S-199
Assessment of ground-water quality trends under the
U.S. Geological Survey National Water-Quality Assessment
Program (NAWQA) included the analysis of samples collected
on a quarterly basis for 1 yr between 2001 and 2005. he
purpose of this quarterly sampling was to test the hypothesis
that variations in the concentration of water-quality parameters
of selected individual wells could demonstrate that the intra-
annual variation was greater or less than the decadal changes
observed for a trend network. Evaluation of more than 100
wells over this period indicates that 1 yr of quarterly sampling
is not adequate to address the issue of intra-annual variation
because variations seem to be random and highly variable
between different wells in the same networks and among
networks located in different geographical areas of the USA. In
addition, the data from only 1 yr makes it impossible to assess
whether variations are due to univariate changes caused by land
use changes, hydrologic variations due to variable recharge,
or variations caused by ground-water pumping. hese data
indicate that funds allocated to this activity can be directed to
the collection of more effective trend data, including age dating
of all wells in the NAWQA network using multiple techniques.
Continued evaluation of data and updating of monitoring
plans of the NAWQA program is important for maintaining
relevance to national goals and scientific objectives.
Evaluation of Intra-annual Variation in U.S. Geological Survey National Water Quality
Assessment Ground Water Quality Data
Michael R. Rosen,* Frank D. Voss, and Jorge A. Arufe USGS
T
he U.S. Geological Survey National Water Quality Assessment
(NAWQA) program samples ground-water wells for water
quality at different spatial and temporal scales (Rosen and Lapham,
2008). Determining the most cost effective and scientifically
beneficial temporal sampling period is essential to being able to
adequately assess changes in ground-water quality (Rosen, 2004).
Although optimization techniques have been developed for high-
intensity local sampling programs for contaminant plumes (e.g.,
Barcelona et al., 1989; Herrera and Pinder, 2005), optimization
techniques for large networks are less common. A study of the effect
of temporal variability during sample collection on analysis of spatial
variability was conducted over an area of approximately 20 km
2
in Ohio (Ritzi et al., 1993), but this study was designed to assess
whether the length of time it took to sample a spatially distributed
set of wells induced greater variability than the inherent spatial
variability in the water-quality parameters analyzed. However, in all
of these studies, intense sampling, on the order of one sample every
1 wk, were undertaken to determine optimal sampling periods.
One component of the NAWQA ground-water quality sampling
plan is to sample five wells in a 30-well network on a quarterly basis
(see Rosen and Lapham, 2008) to determine if this intra-annual varia-
tion (of selected parameters) could be used to demonstrate whether
decadal changes in water quality were greater or less than the intra-an-
nual variation. Quarterly sampling was designed to be conducted in 1
yr within the decadal sampling period. However, these data were not
collected to establish an optimal temporal sampling strategy. Sampling
for only 1 yr (four sampling points) limits the ability to conduct statis-
tical tests on the data set, but the number of wells sampled nationwide
(>100 wells) allows a comparison of intra-annual variations in differ-
ent land use settings and geographic areas.
Quarterly sampling (the collection of samples once every 3 mo for
1 yr) was deemed important in the NAWQA ground-water trends
design because the magnitude of change within a single year may be
greater than long-term changes. For example, using Monte Carlo sim-
ulations, Yue et al. (2002) showed that the power of the Mann-Ken-
dall test to detect trends depends on the preselected significance level,
the absolute magnitude of the trend, the number of observations, and
the amount of variation within the interval of interest. his means
that for any site that displays seasonal change in chemistry that is simi-
Abbreviations: AgLUS, agricultural land use study; LUS, land use study; MAS, major
aquifer study; NAWQA, National Water-Quality Assessment Program; UrLUS, urban
land use study; VOCs, volatile organic compounds.
M.R. Rosen, USGS, 2730 North Deer Run Rd., Carson City, NV 89701. F.D. Voss, USGS,
934 Broadway, Suite 300, Tacoma, WA 98402. J.A. Arufe, USGS, 160 N. Stephanie St.,
Henderson, NV 89074.
Copyright © 2008 by the American Society of Agronomy, Crop Science
Society of America, and Soil Science Society of America. All rights
reserved. No part of this periodical may be reproduced or transmitted
in any form or by any means, electronic or mechanical, including pho-
tocopying, recording, or any information storage and retrieval system,
without permission in writing from the publisher.
Published in J. Environ. Qual. 37:S-199–S-208 (2008).
doi:10.2134/jeq2007.0052
Received 29 Jan. 2007.
*Corresponding author (mrosen@usgs.gov).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA
SPECIAL SUBMISSIONS