Estimating groundwater withdrawals for selected principal aquifers in the southeastern United States: Floridan aquifer system and Biscayne aquifer
Project Chief: J.W. Grubbs
Cooperator: None
Period of Project: October 2009-September 2012
Problem Statement
Figure 1. An example of determining optimal sample size using a hybrid approach that combines a census of the largest users (top stratum) with random sampling of the remaining users. This figure is reproduced from the Committee on USGS Water Resources Research (2002, chapter 5, p. 94) and based on data from irrigation withdrawals in Arkansas during 1997, which were obtained from the USGS Arkansas District Office.
Detailed accounting of the Nation’s water use is carried out by compiling data from all known users. Sampling techniques provide an alternative approach, in which data are collected from a subset of water users and used to estimate the total withdrawal for a specified area, aquifer, and set of water-use categories. To assess whether such an approach is feasible, these sampling techniques must be evaluated by quantifying the number of samples required to achieve acceptable levels of accuracy for estimating withdrawals from regionally significant aquifers.
A separate problem associated with water-use assessment is estimating site-specific withdrawals for agricultural irrigation. Estimates of agricultural irrigation are commonly required because irrigation systems are often unmonitored. In these cases, irrigation withdrawals are typically accounted for at a county level, using irrigation-demand estimates that are based on weather conditions, crop type, and areas of irrigated crops that are reported as aggregate totals for individual counties. Site-specific withdrawals for individual irrigation systems can be estimated using simple approaches for ‘disaggregating’ the county totals, but quantitative measures of the accuracy of these site-specific estimates are needed to ensure that the estimates are used appropriately.
Objectives
The primary objective of the project was to evaluate whether the total groundwater withdrawals from two aquifers (the Floridan aquifer system and the Biscayne Aquifer) for public supply, commercial, industrial, mining, and power-generation uses can be estimated accurately using statistical sampling techniques.
A secondary project objective was to evaluate the accuracy of estimating site-specific groundwater withdrawals from withdrawal totals that have been aggregated at the county level.
Approach
The primary objective of the project was accomplished by first compiling a dataset of individual public supply, commercial, industrial, and mining withdrawals in calendar years 2000 and 2005. This dataset will be used to determine the total number of random samples needed to estimate the total withdrawals for these water-use categories from either the Floridan aquifer system or the Biscayne Aquifer to an accuracy of 1, 2, 5, and 10 percent of the total withdrawal. These water-use categories were used as sampling categories or “strata”, to try to improve sampling efficiency (fig. 1). The project also evaluated whether sampling efficiency can be improved by further dividing these strata into a “top” substratum, in which a census approach is applied to a limited number of major users, and a “bottom” substratum, in which the remaining, more numerous minor users are randomly sampled.
The secondary project objective was accomplished by first estimating site-specific withdrawals for a selected set of counties in the study area (the area covered by the Floridan aquifer system and Biscayne Aquifer) for which metered data exist at individual irrigation wells. This was accomplished by scaling the county-level irrigation estimate by the ratio of the permitted withdrawal at a given site to the total permitted withdrawal amount for irrigation in that county. Summary statistics of the difference between estimated and measured site-specific withdrawals were then computed to assess the statistical properties of the site-specific differences (for example, bias and precision).
Results
Not yet available
Information Product
A journal article will describe the estimation of groundwater withdrawals for selected principal aquifers in the southeastern United States through stratified, random sampling.
A USGS data-series report will summarize the demand-based estimates of irrigation withdrawals from the Floridan aquifer system.
| 
U.S. Department of the Interior |
U.S. Geological Survey
