Joel E. Atwood and Marek Zreda
Department of Hydrology and Atmospheric Science
The University of Arizona, Tucson, AZ
Rainfall distribution has been shown to be highly heterogeneous and therefore point measurements can poorly represent rainfall when applied to field scales. We modified a cosmic ray neutron detector (CRD) to produce soil moisture data for a 10-cm control volume. From the soil moisture time series, we developed an empirical relationship relating soil moisture to the flux of moisture leaving the control volume due to drainage. Using mass conservation principles, a model was developed using superposition of drying curves to inverse model the rainfall fluxes needed to produce the soil moisture time series. This approach was applied to the monsoonal period for three years, from 2010 to 2012, at the Manitou research site. The model has trouble predicting the exact timing of events on an hourly scale but performs better at 6-hour, 12-hour, and 24-hour integrated scales. The model tends to overpredict the magnitude of events smaller than 5 mm and underpredict larger magnitude events. Applying a scaling function that takes into account the changes in the CRD sensitivity as a function of water content improves results. These estimations represent rainfall, spatially averaged at a hectare scale, and further research is needed to assess the validity of these estimations.