Skip to main content
ARS Home » Research » Publications at this Location » Publication #70361
Title: ESTIMATING THE SOIL WATER RETENTION CURVE WITH SOIL BULK DENSITY AND -33 K PA VALUE.

Author
item Williams, Robert
item Ahuja, Lajpat
item Timlin, Dennis

Submitted to: ARS Workshop on Real World Infiltration
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/26/1996
Publication Date: N/A
Citation: N/A

Interpretive Summary: Computer models of the complex soil-water process have a demonstrated potential for improving management. However these models require knowledge of the basic soil hydraulic properties which are time consuming and tedious to measure. Here we summarize our recent work using similar-media scaling and the one-parameter model to determine soil water content (or the soil water retention curve) using soil properties which are easily measured or that can be readily obtained from the literature. These simple techniques can also be used to obtained values required for estimating infiltration parameters, as well as to determine the spatial and temporal variation of soil hydraulic properties. Both the similar-media scaling and the one- parameter model are simple to use and can be easily incorporated into the various soil-water erosion models.

Technical Abstract: Two methods to estimate the soil water retention curve are summarized: similar-media scaling and a one-parameter model. Both require the saturated soil water content, or its estimate from the soil bulk density, plus one measured point on the curve (such as the water content at a matric potential of -33 kPa). Similar-media scaling also requires knowledge of one complete curve to serve as a reference. The one-parameter model is based on the log-log form of the soil water retention curve below the air- entry value of the soil (psi), and requires a generalized slope-intercept relationship. Although the methods estimate the soil water retention curve fairly well, visual comparison of the calculated vs measured soil water content showed less scatter in the relationship to the 1:1 line, concomitant with smaller calculated error terms, for the one-parameter model.