Skip to main content
ARS Home » Research » Publications at this Location » Publication #68198

Title: APPLICATIONS OF GIS AND LOGISTIC REGRESSION IN EVALUATING SALINITY DEVELOPMENT IN IRRIGATED LANDS

Author
item WANG, H - GRESHAM, OR - PRIVATE CO.
item Corwin, Dennis
item LUND, L - UCR, RIVERSIDE, CA
item Rhoades, James
item VAUGHAN, PETER - UCR, RIVERSIDE, CA
item CONE, D - BROADVIEW WATER DIST.

Submitted to: Journal of Environmental Quality
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/10/1995
Publication Date: N/A
Citation: N/A

Interpretive Summary: Soil salinization constitutes a major cause for reduced crop yield and degraded groundwater quality in irrigated lands of arid and semi-arid regions. The objectives of this study were to identify factors which strongly influence salinity development in the rootzone of soil in the Broadview Water District (~6000 acre) of the central valley of California and to develop a statistical model known as a logistic regression model based on these salinity development factors. The logistic regression model was incorporated into geographic information system (GIS) technology to create a practical, management-oriented model that visually displayed maps of high soil salinity for irrigated lands. The main findings of this study included:(1) Soil saturation percentage and irrigation amount are the two most important factors in soil salinity development in the Broadview Water District. (2) Logistic regression is a viable means of modeling salinity development on a site-specific basis when categorizing into salt-affected and unaffected soil.(3) Neighborhood effects are significant for the prediction of soil salinity in the top 1.2m of soil.(4) Agricultural management has a strong influence on the prediction of soil salinity in the top 0.6m of soil.

Technical Abstract: Salinity poses a potential agricultural and environmental threat to soil and water resources by reducing crop yields and degrading groundwater quality. The identification of potentially high areas of salinization within the rootzone of irrigated lands is a useful tool for mitgating the deleterious effects of salinity before they manifest themselves. A geographic information system, ARC/INFO, and a statistical software package, STATA, were used to develop a spatially-referenced, logistic regression model which identified areas of high soil salinity on a field scale. The study area was 2,395 hectares of the Broadview Water District in the San Joaquin Valley of central California. The significant variables in estimating soil salinity development were identified. The following results were found: (1) soil saturation percentage and irrigation amount were the two most important factors in soil salinity development in the Broadview Water District, (2) neighborhood effects were significant for the prediction of soil salinity in the top 1.2 m of soil, and (3) agricultural management had a stronger influence in the prediction of soil salinity in the top 0.6 m of soil than in the top 1.2 m of soil.