Virginia Tech presents preliminary groundwater assessment showing recharge variability, recommends monitoring and gauges

Virginia Tech researchers gave the Floyd County Board of Supervisors a preliminary groundwater assessment on March 25 that summarizes thousands of well reports, maps groundwater flow within subwatersheds and models how precipitation becomes recharge across the county.

The presentation, led by Dr. Mark Wittison, a Virginia Tech engineer, and co‑presenter Ed Mendez, showed the project compiled roughly 3,660 well records from the Virginia Department of Health and county sources and used the USGS soil‑water balance (SWB) model on a 30‑meter grid to estimate recharge across 15 subwatersheds. "The average recharge rate through the period of record was about 11 inches per year," Wittison said, but drought years produced far lower numbers: the extended drought regionally analyzed dropped average recharge to about 5.4 inches per year and the single worst year studied (1988) produced roughly 3.4 inches per year.

The researchers summarized well construction data and reported averages from the compiled well logs: mean total well depth about 325 feet, mean depth to bedrock about 62 feet, mean depth to the upper water‑bearing zone about 76 feet and a mean depth to the bottom of reported water‑bearing zones about 238 feet. The presenters emphasized that groundwater in Floyd County sits mostly in fractured and weathered rock near the surface and that productive wells commonly tap multiple water‑bearing fracture zones. In the dataset, Wittison said, "the more water‑bearing zones that you hit, the higher your yield is going to be."

Ed Mendez described the SWB modeling approach and key inputs: daily precipitation and temperature records (1980–February 2023), land‑cover (2021), soil groups and available water content. The model runs daily on grid cells and accounts for interception, root‑zone storage, evapotranspiration and runoff before counting recharge to deeper groundwater. Mendez showed spatial maps: recharge is higher where precipitation and forested cover are greater and lower across the county during drought snapshots. He noted Beaver Dam Creek, the county's largest subwatershed, had average modeled recharge near 10 inches per year across the period of record but dropped to roughly 2.8–5.2 inches during the drought snapshots studied.

Researchers framed a conceptual groundwater model that ties recharge at higher elevations and forested areas to groundwater flow toward local streams and springs. "A lot of that water is flowing toward these creeks," Wittison said of subsurface flow paths visualized in the study. They stressed the local nature of recharge for fractured‑rock systems: water supplying a well usually comes from a nearby recharge zone rather than a countywide aquifer.

Recommended next steps focused on adding continuous observations to reduce uncertainty. The team proposed: (1) installing streamflow gauges at selected subwatershed outlets, (2) deploying additional real‑time rain gauges, and (3) instrumenting a network of representative wells with recording transducers and telemetry to capture continuous water‑level records. Mendez said ground‑truthing the model to streamflow and monitored water‑level records is how they would calibrate and improve the estimates. Wittison added Virginia Tech can help instrument wells and provide public access to real‑time data.

County staff noted the board already funded seven stream gauges and that the Virginia Department of Emergency Management had installed three real‑time rain gauges and a flood gauge, which the presenters said would be useful for the proposed monitoring package. Supervisors pressed presenters on where gauges should sit and on how much of the county needed instrumentation; presenters said a balanced network of stream gauges, rain gauges and a selection of monitored wells would greatly improve the county's ability to track recharge and identify drought impacts.

The presenters emphasized the study was preliminary and that additional VDH well records (dating to the late 1970s) were being added to the dataset to increase the sample size and detect which wells failed during past droughts. They urged the county to prioritize monitoring to allow a quantitative water budget and to inform land‑use and stormwater decisions that could help retain or increase recharge.

The presentation concluded with the board expressing appreciation and with staff and Virginia Tech agreeing to refine the model and seek calibration data and monitoring funding.