By Topic

Validation of SMOS L1C and L2 Products and Important Parameters of the Retrieval Algorithm in the Skjern River Catchment, Western Denmark

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

3 Author(s)
Simone Bircher ; Department of Geography and Geology, University of Copenhagen, Copenhagen K, Denmark ; Niels Skou ; Yann H. Kerr

The Soil Moisture and Ocean Salinity (SMOS) satellite with a passive L-band radiometer monitors surface soil moisture. In addition to soil moisture, vegetation optical thickness \tau _{\rm NAD} is retrieved (L2 product) from brightness temperatures ( T_{B} , L1C product) using an algorithm based on the L-band Microwave Emission of the Biosphere (L-MEB) model with initial guesses on the two parameters (derived from ECMWF products and ECOCLIMAP Leaf Area Index, respectively) and other auxiliary input. This paper presents the validation work carried out in the Skjern River Catchment, Denmark. L1C/L2 data and the most sensitive algorithm parameters were analyzed by network and airborne campaign data collected within one SMOS pixel (44 km diameter). The SMOS retrieval is based on the prevailing low vegetation class. For the L1C comparison, T_{B} 's were calculated from in situ soil moisture using L-MEB. Consistent with worldwide findings, the initial/retrieved SMOS soil moisture captures the in situ dynamics well but with significant wet/dry biases and too large amplitudes in case of the latter. While the initial \tau _{\rm NAD} is in range with an in situ estimate for low agricultural vegetation, the retrieved \tau _{\rm NAD} is too high with too pronounced temporal variability. A filter based on L2 criteria removed radio frequency interference (RFI) and improved the R^{2} between retrieved and network soil moisture from 0.49 to 0.61, while the bias remained (-0.092/-!0.087 \hbox {m}^{3}/\hbox {m}^{3}) . L- kely error sources include the following: 1) still present RFI; 2) potential link between high retrieved \tau _{\rm NAD} and other L-MEB parameters, e.g., low roughness parameter (H_{R}) ; 3) \sim 18% lower sand and \sim 8% higher clay fractions while \sim!!0.35 \hbox {g/cm}^{3} lower bulk density in SMOS algorithm than in situ; and 4) caveats in the Dobson dielectric mixing model implemented in the L-MEB model. A previous study at the Danish validation site had revealed superior performance of the Mironov dielectric mixing model at the 2 \times 2 km scale. Studies are ongoing to address the aforementioned issues, and the role of organic surface layers will be investigated.

Published in:

IEEE Transactions on Geoscience and Remote Sensing  (Volume:51 ,  Issue: 5 )