Estimating surface soil moisture and soil roughness over semiarid areas from the use of the copolarization ratio R.D. Magagi a, *, Y.H. Kerr b a INRS-Eau, 2800 rue Einstein, Case Postale 7500, Sainte-Foy, Quebec, Canada G1V 4C7 b CESBIO, 18 avenue E. Belin, BPi. 2801, 31401 Toulouse cedex 4, France Received 28 February 2000; accepted 21 August 2000 Abstract This paper presents a new method to retrieve soil moisture and roughness from ERS-1. Wind scatterometer (WSC) data measured over the HAPEX-Sahel area (semiarid environment). The retrieval algorithm makes full use of the multiangular acquisitions and the high temporal repetition of the measured backscattering coefficients. The vegetation contribution to the signal is taken into account through a first-order radiative transfer model. The soil moisture and roughness are subsequently retrieved, throughout the rainy season, using the copolarization ratio as expressed by Oh et al. [IEEE Transactions on Geoscience and Remote Sensing GE-30 (1992) 370±381.]. The paper describes the data and the approach used, together with the results gained. A good sensitivity of the backscattering coefficient to soil moisture is obtained. The results are compared with data collected during the HAPEX-Sahel campaign. D 2001 Elsevier Science Inc. All rights reserved. 1. Introduction Soil moisture measurements are a prerequisite for many research and application topics. In this paper, the term soil moisture (obtained from microwave data) will refer to the amount of water contained in the very first centimeters of the soil. Actually, soil moisture is of crucial importance for characterizing the water exchanges between the surface and the atmosphere (Chanzy & Kustas, 1995), for the derivation of fluxes at the surface atmosphere interface, and as input (Pielke, Dalu, Snook, Lee, & Kittel, 1991) in global circulation models (GCMs), to name but a few. Beljaars, Viterbo, Miller, and Betts (1996) showed that the knowl- edge of soil moisture significantly improves the accuracy in rainfall estimates. Surface soil moisture is also of prime importance in surface hydrology since, coupled with the roughness, it sets the conditions for runoff/infiltration as a function of rain rates. Similarly, in vegetation growth models, it is necessary to have an estimate of all the growth limiting factors, i.e., water, temperature, insolation, and nutrients. In arid and semiarid areas the most limiting factor is water availability. Even though soil moisture is not a solution for water availability, methods have been developed for the derivation of root zone water content through the knowledge of surface soil moisture, rainfall and soil characteristics (Bernard, Vauclin, & Vidal-Madjar, 1981; Jackson, 1986). However, in spite of the high demand, there are no means for the time being to gather information on soil moisture on a global scale. The previous nonimaging spaceborne scatte- rometers were in orbit during a short period of time (summer 1973 for Skylab and summer 1978 for SEASAT) and operated in K-band (13.8 GHz). Therefore, they were not suitable to allow useful temporal monitoring of soil moist- ure. It seemed useful to test the potential of the low- frequency active microwave satellite system, the ERS-1 wind scatterometer (WSC), to derive surface soil moisture. Due to the relationships between the backscattering coeffi- cient (s 0 ) and surface characteristics (soil roughness, soil moisture, and vegetation), many studies have shown the possibility of inferring land surface parameters from active microwave data (Alphonse, 1988; Benallegue, Taconet, Vidal-Madjar, & Normand, 1995; Bouman, 1991; Dobson & Ulaby, 1986a, 1986b; Engman 1991; Mo, Wang, & Schmugge, 1988; Prevot, Champion, & Guyot, 1993; Pre- vot, Dechambre, et al., 1993). However, these studies mainly showed that radar signal could be related to soil * Corresponding author. Tel.: +1-418-654-3784; fax: +1-418-654- 2600. E-mail address: magagira@inrs-eau.uquebec.ca (R.D. Magagi). www.elsevier.com/locate/rse Remote Sensing of Environment 75 (2001) 432 ± 445 0034-4257/00/$ ± see front matter D 2001 Elsevier Science Inc. All rights reserved. PII:S0034-4257(00)00185-1