MONITORING STORMWATER INFILTRATION IN A VACANT LOT COMPARING TIME-LAPSE ELECTROMAGNETIC INDUCTION AND ELECTRICAL RESISTIVITY TOMOGRAPHY

Abstract

Vacant lots in cities and surrounding urban areas can potentially be used for stormwater management because they are pervious. However, the extent to which vacant lots provide pervious cover to increase infiltration and reduce stormflow is poorly understood. The goal of this study was to develop faster methods for monitoring stormwater infiltration to improve characterization of heterogeneous urban systems. Geophysical techniques are capable of mapping and characterizing subsurface materials, but are often limited by time and sensitivity constraints. In this study, the infiltration characteristics of a vacant lot created by the demolition of a house was characterized using a series of modeling, field and lab experiments. Site characterization under background conditions with an EM Profiler was used to map zones of different fill materials. Three zones were identified in the study site: grass area, driveway area, and a former house area. Transient soil moisture conditions were monitored during irrigation tests using two geophysical methods (electrical resistivity tomography [ERT] and electromagnetic induction [EM]) to evaluate method sensitivity and differences between the three zones. ERT proved more sensitive than EM profiling at detecting changes in the three zones. Soil moisture changes in the driveway area were particularly difficult to detect using EM. The EM Profiler showed a reduction rather than increase in conductivity at the start of irrigation and storms, which was attributed to flushing of high conductivity pore fluids by dilute irrigation or rain water. This explanation was supported using Archie’s Law to model the response of apparent conductivity under highly conductive pore fluid conditions. The EM Profiler was also used under natural precipitation conditions to quickly monitor areas too large for the ERT to reasonably survey. The results suggested that EM instrument drift needs to be corrected to make the method more sensitive. It was difficult to detect differences in hydrologic characterization between areas of the vacant lot using traditional soil point measurements because of the inherent spatial variability. The most useful point measurement was soil moisture loggers. Data from soil moisture loggers was used to parameterize the model; in addition, the soil moisture loggers showed a slow drying period. By combining the EM Profiler method with soil moisture data and applying corrections for drift, some improvement in sensitivity might be achieved. Quantitative characterization of fill material was shown by ERT, which detected more heterogeneous infiltration in the area of the former house than in the grass area.