&EPA
United States Office of Acid Deposition, Environmental EPA/600/4-88/023
Environmental Protection Monitoring and Quality Assurance June 1988
Agency Washington DC 20460
Research and Development
National Stream Survey -
Phase I:
Field Operations Report
iT£ w XT*»* '*j& * / " § jr*
4 ? ^^r™$f^'1^ i^^" f'
*>,
-, ^•»»,
-------
SUBREGIONS OF THE NATIONAL STREAM SURVEY-PHASE I
Northern
Appalachians (2Cn)
Valley and Ridge (2Bn)
Southern Blue Ridge (2As)
(Pilot Study)
Poconos/Catskills (ID)
NY\
Ozarks/Ouachitas (2D)
Mid-Atlantic
Coast'al Plain (3B)
Southern Appalachians (2X)
-------
EPA 600/4-88/023
June 1988
National Stream Survey
Phase I
Field Operations Report
A Contribution to the
National Acid Precipitation Assessment Program
U.S. Environmental Protection Agency
Office ol Research and Development
Washington, DC 20460
Environmental Monitoring Systems Laboratory - Las Vegas, NV 89119
Environmental Research Laboratory - Corvallls, OR 97333
tionmental Protection Agenof.
Library (5PL-16)
i;jrn Street, Room 137Q-
-.1 60604
-------
Notice
The information in this document has been funded wholly or in part by the US
Environmental Protection Agency under Contract No. 68-03-3249 and 68-03-3050 to Lock-
heed Engineering and Sciences Company, Inc., No. 68-03-3246 to NSI, No. 68-03-3439 to
Kilkelly Environmental Associates, No. 68-02-3889 to Radian Corporation, and Interagency
Agreement No. 40-1441-84 with the U.S. Department of Energy. Y
Mention of corporation names, trade names, or commercial products does not con-
stitute endorsement or recommendation for use.
™s document is one volume of a set which fully describes the National Stream
Survey - Phase I. The complete document set includes the major data report quality
assurance plan, analytical methods manual, field operations report, processing laboratory
operates report and quality assurance report. Similar sets are being produced for each
Aquatic Effects Research Program component project. Colored covers, artwork and the
use of the project name in the document title serve to identify each companion 'document
SGI.
The correct citation of this document is:
Hagley C. A., C. L. Mayer, and R. Hoenicke. 1988. National Stream Survey - Phase I
Vegas nSReP°rt' ^ 6°°/4-88/023- U'S' Environmental Protection Agency, Las
-------
Abstract
The National Stream Survey was conducted during the spring of 1986 as a synoptic
chemical survey to characterize streams in the mid-Atlantic and southeastern regions of
the United States which were thought to be potentially susceptible to acidic deposition.
The survey included three distinct parts: a Phase I survey of streams in the mid-Atlantic
region; a Screening survey designed to assess the need for future Phase I studies in the
United' States; and an Episodes Pilot survey designed to provide a preliminary assessment
of the frequency, duration, and characteristics of storm episodes in the mid-Atlantic states.
The Episodes Pilot survey was conducted on a subset of Phase I streams and replaced
normal Phase I sampling during rain events. It also served to evaluate sampling designs
and logistical protocols for future episodes studies.
This report describes the survey planning, protocol development, personnel require-
ments, field operations, and logistical aspects of all three components of the National
Stream Survey. Because of the large scope and geographical area covered by the survey,
sampling regions were subdivided into four areas, each containing approximately the same
number of streams. Samples were collected, shipped at 4 °C, and received within 24 hours
by a central processing laboratory. Sampling was completed on schedule, and 447 out of
a total of 479 streams were sampled. A detailed evaluation of episodes sampling is pro-
vided with recommendations for future consideration.
This report was submitted in fulfillment of Contract No. 68-03-3249 by Lockheed
Engineering and Sciences Company, Inc. under sponsorship of the U.S. Environmental
Protection Agency. This report covers a period from October 1984 to June 1986 and work
was completed as of June 1988.
