Quantifying Groundwater Recharge Dynamics Using a Process-Based Distributed Hydrologic Model

Event Date/Time: 
December 1, 2017 - 10:00am
Event Location: 
3405B Engineering
Guoting Kang
Ph.D Dissertation Defense


Groundwater – the lifeblood of groundwater-dependent ecosystems and societies – is facing unprecedented threats from over-extraction, contamination, and changing climate. Groundwater recharge provides a sustainable source of water for aquifers and plays an important role in both surface and sub-surface domains. Understanding and accurately estimating the rate, location, and timing of major recharge events and their seasonal and inter-annual variability is key to safely matching societal needs of water and to maintaining healthy groundwater-dependent ecosystems. This work attempts to understand and quantify recharge dynamics in an agricultural watershed in the Ottawa County, Michigan using field observations of baseflows, groundwater heads, satellite-based evapotranspiration (ET) products and an integrated, process-based hydrologic model. Specific objectives of the work are to: (1) understand the spatial and temporal distribution of high- and low-recharge events and (2) assess the relative impacts of climate, land use, soils, and topography on the spatiotemporal distribution of recharge within the region. County-wide synoptic and time-series baseflow data collected from over 40 small streams between July and November of 2015 were used to quantify the uncertainties in recharge estimation. Precipitation data represent important inputs to hydrologic models and have a major influence on model performance and the estimated recharge. Compared to data from a typical network of rain gauges, the Next-Generation Weather Radar (NEXRAD) provides precipitation data at a much higher spatial resolution. NEXRAD data were blended with traditional rain gauge data to estimate recharge and to evaluate differences relative to recharge estimated using rain gauge data alone. Results indicate that caution should be exercised in using NEXRAD precipitation data for recharge estimation. The representation of recharge and its variability within a numerical model are closely related to the representation of meteorological forcing fields and their spatial structure, land use and land cover, the hydraulic properties of underlying soils and aquifers as well as topography -- all of which are represented to varying degrees of accuracy depending on the mesh resolution employed and the algorithms used to represent subgrid-scale processes. To understand the effects of grid resolution on recharge and to identify optimal resolution relative to the size of the watershed, models were setup with different grid resolutions.  Recharge patterns follow precipitation patterns more closely at coarse grid sizes, since the characteristics of LULC, terrain and hydraulic properties are smoothed at this resolution. Insights gained from the study are expected to aid in sustainable management of natural resources, particularly groundwater-dependent ecosystems.