Monitoring Tree Roots over Time with GPR

For over ten years, ground-penetrating radar (GPR) has been used in a variety of applications to detect tree roots, quantify their spatial distribution, and to estimate root diameter and biomass. Most surveys are performed once, offering a point-in-time assessment of root parameters. My own approach has been to link a small destructive sample using soil cores to GPR data collection, allowing direct comparison of target and acquired signal. All collection parameters (range, gain, dielectric, etc.) are held constant, since the data are collected over the course of a day or two, there is minimal temporal variation. for measures spread out over a longer period of time, differences in soil moisture, water usage, and drainage patterns will likely play havoc with root resolution and parameter estimation, making time series comparisons (months to years) problematic unless such environmental “noise” can be accounted for. <br>to better understand seasonal variation in root detection, several 20 m transects were established in a 25 year-old balsam fir (Abies balsamea) plantation in South Burlington, Vermont. the transects were located on a deep sandy loam and were scanned 20 times with a 1500 MHz antenna between March and November 2010 using identical settings. A reference transect was created by horizontally inserting 8 mm aluminum rods into undisturbed soil at several depths to accurately measure signal velocity and variation with depth. Soil moisture ranged from 7% to 32% creating dielectric values from 6 to 23. After processing to reduce clutter and isolate root biomass, the radargrams were analyzed with image analysis to quantify the area occupied by roots. Preliminary results are encouraging; GPR data (area) under one tree were compared to soil moisture and dielectric values for 20 different days. Though variation in GPR data was large (order of magnitude) over the season, it was exponentially related to both soil moisture (R2 0.48) and soil dielectric (R2 0.81).