State Water Resources Research Institute Program

**Project ID:** 2007DC86B

**Title:** Gradual Variation Analysis for Groundwater Flow in the District of Columbia

**Project Type:** Research

**Start Date:** 3/01/2007

**End Date:** 2/29/2008

**Congressional District:** DC

**Focus Categories:** Groundwater, Methods, Models

**Keywords:** Groundwater flow, Gradual Variation, Analysis, Computer Simulation, Real data processing, MODFLOW

**Principal Investigator:** Chen, Li

**Federal Funds:** $ 15,000

**Non-Federal Matching Funds:** $ 39,409

**Abstract:** Groundwater flow in the District of Columbia greatly influences the surface water quality in urban areas. The current methods of flow estimation, based on Darcy's Law and the groundwater flow equation, can be described as the diffusion equation (the transient flow) and the Laplace equation (the steady-state flow). The Laplace equation is a simplification of the diffusion equation under the condition that the aquifer has a recharging boundary. The practical way of calculating is to use numerical methods to solve these equations. The most popular system is called MODFLOW, which was developed by USGS. MODFLOW was based on the finite-difference method on rectangular Cartesian coordinates. MODFLOW can be viewed as a "quasi 3D" simulation since it only deals with the vertical average (no z-direction derivative). Flow calculations between the 2D horizontal layers use the concept of leakage.

In this proposal, we will establish a true 3D simulation model based on gradually varied functions. These functions do not rely on a rectangular Cartesian coordinate system. A gradually varied function can be defined in a general graph or network. Gradually varied functions are suitable for arbitrarily shaped aquifers. However, to ensure the accuracy of the calculation, we will add the finite-difference method to our research. That is to use MODFLOW to calculate local and small-region flow, and to use gradual variation to compute the regional or global data. The proposed research will have three steps: (1) Use the gradually varied model and simulation based on the groundwater flow in DC. (2) Decompose the DC aquifer into major 3D cells, by using MODFLOW for each cell and then the gradual variation method on the entire region. (3) Compare the results of MODFLOW, Gradual variation, and the mixed method.

The second phase of the project will add a time dimension, i.e. we will compute the results of the groundwater flow in DC for 10 years and build a boundary variation model to predict the groundwater flow in the District of Columbia for the next 5-10 years.

Progress/Completion Report, PDF