ABSTRACT
Soil moisture observations from satellite sensors are becoming widely available at the global scale with a good accuracy and a spatial-temporal resolution suitable for many applications (Wagner et al. 2007; Seneviratne et al. 2010). However, the climatology of satellite-derived and modeled/in situ soil moisture observations can be very different because of uncertainties affecting both data sources (Koster et al. 2009; Entekhabi et al. 2010). Moreover, the difference between the spatial extent of in situ (~0.1 m²) and coarseresolution satellite (~600 km²) observations is very large. Because of the high soil moisture spatial variability (Famiglietti et al. 2008; Brocca et al. 2010b), point measurements are expected to be not representative of the actual absolute value (i.e., in m³ m-³) of average soil moisture at the satellite pixel scale. Anyhow, the well-known scaling properties of soil moisture (Vachaud et al. 1985; Brocca et al. 2012b) have showed that in situ measurements
CONTENTS
17.1 Introduction ........................................................................................................................ 411 17.2 Scaling and Filtering Techniques Methods: Overview ................................................ 412
17.2.1 Scaling Techniques ................................................................................................ 413 17.2.1.1 Linear Regression .................................................................................... 413 17.2.1.2 CDF Matching ......................................................................................... 413 17.2.1.3 Linear Rescaling ...................................................................................... 413 17.2.1.4 Min/Max Correction .............................................................................. 414
17.2.2 Filtering Techniques .............................................................................................. 414 17.2.2.1 Exponential Filter .................................................................................... 414 17.2.2.2 Nonlinear Exponential Filter ................................................................. 415 17.2.2.3 Moving Average ...................................................................................... 415
17.3 Case Studies Datasets ........................................................................................................ 415 17.3.1 In situ Soil Moisture Observations ...................................................................... 415 17.3.2 Satellite Soil Moisture Observations ................................................................... 416
17.4 Results and Discussions ................................................................................................... 417 17.4.1 Scaling Techniques ................................................................................................ 417 17.4.2 Filtering Techniques .............................................................................................. 421
17.5 Conclusions .........................................................................................................................423 References .....................................................................................................................................423
can capture the large-scale temporal dynamics. Another matter is that satellite sensors are able to monitor only a very thin soil layer (less than 5 cm), providing soil moisture information that is difficult to be assimilated into hydrological and meteorological models (Brocca et al. 2010a, 2012a; Dharssi et al. 2011).
