2007 GSA Denver Annual Meeting (28–31 October 2007)

Paper No. 12
Presentation Time: 4:50 PM

WATER PARTITIONING WITHIN THE UNSATURATED ZONE AFTER MULTIPLE NATURAL PRECIPITATION EVENTS DETECTED BY GPR


JACOB, Robert W. and HERMANCE, John F., Department of Geological Sciences, Brown University, 324 Brook Street, Providence, RI 02912-1846, Robert_Jacob@Brown.Edu

Soil water in the vadose zone connects atmospheric water and other surface water to the saturated subsurface. We report short-term changes in the vertical distribution of soil water content (SWC) after multiple storm events at a long-term geophysical test site in Southeastern New England using ground penetrating radar (GPR). GPR is a non-invasive geophysical technique which is sensitive to the high dielectric polarizability of the water molecule. The typical analysis of multiple offset reflection/refraction soundings consists of using the reflected signal traveltimes (hence velocity and depth) to produce the vertical profile of GPR velocity. We show that with careful analysis of high quality soundings it is possible to determine the SWC with precisions on the order of ±1% (water volume per unit soil volume). We have additionally found that the information from the air refracted and ground refracted phases can be essential when observing the redistribution of SWC.

The time-dependent partitioning of SWC between the nominally 1 m thick topmost organically-rich soil layer and an underlying nominally 3 m thick organically-poor gravelly sand is described in relation to five precipitation events. These events range in total rainfall amount from 1.6 cm to 7.9 cm. The SWC increased, as expected, for each rain event in the soil layer. Regardless of differences in the total rainfall amount, the SWC of the soil layer increased to 26%. Thus, we conclude that the field retentivity of the soil layer is reached for all storms, and that any excess water infiltrates deeper into the gravelly sand. This is confirmed for 4 events by the ground refraction from the top of the gravelly sand layer. In addition, the SWC of entire gravelly sand layer is observed to increase in response to only 3 storms. We find that the lack of response in the deeper layer is due to the antecedent SWC of the topmost layer and the elapsed time between rainfall and GPR data collection.

GPR proved to be invaluable for this research because it is a completely non-invasive technique that samples over a larger volume than traditional SWC techniques. Furthermore, we show that the time-dependant vertical redistribution of SWC after precipitation events make it possible to observe changes in subsurface layering due to hydrologic properties of the geologic material.