Location: Hydrology and Remote Sensing LaboratoryTitle: L-band radar experiment and modeling of a corn canopy over a full growing season
|SHARMA, A. - George Washington University|
|LANG, R.H. - George Washington University|
|KURUM, M. - Goddard Space Flight Center|
|O'NEILL, P.E. - Goddard Space Flight Center|
Submitted to: IEEE Transactions on Geoscience and Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/1/2020
Publication Date: 2/15/2020
Citation: Sharma, A., Lang, R., Kurum, M., O'Neill, P., Cosh, M.H. 2020. L-band radar experiment and modeling of a corn canopy over a full growing season. IEEE Transactions on Geoscience and Remote Sensing. 1-15. https://doi.org/10.1109/TGRS.2020.2971539.
Interpretive Summary: Microwave remote sensing of soil moisture for agricultural purposes is confounded by vegetation water content, but this can be overcome with accurate modeling of the microwave signal through the various crop canopies. One of the most challenging agricultural crops in this respect is corn, which has a significant amount of water content. Usually complex modeling of each vegetation component is necessary, but sufficient data were collected during a field experiment in 2012. As a result, a simplified model was validated to a high degree of confidence. This model would be of use to future microwave remote sensing models to simplify calculations and provide efficient and accurate estimation of surface soil moisture through agricultural canopies.
Technical Abstract: Modeling L-Band backscatter from a corn canopy continues to be a challenge due to the complex dynamics in both plant phenology and the underlying soil. An experiment has been conducted to better understand the relationship between L-Band backscatter and canopy parameters such as soil moisture, vegetation water content, dew and periodic rows. The experiment consists of field measurements that take into account plant phenology and are concurrent with L-Band backscatter returns from a corn canopy over a full growing season. The field measurements of the corn plants’ constituents highlight modeling complexities, such as an inhomogeneity in the dielectric constant of the stalk and cobs. A simple method to replace the stalk and cob with a homogeneous dielectric constant is validated. Using the field measurements in a scattering model developed at George Washington University (GW), both coherent and incoherent backscatter is computed. The results show coherent effects contributing to enhanced backscatter by up to 2.7 dB for both HH-pol and VV-pol. The coherent model and the detailed measurements, especially, the dielectric constant of the stalks, resulted in good agreement with the measurements. These measurements have an average root mean square difference (RMSD) with the results from the coherent model of around 1 dB for both HH and VV-pol over the entire growing season. The incoherent mode does not perform as well.