| Abstract: |
In the past the effect of soil roughness was often considered secondary within the determination of soil moisture from remote sensing instruments. But several studies showed, that an accurate determination of soil roughness leads to an improved estimation of soil moisture [1].
Soil roughness characterizes the geometric properties of the soil surface. Two standard parameters (used in remote sensing) to describe the surface roughness are the standard deviation of the surface height variation s with its corresponding autocorrelation function, and the surface correlation length l [2]. Both parameters (s, l) affect the emissivity measured by radiometers as well as the backscattering observed by radars [1], and an increase of soil roughness produces an increase in both, emissivity and backscattering [3].
In this study, we developed a novel approach to retrieve surface roughness parameters by combining polarimetric radar and radiometer microwave signatures through a linear functional relationship based on active-passive signal covariation. The algorithm is physics-based, containing a forward model and a retrieval algorithm [4], [5].
To test the roughness retrieval algorithm, we used active/passive microwave time series data (06/04/2012 to 06/29/2012) measured with the ComRAD truck-based SMAP simulator at L-band. The acquisitions were carried out under a 40° incidence angle during the APEX12 field experiment (2012) over two bare soil fields at the USDA test site in Beltsville, MA, USA [6]. In situ measurements of soil roughness (s, l) were conducted for each field, but only once during the entire campaign period.
The results of the roughness estimation and the comparison with corresponding field measurements on ground show that s and l can be estimated simultaneously when using the physics-based retrieval approach. After preliminary validation, the RMS-Error of s between the microwave-based and field-measured surface roughness parameters is 0.274 cm and 0.365 cm, respectively. The RMS-Error of l amounts to 1.658 cm and 1.550 cm for both fields. A complete validation of all results is still ongoing and will be presented at the conference. However it reveals already that the influence of the autocorrelation function, needed within the retrieval, is distinct. Moreover, the temporal variation of the roughness estimates over days, where environmental influences, like rain and wind, were little, is another observation that will be analyzed.
[1] Saatchi, S.S., Njoku, E.G., Wegmueller, U. (1994). ”Synergism of active and passive microwave data for estimating bare soil surface moisture.” In: Passive Microwave Remote Sensing of Land-Atmosphere Interactions (Edts. Choudhury, Kerr, Njoku & Pampaloni), 205-224.
[2] Ulaby, F.T., Long, D.G., Blackwell, W., Elachi, Ch., Fung, A., Ruf, C., Sarabandi, K., Zebker, H., van Zyl, J.J. (2014). Microwave radar and radiometric remote sensing. Ann Arbor, USA, The University of Michigan Press.
[3] Guerriero, L., Ferrazzoli, P., Vittucci, C., Rahmoune, R., Aurizzi, M., & Mattioni, A. (2016). L-Band Passive and Active Signatures of Vegetated Soil: Simulations With a Unified Model. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 9 (6), 2520-2531.
[4] Jagdhuber, T., Konings, A.G., McColl, K.A., Alemohammad, S.H., Das, N.N., Montzka, C., Link, M., Akbar, R., Entekhabi, D. (2017): Physics-Based Modeling of Active and Passive Microwave Covariations Over Vegetated Surfaces. IEEE Transactions on Geoscience and Remote Sensing, in review.
[5] Jagdhuber, T., Das, N.N., Entekhabi, D., Baur, M., Link, M., Piles, M., Akbar, R., Konings, A.G., McColl, K.A., Alemohammad, S.H., Montzka, C., Kunstmann, H. (2016): A data-driven and physics-based single-pass retrieval of active-passive microwave covariation and vegetation parameters for the SMAP mission. AGU Fall Meeting, San Francisco, USA.
[6] O'Neill, P., Joseph, A., Srivastava, P., Cosh, M., Lang, R. (2014). Seasonal parameterizations of the tau-omega model using the ComRAD ground-based SMAP simulator. International Geoscience and Remote Sensing Symposium (IGARSS). 2423-2426. |
