The effect of rock cover fraction on the retrieval of surface soil moisture at L-band

Nan Ye, Jeffrey Walker, Rocco Panciera, Dongryeol Ryu, Christoph Rudiger, Robert Gurney

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review


The Soil Moisture and Ocean Salinity (SMOS) mission, developed by the European Space Agency (ESA), will be launched in the second half of 2009. It will be the first L-band (~1.4 GHz) passive microwave satellite specifically designed for global soil moisture observations, with an expected accuracy for the retrieved soil moisture of ~0.04 m3/m3. While passive microwave observations have been widely acknowledged to give the most accurate information on soil moisture, these sensors are characterized by low spatial resolution footprints, being on the order of 50km. One of the key difficulties with observations at this scale is the heterogeneity that exists in land surface features. However, past and current soil moisture retrieval algorithms have typically assumed a homogeneous pixel approach. Thus, in order to maximize the soil moisture retrieval accuracy, the various land surface features that exist in a satellite footprint should be taken into account. While the SMOS retrieval algorithm distinguishes between three different surface types (bare soil, herbaceous and woody vegetation), at this stage it does not take into consideration other sub-pixel effects such as the brightness temperature contribution from open water bodies, rock cover or urban areas, which are expected to affect the overall soil moisture retrieval accuracy for many parts of the world. This study explores the impact of surface rock on the retrieval of surface soil moisture for short vegetation covered fields by comparing retrieved soil moisture estimates with and without accounting for the presence of rock in a synthetic framework. First, the microwave “observation” used to retrieve soil moisture is simulated by accounting for the contribution of rocks to the overall emission, assuming that rock behaves like very dry soil with a fixed dielectric constant and a smooth surface. The soil moisture is then retrieved using the homogeneous pixel approach. The simulation of microwave emission is based on a representative short grass field with various surface rock cover fractions, soil moisture contents and vegetation conditions. The results illustrate that rock induced soil moisture retrieval error is dependent on soil moisture and vegetation water content since the brightness temperature difference between soil and rock impacts the soil moisture retrieval error. The omission of rock cover from the retrieval algorithm leads to an overestimation of the bulk soil moisture content for low soil moisture conditions and an underestimation for high soil moisture conditions. Taking 30% rock cover fraction as an example, the maximum error in the bulk soil moisture estimation is as much as 0.04 m3/m3 in bare soil and up to 0.10 m3/m3 in wet soil covered by short grass vegetation. It should be noted however that these results may be highly dependent on two key assumptions of this paper; i) that rock can be modelled as a smooth surface, and ii) that there is no vegetation cover over the rock.

Original languageEnglish
Title of host publicationThe 18th World IMACS Congress and MODSIM09 International Congress on Modelling and Simulation
EditorsR S Anderssen, R D Braddock, L T H Newham
Place of PublicationNedlands WA
PublisherUWA Publishing
Pages3761 - 3767
Number of pages7
ISBN (Electronic)9780975840078
ISBN (Print)9780975840078
Publication statusPublished - 2009
Externally publishedYes
EventInternational Congress on Modelling and Simulation 2009: Interfacing Modelling and Simulation with Mathematical and Computational Sciences - Cairns Australia, Cairns, Australia
Duration: 13 Jul 200917 Jul 2009
Conference number: 18th

Publication series

Name18th World IMACS Congress and MODSIM 2009 - International Congress on Modelling and Simulation: Interfacing Modelling and Simulation with Mathematical and Computational Sciences, Proceedings


ConferenceInternational Congress on Modelling and Simulation 2009
Abbreviated titleMODSIM09


  • Passive microwave
  • Remote sensing
  • Rock fraction
  • Soil moisture

Cite this