Skip to main content
ARS Home » Pacific West Area » Pullman, Washington » Northwest Sustainable Agroecosystems Research » Research » Publications at this Location » Publication #246056
Title: Spatio-temporal patterns of soil water storage under dryland agriculture at the watershed scale

Author
item IBRAHIM, HESHAM - Washington State University
item Huggins, David

Submitted to: Journal of Hydrology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/20/2011
Publication Date: 5/4/2011
Citation: Ibrahim, H.M., Huggins, D.R. 2011. Spatio-temporal patterns of soil water storage under dryland agriculture at the watershed scale. Journal of Hydrology. 404:186-197.

Interpretive Summary: The availability of soil water is critical for crop growth and also influences nutrient use and losses. Soil water is not uniformly distributed across fields and patterns in soil water availability can develop due to precipitation, soil properties, topographic features and land use. We used a new statistical technique called empirical orthogonal function (EOF) analysis to see if we could characterize and better predict patterns of soil water across a 37-ha field of the Washington State University Cook Agronomy Farm near Pullman, WA. During the period 1999-2006, soil water (0 to 1.5 m depth) was measured in the spring prior to planting and again in the fall after harvest on 123 geo-referenced points. Application of the EOF analysis showed that we could explain 73 to 75% of the soil water variability. Furthermore, we discovered that the soil water patterns varied between wet and dry time periods. Under wet spring conditions, topographic attributes were important factors in explaining the field patterns of soil water. In contrast, under dry fall conditions, soil properties become more related to the patterns of soil water. We also compared the EOF analysis to a more standard statistical method, stepwise multiple linear regression (SMLR) and found that both methods accurately estimated average soil water over the entire watershed. However, the EOF-based method has the advantage of estimating the soil water variability at different times, whereas in SMLR only soil water measurements at a single time can be used. We concluded that the EOF-based method will be very useful to scientists and potentially growers for predicting field patterns of soil water and therefore aiding predictions of crop productivity, fertilizer needs, and nutrient use and loss across watersheds.

Technical Abstract: Soil water patterns vary significantly due to precipitation, soil properties, topographic features, and land use. We used empirical orthogonal function (EOF) analysis to characterize the spatial variability of soil water across a 37-ha field of the Washington State University Cook Agronomy Farm near Pullman, WA. During the period 1999-2006, soil water (0 to 1.5 m depth) was measured gravimetrically in the spring prior to planting and again in the fall after harvest on one third of the established 369 geo-referenced points, representing one of three similar field blocks, A, B and C. The first EOF generated from the three blocks explained 73 to 75% of the soil water variability. Soil water patterns based on EOF interpolation varied between wet and dry conditions. Under wet spring conditions, elevation and wetness index were the dominant factors in controlling the spatial patterns of soil water. In contrast, under dry fall conditions, soil properties (apparent electrical conductivity and bulk density) become more related to spatial patterns of soil water. The EOFs generated from block B, which represented average topographic and soil properties, provided better estimates of soil water over the entire watershed with larger Nash-Sutcliffe Coefficient of Efficiency (NSCE) values, especially when the first two EOFs were retained. The EOF interpolation method to estimate soil water variability worked slightly better during spring than fall, with average NSCE values of 0.23 and 0.20, respectively. A comparison between the EOF-based soil water estimation method and direct stepwise multiple linear regression (SMLR) of soil water measurements against topographic and soil attributes showed that both methods accurately estimated average soil water over the entire watershed. The EOF-based method has the advantage of estimating the soil water variability based on soil water data from several measurement times, whereas in SMLR only soil water measurements at a single time are used. The EOF-based method can also be used to estimate soil water at any time other than measurement times, assuming the average soil water of the watershed is known.