Comparison of Land C Bands Radar Time Series for the Monitoring of Sahelian Area
FRISON, Pierre-Louis¹; MERCIER, Gregoire²; MOUGIN, Eric³; HIERNAUX, Pierre³; LARDEUX, Cedric¹; RUDANT, Jean-Paul^1
1Université Paris-Est, FRANCE; ²Telecom Bretagne, FRANCE; ³Geosciences Environnement Toulouse, FRANCE

The aim of this study is to compare the contribution of L and C band time series for the monitoring of a Sahelian area. This latter belongs to the Gourma region (14.5-17°N, 1-2°W), located in Mali. The Gourma region is located entirely within the Sahel bioclimatic zone and is mainly a pastoral region enclosed by the annual average 500 and 150 mm isohyets. The rain distribution is strictly mono-modal with rainfall starting in June and ending in September with a maximum in August. The rainy season is then followed by a long dry season characterized by the absence of green vegetation apart from some scattered trees and shrubs. Rangeland vegetation is composed of a herbaceous layer and a sparse woody plant population. Herb growth is strongly influenced by the pattern and magnitude of rainfall events and by the soil moisture regime that results from them and from run-off influenced by topography and soil texture. Annual herbs germinate after the first rains, in June or July, and unless the plants wilt before maturity owing to a lack of rainfall, the senescence coincides approximately with the end of the rainy season. Numerous field campaigns have been realised, allowing continuous and repetitive in situ measurements for over twenty years [1]. 23 ALOS PALSAR acquisitions have been made over the Gourma region between January 2007 and April 2009. This sensor operates at L band (lambda =23.6 cm) in different polarimetric and spatial resolution modes. The concerned time series is composed of seven acquisitions in Fine Beam Single (FBS) mode (in HH polarisation), five in Fine Beam Dual (FBD) in HH/HV polarisation, seven in ScanSAR mode (WB1) in HH polarisation, and four in fully polarimetric (PLR) mode. These data are compared to ENVISAT-ASAR data, that have been acquired at C-band (F U¨ =5.7 cm) in Wide Swath mode (HH or VV polarization) and Alternate Polarisation mode (HH/HV). First results show that L band data are closely linked to geophysical features. In particular, the remnant of alluvial system and lacustrine depressions are better discriminated with PALSAR data. When only data acquired during the dry season are analysed, the temporal changes observed over both time series are only encountered over the ponds. The pond detection and their extent estimation is prime interest. First it is the main water resource for livestock and population. It is also a good proxy for hydrological and climate studies [2]. The objective of this study is to make a deeper analysis of temporal changes. When considering the data in a 23-dimensional feature space, those changes may be characterized by a shift which is orthogonal to the main variability of the data. Moreover, it has been found that the saturation channel of the Hue-Saturation-Value (HSV) color transform yields a more robust and accurate change indicator than the second component of the Principal Component Analysis (PCA). Unfortunately, HSV transformation is defined in a 3D feature space only. To cope with this constraint, a Random Projection [3] has been applied iteratively to reduce the initial 23-D data to a 3D space in order to extract the saturation component of the HSV transform. Then, the change measure is defined by the sum of the saturation values. A preprocessing step has been applied in order to reduce dramatically the speckle noise of each image. A local averaging is performed to yield a equivalent number of look greater to 200. Comparison between change detection in L and C bands time series are presented [4]. REFERENCES [1] E. Mougin, P. Hiernaux, L. Kergoat, M. Grippa, P. de Rosnay, F. Timouk, V. le Dantec, V. Demarez, M. Arjounin, F. Lavenu, N. Soumaguel, E. Ceschia, B. Mougenot, F. Baup, F. Frappart, P.-L. Frison, J. Gardelle, C. Gruhier, L. Jarlan, S. mangiarotti, B. Sanou, Y. Tracol, F. Guichard, V. Trichon, L. Diara, A. Soumaré, M. Koité, F. Dambielé, C. Lloyd, N. Hanan, C. Damesin, C. Delon, D. Serça, C. Galy-Lacaux, J. Seghiéri, S. Becerra, H. Dia, F. Gangneron, P. Mazzega, 2009 : The AMMA-CATCH Gourma observatory site in Mali : Relating climatic variations to changes in vegetation, surface hydrology, fluxes and natural resources. Journal of Hydrology, vol. 375, nc X1-2, pp14-33. [2] Gardelle, J., Hiernaux, P., Kergoat, L., and Grippa, M.: Less rain, more water in ponds: a remote sensing study of the dynamics of surface waters from 1950 to present in pastoral Sahel (Gourma region, Mali), Hydrol. Earth Syst. Sci., 14, 309-324, doi:10.5194/hess-14-309-2010, 2010. [3] S. Santosh Vempala, "The Random Projection" American Mathematical Society, DIMACS series in discrete mathematics and theoretical computer science, vol 65, 2004. http://books.google.fr/books?id=L5L2J0UEY4QC&printsec=frontcover&dq=random+projection&source=bl&ots=mBtbS4ROWF&sig=Xev8WALb19xd2YbzhlZMEnqMCrg&hl=fr&ei=HAAmTbfNHcvZ4gadouWdCg&sa=X&oi=book_result&ct=result&resnum=9&ved=0CH4Q6AEwCA#v=onepage&q&f=false [4] Frison P.-L., Mercier G., Faye G., Mougin E., Hiernaux P., Lardeux C., Rudant J.-P., 2013: Analysis of L and C-bands SAR images time series over a Sahelian area. Accepté dans IEEE Geosc. Remote Sensing Letters, doi 10.1109/LGRS.2012.2227931.