-band (1–2 GHz) . At the corresponding sensitive SMOS wavelength λ ≈ 21 cm, thermal emission T pB,m(θk) is proven to be highly sensitive to changes in liquid water in the soil, even in the presence of moderate vegetation or dry snow-covers. This is because L-band measurements are less susceptible to attenuation and volume scattering than corresponding observations at higher frequencies . For the same reason, L-band emission is almost entirely unaffected by atmospheric conditions, allowing for all-weather measurements T pB,m(θk) of the Earth’s surface.
Furthermore, microwave emissions T pB,m(θk) of the land surface are almost entirely independent of solar radiation, which is particularly important for corresponding satellite observations of high latitude regions.
Beyond the global provision of SMOS data, defined as the SMOS main objectives, a generally important aspect for
SMOS is the extended and advanced exploitation of T pB,m(θk) in compliance with the challenges identified in ESA’s scientific strategy for the Living Planet Programme, “The Changing
Earth.” Among other areas, this includes the characterization of terrestrial snow cover and freeze/thaw cycles across midto high latitudes. In this context, a number of research activities have been started in support of using passive L-band observations to derive information on, e.g., 1) drought and flooding; 2) surface roughness ; 3) sea ice thickness ; 4) vegetation optical depth ; and 5) soil freeze/thaw states . Likewise, the retrieval of global information on groundsurface states is a key component of the National Aeronautics and Space Administration (NASA) Soil Moisture Active and
Passive (SMAP) mission .
In the course of recent model-oriented research on the Lband emission of snow-covered ground, it has been found that mass density ρS of dry snow just above the ground surface affects L-band brightness temperatures significantly . Since 1939-1404 © 2015 EU 3834 IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, VOL. 8, NO. 8, AUGUST 2015 penetration depth in dry snow is >100m , , the identified sensitivity of L-band T pB,m(θk) with respect to ρS was not expected a priori. The analysis of the relevant propagation effects revealed that refraction and impedance matching caused by dry snow in contact with the ground surface explains this sensitivity of T pB,m(θk). In spite of this suggested sensitivity, no previous effort has been made to retrieve dry-snow massdensity ρS from L-band T pB,m(θk). Instead, current satellite derived snow mass datasets utilize shorter wavelength passive microwave measurements , while proposed missions have focused on active microwave sensors at similar wavelengths.