Location: Hydrology and Remote Sensing LaboratoryTitle: Investigating the SMOS dry bias in the corn belt of the United States
|WALKER, V.A. - Iowa State University|
|HORNBUCKLE, B. - Iowa State University|
Submitted to: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/1/2018
Publication Date: 9/5/2018
Citation: Walker, V., Hornbuckle, B., Cosh, M.H. 2018. Investigating the SMOS dry bias in the corn belt of the United States. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. 11(12):4664-4675. https://doi.org/10.1109/JSTARS.2018.2864897.
Interpretive Summary: Soil moisture remote sensing is growing as a resource for agriculture and water planning with the recent launch of dedicated L-band satellites. The data produced from these satellites has been shown to be accurate for the majority of land covers around the world, however, there is still a larger than expected error over agricultural regions such as Iowa. This error is related to a bias from the poor parameterization of the soil roughness. Soil roughness is often considered constant in satellite remote sensing, however, the tillage practices in agriculture have introduced a dynamic aspect to the surface that is not fully understood in satellite remote sensing. This study is of importance to remote sensing scientists who must address the land surface characterization in new algorithms that are under development.
Technical Abstract: The Soil Moisture Ocean Salinity (SMOS) satellite mission is currently making global observations of soil moisture: a key component of agriculture that both provides water to growing crops and influences precipitation. When compared to the South Fork Iowa River (SFIR) in situ network, located in the heart of the Corn Belt, SMOS soil moisture is too dry and too noisy. The dry bias could be caused by radio frequency interference (RFI) or invalid soil moisture retrievals. Removing these retrievals did not improve the bias. A cold bias in auxiliary surface temperature would create a dry bias, but the temperature used by SMOS does not have a strong enough bias to impact soil moisture retrievals. While errors in assumed clay content could result in a soil moisture bias, the soil texture maps used by SMOS are similar to those compiled from United States Department of Agriculture soil surveys. SMOS currently does not account for scattering that occurs in a corn/soybean canopy such as the SFIR. Adding this component into soil moisture retrievals during a high vegetation test case worsens the dry bias but improves the noisiness and the correlation coefficient. While SMOS currently parameterizes the soil as moderately smooth, increasing roughness improves the bias for both low and high vegetation test cases at the cost of decreasing sensitivity to actual soil moisture. We hypothesize that the SMOS dry bias can be improved by the use of a dynamic soil roughness that varies when vegetation is limited.