The Spatiotemporal Variability of Measured Soil Moisture Across Multiple Cropped Agricultural Fields.

The ability to characterize drought within a given field or initiate irrigations from monitored soil moisture data hinges upon the ability of the instrument to characterize soil moisture at the sampled point and extrapolate that information across the landscape and time.  Therefore, the objective of this study was to determine if a limited number of soil moisture sensors deployed into a dryland variety trial could accurately characterize the volumetric water content (VWC) at a given point within the field and if this measurement could be extrapolated out to the field scale from the very small sphere of influence characterizing the utilized soil moisture sensors.  During the 2013 growing season soil moisture sensor deployments were made into seven dryland cotton (Gossypium hirsutum, L.) variety trials across the Mid-South and Southeastern Regions of the U.S.  Inference on VWC of each monitored profile was determined by (4) Decagon EC-5 sensors (Decagon Devices Inc., Pullman, WA) installed at depths of 7.5, 22.5, 45, and 75 cm.  Sensor reported VWCs related well to soil water content measured at installation (r²=0.617).  Relationships between within-location nodes varied but were typically moderate to strong (0.646<r²<0.988).  Mean relative difference analysis indicated a minimum of one node in six of seven instrumented trials was characterized as temporally stabile (σ<5%). The moderate relationships noted between sensor-estimated and measured VWC and the presence of temporal stability noted at nearly all locations suggests a limited number of soil moisture sensors deployed into a dryland variety trial could accurately characterize the VWC of the trial.  Still, further research should be conducted to see if trends in temporal stability remain through multiple seasons and to investigate the relationship between varietal yield responses to the calculated drought stress-indices in more rainfall-normal years.