A Drought Index for Maryland Using Current Precipitation, Ground-Water, and Streamflow Data
WRD
PROJECT #: MD137
PROJECT CHIEF: Fisher, Gary
BEGIN DATE: 01-June-2000
END DATE: 30-September-2000
Customers currently supporting the project:
Maryland Geological Survey
U.S. Geological Survey
Problem
Maryland experienced a statewide drought in 1999 that was part of a regional drought affecting much of the eastern United States. Ground-water levels and streamflow data show that signs of drought were also present in 1998, although masked for much of the year by excess precipitation early in the year. Uncertainty about when a serious drought is occuring makes it difficult for water-resource managers to make informed and timely decisions regarding drought management. In particular, it is difficult to evaluate potential impacts of extended dry-weather periods on water resources and to evaluate recovery when precipitation returns to a near-normal pattern.
There have been at least three major regional droughts in the 20th century that have impacted Maryland, as well as numerous smaller droughts that have affected individual counties or groups of counties. Historically, major droughts have lasted for at least several years, and a "single-year" major drought is unprecedented. There is concern that the drought of 1999 may not have ended, but may be in a period where impacts are not as apparent as during the growing season. Long-term ground-water and surface-water records can provide an index that can be used to identify the beginning, end, and trend of a hydrologic drought. A study that will use historical hydrologic records to define empirical relationships between precipitation, streamflow, and ground-water levels during extended dry-weather periods is needed. Such relations will provide a basis for evaluating streamflow and ground-water responses to changing precipitation and will guide decision-making by water-resource managers. Understanding these relations will enable the development of numerical indices or scoring criteria that will describe the status of drought in Maryland.
Objectives
- Develop a drought-index model for Maryland using historical records of precipitation, streamflow, and ground-water data.
- Demonstrate application of the drought-index model to identify the beginning of the current drought in Maryland and its trend or end point.
Approach
Streams and water-table wells will be selected for several major physiographic regions in Maryland to provide data for an ongoing drought index determination. Relations among base-flow recession in streams, ground-water discharge, and drought have been well documented by Riggs (1963), Hanson (1987), and others (Hal, 1968). Base-flow recessions will be identified and rates will be quantified to provide a first-cut forecast of future streamflow, such as described by Riggs and Hanson (1969). Such an approach is useful for managing water resources during a drought, but can only be applied after ground water has discharged to streamflow. Additional lead time for forecasts could be achieved if suitable empirical relationships can be defined between ground-water levels and streamflow response, and between precipitation, and ground-water response. Comparisons will be made between precipitation, streamflow, and ground water and lag times, and magnitude of response to inputs will be determined. Candidate streams and wells for analysis in the selected regions are:
| Region | Stream-Gaging Station | Well |
| Appalachian | 03076600 Bear Creek at Friendsville | GA Bc 1 |
| Valley and Ridge | 01610155 Sideling Hill Creek at Bellgrove | WA Ac 1 |
| Great Valley | 01619500 Antietam Creek near Sharpsburg | WA Bk 25 |
| Piedmont (west) | 01643500 Bennett Creek at Park Mills | FR Eh 11 |
| Piedmont (east) | 01583600 Beaverdam Run at Cockeysville | BA Dc 444 |
Streams and wells will not have to be located in the same drainage basins, as long as the wells are located in surficial geologic formations that are representative of those draining to the stream-gaging station. However, precipitation data must represent input to both the well and stream. Respective precipitation data from the Confluence (Pa.), Hancock, Md. Hagerstown, Md. Frederick, Md. and Towson, Md. weather stations will be compared with streamflow and ground-water-level data.
Investigation will seek to identify empirical relationships between precipitation and ground-water response and between ground-water levels and streamflow response, including lag times and relative magnitudes.
