Multi-scale spatial correlation and scaling behavior of surface soil moisture

Abstract

The scaling behavior of surface soil moisture variance is presented across spatial scales of observation from 1 × 1 km2 to 140 × 140 km2. Semi-variograms were used to analyze the spatial correlation in remotely sensed soil moisture data from SGP97. Results indicate that a nested correlation structure exists within regional scale fields. Three representative semi-variograms were chosen to calculate the scaling of soil moisture variance with increasing support scale. While previous research at smaller scales (<1 km2) revealed that soil moisture variance follows a power law decay with increasing support area, we show that in the presence of significant spatial correlation at the larger scales studied here, a power law relationship no longer holds. Rather, the type of correlation function and correlation length play an important role in the relationship between soil moisture variance and support area. We suggest that understanding the spatial correlation pattern of soil moisture across scales is critical for characterizing the scaling behavior of soil moisture variance. Copyright 2006 by the American Geophysical Union.