USE AND LIMITATIONS OF GROUND PENETRATING RADAR TO MAP INFILTRATION IN AN URBAN STORMWATER BASIN

Green infrastructure designed to increase stormwater infiltration is an increasingly common element of new construction in urban areas, but seldom includes monitoring and performance assessment. Temple’s new Science Education and Research Center has a stormwater basin that is lined, filled with drainage pipes and gravel, topped with approximately 1m of urban soil and sod, and equipped with monitoring wells. Despite the layered design and uniform appearance of the grass-covered surface, water levels responded differently to the same recharge events. We used ground-penetrating radar (GPR) to look for uneven infiltration through the shallow urban soil atop the basin. GPR can be effective for mapping soil moisture distributions, but urban soil poses unique challenges. An unknown mixture of glass, rocks, bricks, and other debris can lead to signal attenuation, absence of continuous reflection horizons, and high background clutter. Furthermore, clay content limits depth of penetration. These complications become increasingly relevant when time-lapse GPR is used to look for temporal changes. A 15 m by 5 m grid was laid out and two sprinkler heads were used to water the grid for two hours. Five 3D GPR surveys with an 800 MHz antennae were conducted: one background and four post-irrigation to monitor the soil moisture recovery. Five capacitance sensors were placed in the soil to provide continuous volumetric water content (VWC) for comparison. We were able to see the wetting and subsequent drying of the soil in both time-lapse GPR and VWC sensor data. Both showed a spatially heterogeneous pattern of infiltration, but the GPR did not provide a continuous map of the change due to the heterogeneity of the signal. A subsequent terrestrial LiDAR survey suggests that in addition to heterogeneous properties of urban soil, microtopography may contribute to the non-uniform stormwater infiltration. Our study shows that the storm response of engineered structures designed to decrease urban run-off is complex and performance should be monitored, particularly as the effectiveness may change over time. Geophysical methods such as time-lapse GPR show promise as a non-intrusive source of monitoring information but the complexity of urban soils complicates interpretation.