Examining Soil Moisture Variability and Field Mean Estimation Methods using Nested Observations

Thursday, 18 December 2014
Amber Peterson, Warren Helgason and Andrew M Ireson, University of Saskatchewan, Saskatoon, SK, Canada
Information about soil moisture is typically required at the field scale. Direct measurements of soil moisture at this scale are not possible, though there are a number of promising indirect methods (e.g. remote sensing methods and cosmic-ray neutrons). Methods for obtaining point scale measurements of soil moisture are well established. However, variability of soil moisture, in both space and time, makes accurately determining field scale soil moisture from point measurements difficult. Understanding sub-field scale variability is a key step in determining how to upscale point measurements, and in particular to identify the minimum number of point measurements necessary to represent field scale mean soil moisture. Objectives of this study are to: (1) examine the spatial variability of soil moisture with time, and (2) compare field scale soil moisture estimation methods. Nested soil moisture measurements provided observations covering a 5002m2 area within a semi-arid prairie pasture site in southern Saskatchewan, Canada. Complementary measurements of the water balance were measured using meteorological and flux instrumentation. Spatial variability of surface and root zone soil moisture were examined using data from gridded dielectric water content probe surveys and a neutron probe array. Field scale surface soil moisture was measured at the site using a cosmic-ray neutron probe. The field scale estimation methods compared are: (1) water balance, (2) upscaling by averaging point scale measurements, (3) upscaling by identification of average representative time stable sites, and (4) extrapolation of shallow soil moisture measured by cosmic-ray neutron probe. Variability of surface soil moisture was found to be smallest under extreme dry and wet conditions, and largest during intermediate moisture conditions. Large spatial variability was found in the root zone, with soil moisture being most temporally variable closer to the surface.