Skip to main content
ARS Home » Research » Publications at this Location » Publication #117018
Title: STATISTICAL PROPERTIES OF SOIL MOISTURE IMAGES REVISITED

Author
item OLDAK, ANNA - USDA-ARS
item Pachepsky, Yakov
item Jackson, Thomas
item Rawls, Walter

Submitted to: Journal of Hydrology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/1/2001
Publication Date: 1/10/2002
Citation: N/A

Interpretive Summary: Microwave remote sensing offers the unique opportunity to measure surface soil moisture over large areas. Different instruments have different resolutions, and it was shown that the resolution affects the spatial variability of soil moisture values inferred from the measurements. This conclusion was obtained from observations of a single drydown period over the Little Washita watershed in Oklahoma. New data provided observations o several drydowns for the same area. Our goal was to see whether the earlier discovered effect of resolution on spatial variability holds. We found that the dependencies of variability on resolution are different for different drydown periods. However, once a dependence on resolution is established at the beginning of a drying period, its shape is maintained during the drydown. The further changes in the dependence are governed mostly by the average soil moisture over the area. Having constant shape is the first approximation to predict the resolution dependencies of soil moisture variability for the days without observations during drydowns. This empirical evidence warrants future studies to understand reasons of the persistence in shapes of scaling dependencies and factors that form this shape prior to drydown.

Technical Abstract: Data from passive microwave remote sensing were used to produce soil moisture images over the area of the Little Washita watershed in Oklahoma. The area of the experiment conducted in 1992 was revisited during the Southern Great Plains 1997 Hydrology Experiment and offered an opportunity to re-evaluate the statistical structure of soil moisture images under different scales of observation. Three types of scaling laws were tested i this work: the power law scaling of the variance of moisture content; multiscaling that relates the moment of soil moisture at a base scale to the comparable moments at a coarser scale, and within-pixel variance. Data on scaling in 1992 and 1997 have one important common feature, once a dependence on resolution is established at the beginning of a drying period its shape is maintained during the drydown. Slopes of the dependencies change only slightly, whereas, the intercepts decrease as the drying progresses. This implies that if the slope is found from an image of the first day of a drydown, it can be used with reasonable accuracy on the following days of the drying period. Parameters of the scaling equations change during drying periods, but having constant slopes and intercepts dependent on average area water contents gives a first approximation to predict scaling laws of soil moisture for the days without observations during drydowns. This empirical evidence warrants future studies to understand reasons of the persistence in shapes of scaling dependencies and factors that form this shape prior to drydown.