Scope and Method of Study: The main objective of this research was to study the influence of weather and topography on the temporal and spatial variability of wheat yields in Oklahoma by examining the soil moisture distribution and storage and evapotranspiration (ET) variability that are induced by topographical changes, and evaluating their correlation with wheat yield variance across the study field. A stratified random sampling method was used to take soil samples in the study field. Elevation data were collected with a GPS unit on a semiregular grid with a distance from 2-8 m. The weather data used for this research was obtained from the Medford Mesonet Station. The seasonal ET data were calculated using weather data and soil data. Topographical factors, aspect, slope and curvature were derived from the DEM using the elevations collected in the study field with the GPS unit. Their correlations with yields were examined. A topography-related soil moisture index (TRSMI) was constructed from combining two topographic variables, slope and curvature, with one soil variable, available soil water content. Its correlations with yields for each year were examined. The correlation between seasonal ET and yields for each growth season was examined.

Findings and Conclusions: The examination of relationships between topographical factors, slope and curvature, soil property, and available soil water content for the entire study field showed that available soil water content has the highest correlation with yield among the factors listed above. This demonstrated that the amount of water soils can retain is an important factor in yield variability on a field scale. The correlation coefficient between TRSMI and yields indicates that a positive correlation exists between yield and TRSM. The correlation coefficient between seasonal ET and yields in dry years and years with adequate precipitation is higher than that in wet years. The patterns of correlation coefficients between yields and the slope, curvature, available soil water content, TRSMI, and seasonal ET are for the sample locations are similar to those for the entire study field, but the correlations were stronger.