Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #326187
Title: Error sources in passive and active microwave satellite soil moisture over Australia

Author
item SU, CHUN-HSU - University Of Melbourne
item ZHANG, JING - University Of Melbourne
item GRUBER, ALEX - Vienna University Of Technology
item PARINUSSA, R.M. - University Of New South Wales
item RYU, D. - University Of Melbourne
item Crow, Wade
item WAGNER, W. - Vienna University Of Technology

Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/1/2016
Publication Date: 9/1/2016
Citation: Su, C., Zhang, J., Gruber, A., Parinussa, R., Ryu, D., Crow, W.T., Wagner, W. 2016. Error sources in passive and active microwave satellite soil moisture over Australia. Remote Sensing of Environment. 182:128-140. doi:10.1016/j.rse.2016.05.008.

Interpretive Summary: Remotely-sensed soil moisture products have the potential to aid a variety of agricultural and climate applications. However, before this potential can be fully met, we need to better understand the sources and magnitudes of error (both systematic and random in nature) within existing soil moisture products. This is critical for attempts to merge soil moisture products derived from multiple satellite platforms into a single, unified product. With his goal in mind, this paper applies a novel mathematical technique to better understand 1) the relationship between soil moisture products derived from different sensors and 2) the errors existing within individual products. With the understanding gained from this analysis, we will be able to more efficiently use existing observations for critical agricultural applications (e.g., agricultural drought monitoring and climate trend analysis).

Technical Abstract: Development of a long-term climate record of soil moisture (SM) involves combining historic and present satellite-retrieved SM data sets. This in turn requires a consistent characterization and deep understanding of the systematic differences and errors in the individual data sets, which vary due to changes in instrument configuration, backscatter/brightness temperature calibration, and retrieval algorithm design and calibration. This work presents a comprehensive assessment of the systematic and random errors in nine passive and active satellite SM products over Australia (time coverage: 1978–present), which are estimated in a consistent manner using lagged instrument variable (LV) analysis. The additive and multiplicative biases (relative to modelled SM) and random errors in satellite SM, which are constituents of the bulk statistics such as mean squared error, are separately determined for SM components at sub-seasonal and seasonal-to-interannual timescales, revealing non-trivial error structures in individual satellite SM products. Further, the spatial variability in these errors can be modelled to some degrees using multiple linear regression models with up to thirteen explanatory variables that correspond to various land-surface and climatic characteristics. The results give cause for conducting satellite SM merging and data assimilation in a multi-scale fashion, explain possible sources for the observed errors, and allow predictions of individual error components and screening for low signal-to-noise ratio (SNR) based on a subset of the explanatory variables.