Author
![]() |
Pachepsky, Yakov |
![]() |
SHEN, M - UNIVERSITY OF MARYLAND |
![]() |
RAWLS, WALTER |
Submitted to: Sring Meeting American Geographical Union Washington DC
Publication Type: Abstract Only Publication Acceptance Date: 5/1/2000 Publication Date: 5/22/2000 Citation: N/A Interpretive Summary: None. Technical Abstract: Many potential user groups have a need in estimating soil hydraulic properties from readily available soil survey data. Empirical regression equations built for this purpose are called pedotransfer functions (PTF). PTFs were developed from various soil databases. The NRCS MUUF pedotransfer model has important advantages, such as geo-referencing, the same set of the soil analysis methods for all data, reflecting effects of clay mineralogy on the hydraulic properties, and the coverage of the whole United States. MUUF uses soil name and textural class or name or soil association identification to give estimates of soil hydraulic properties. Although the potential of the MUUF program is obviously large, its validation has been limited. The objectives of this work were (a) to evaluate MUUF using a representative independent database, (b) to see whether MUUF estimates can be improved if the measured texture will be used dinstead of the textural class. We used the Southern Plains soil hydraulic database that contained 500 records. Water content (vol. percent) at 0.33 bar and at 15 bar capillary pressures was estimated. Using measuring texture improved the MUUF estimates. The 0.33 bar RMSE was circa 4 percent with and without using measured texture in topsoil, and was 4 percent and 7 percent, respectively, in deeper horizons. The 15 bar RMSE was circa 2 percent (3 percent) and 3 percent (7 percent) in topsoil (subsoil) with and without measured texture. With this accuracy, regional applications may use MUUF as a reasonable source of soil hydraulic properties. |