Proceedings of the 2022 Mississippi Water Resources Conference

Year: 2022 Authors: Guthrie G.M., Jin G.

Groundwater data is commonly used by state and federal organizations as a component of drought assessment during dry seasons. The incorporation of this data in monthly assessments may be of limited usefulness due to the delayed response, or lag time, of water levels to precipitation events resulting from the time required for recharge waters to migrate to the saturated zone. Daily water-level measurements are traditionally presented as time series hydrographs that are seemingly complex, noisy, and thus difficult to interpret due to diurnal, seasonal, and even decadal variations. Continuous water level (WL) data from eight drought monitoring wells in different hydrogeologic provinces and precipitation (PPT) data from nearby stations have been analyzed using spectral analysis by discrete Fourier transform to evaluate primary seasonal and trend events. The technique can reveal both the periods and magnitudes of all seasonal, cyclical, and/or long-term (more than one year) wet and drought events and provide important parameters to decompose time series data into seasonal, trend, and random components using seasonal and trend decomposition. The resulting trend components are then evaluated with correlation analysis to establish groundwater lag time in the wells in response to precipitation events. Lag time is determined from cross-correlation diagrams as the amount of time shift required to attain a peak correlation coefficient for both WL and PPT trend curves. Lag time is not consistent among wells, ranging from 60 to 180 days, and the differences may reflect different environmental and hydrological properties of the wells, such as elevation, confining conditions, distance from recharge area, permeability, hydraulic conductivity, unsaturated zone hydraulic properties. Lag time analysis provides an invaluable tool to accurately model/forecast groundwater changes using machine learning and statistical techniques and the ability to predict anticipated groundwater levels at the beginning of the drought season. Using this methodology will be an invaluable tool for water managers and stakeholders to assess potential water availability issues.