Home > Research > Publications & Outputs > Explaining temporal and spatial variation in so...


Text available via DOI:

View graph of relations

Explaining temporal and spatial variation in soil moisture in a bare field using SAR imagery

Research output: Contribution to journalJournal articlepeer-review

<mark>Journal publication date</mark>2003
<mark>Journal</mark>International Journal of Remote Sensing
Issue number15
Number of pages16
Pages (from-to)3059-3074
Publication StatusPublished
<mark>Original language</mark>English


Previous studies to determine the utility of satellite remote sensing for soil moisture monitoring have concentrated on field scale assessment. The European Remote Sensing satellites (ERS-1 and ERS-2) Synthetic Aperture Radar (SAR) instruments have a spatial resolution of less than 30 m in azimuth and 26 m in range, and potentially offer a considerably finer scale of soil moisture assessment. A series of soil moisture measurements were taken within a 15 ha bare field on the Essex (UK) coast at times to coincide with ERS-1 overpasses in the autumn and winter of 1995-96. This dataset has been used to investigate the ability of ERS-1 SAR backscatter response to detect spatial variation in soil moisture within the field. Analysis of variograms of SAR backscatter show that spatial variations in the ERS SAR data can be detected within the field and that spatial dependency increases when the soil is wet. Field measurements of soil moisture do not reflect these findings in quite the same way; during periods of low soil moisture content, the spatial variation of soil moisture cannot be easily characterized by the variograms. When the field is wet, spatial autocorrelation is more evident in the variograms. Regression analysis of paired observations of measured soil moisture and ERS SAR backscatter demonstrated that no significant relationship could be obtained for each day. However, when paired data for different dates were combined, a significant relationship between SAR backscatter and soil moisture could be obtained. The results suggest, therefore, that for this case study, soil moisture can be predicted at the field scale but not at the within-field/pixel scale.

Bibliographic note

M1 - 15