in
-------
-------
Confenfs
Notice jj
Abstract jjj
Figures vii
Tables viii
Acknowledgements ix
Introduction 1
Phase I Survey 1
Screening Area Survey 2
Episodes Pilot Survey 3
Preparation for Field Operations 4
Overview 4
Survey Planning 4
Division of Study Area 5
Sample Site Information 5
Sampling Sequence and Scheduling 5
Laboratory Location 7
Protocol Development 7
Laboratory Protocol 7
Sampling Protocol 7
Guidelines for Sampling 8
Personnel 8
Staffing Requirements 8
Personnel Duties 9
Personnel Selection and Training 9
Field Operations 10
Overview 10
Daily Base Site Operations Summary 10
Sampling 10
Daily Sampling Operations 10
Sampling Methods 13
Quality Assurance of Field Operations 14
Logistics 15
Communications 15
Shipping 16
On-Going Scheduling Considerations . . 17
Episodes Pilot Operations 18
Initiation of Episodes Sampling 18
Episodes Logistics 18
Episodes Sampling 19
-------
Summary of Results 22
Phase I 22
Upper Mid-Atlantic 22
Lower Mid-Atlantic 22
Screening 23
Southern Appalachians 23
Arkansas/Florida 23
Episodes Pilot 23
Upper Mid-Atlantic 23
Lower Mid-Atlantic 24
Observations and Recommendations 25
Scheduling 25
Site access 25
Stream Site Location 25
Stream Channel and Flow
Measurements 30
Shipping 31
Equipment 31
Safety 32
Episodes Pilot 32
Summary 33
References 34
Appendix A 35
Data Forms Used in the National Stream Survey 35
VI
-------
Figures
1. Regions and subregions sampled in the National Stream Survey, 1986 2
2. Summary of daily base site operations 11
3. Summary of sampling operations 12
4. Flow chart for episodes sampling 20
A-1. NSWS Form 4 35
A-2. NSWS Form 4A. 36
A-3. NSWS Form 6 37
A-4. NSWS Form 7 38
VII
-------
Tables
1. Summary of Base Site Movement and Number of Streams
Sampled: Upper Mid-Atlantic 6
2. Summary of Base Site Movement and Number of Streams
Sampled: Lower Mid-Atlantic 6
3. Summary of Base Site Movement and Number of Streams
Sampled: Southeast Screening 7
4. Summary of Base Site Movement and Number of Streams
Sampled: Arkansas/Florida 7
5. Physical and Chemical Parameters Measured in the National Stream Survey . . 8
6. Summary of Streams Visited 22
7. Incompletely Sampled Streams with Explanations 23
8. NSWS Episodes Pilot Summary 24
9. Summary of Problems, Solutions, and Recommendations for the NSS 26
10. Weather Predictions for the Lower Mid-Atlantic 32
VIII
-------
Acknowledgements
This study was conducted under the technical direction of Dr. J. J. Messer (U.S.
Environmental Protection Agency, Corvallis, Oregon). Logistical support for field oper-
ations was coordinated by R. E. Crowe (U.S. Environmental Protection Agency, Las Vegas,
Nevada, retired). Project management of field operations was performed by S. L. Pierett
and J. R. Baker (Lockheed Engineering and Sciences Company, Inc., Las Vegas, Nevada).
Recognition belongs to W. L. Kinney (U.S. Environmental Protection Agency, Las Vegas,
Nevada) who served as project officer for this survey.
Comments on this manuscript came from D. J. Chaloud and J. R. Baker, and J. M.
Nicholson (Lockheed Engineering and Sciences Company, Inc., Las Vegas, Nevada) served
as technical editor. Excellent reviews were received from D. Newbold (Stroud Water
Research Center, Avondale, Pennsylvania) and M. Bowman (State of Maryland, Department
of Environmental Quality, Baltimore, Maryland). A. H. Hall, P. F. Showers, and B. J. McRae
(Lockheed Engineering and Sciences Company, Inc., Las Vegas, Nevada) provided typing
and word processing support.
IX
-------
-------
Section 1
Introduction
The National Surface Water Survey
(NSWS) was initiated in 1983 by the U.S.
Environmental Protection Agency (EPA).
It was designed to provide a base of uni-
formly collected, processed, and analyzed
data on surface waters in the United States
potentially susceptible to change resulting
from acidic deposition. The program consists
of the National Lake Survey (NLS) and the
National Stream Survey (NSS) (Linthurst et
al., 1986). Phase I activities of the NSWS
provided information to determine the cur-
rent chemical status of lakes and streams.
These activities focus on areas where geo-
chemical data indicate a preponderance of
surface waters having low acid neutralizing
capacity (ANC). Phase II activities describe
seasonal variability in regional surface water
chemistry identified in the Phase I survey.
A pilot streams survey (NSS-PS) was
conducted in 1985 to develop sampling and
logistical protocols (Messer et al., 1986;
Knapp et al., 1987). The 1986 NSS was a
full-scale effort that included three distinct
parts: a Phase I survey of streams in the
mid-Atlantic region, a screening survey
designed to assess the need for future Phase
I studies in the southeastern United States,
and an episodes pilot survey conducted on
a subset of Phase I streams. The pilot
survey tested protocols for sampling during
rain events.
The research plans for the Phase I
and Episodes Pilot surveys were developed
by the EPA Environmental Research Labor-
atory in Corvallis, Oregon. Survey oper-
ations, which included sample collection,
processing, preparation and quality assur-
ance, were developed and completed by
Lockheed Engineering and Management
Services Company, Inc. (Lockheed-EMSCO),
under contract to the EPA Environmental
Monitoring Systems Laboratory in Las Vegas,
Nevada. The initial sample processing and
aliquot preparation was performed by the
Lockheed-EMSCO laboratory in Las Vegas.
Further chemical analyses of samples were
performed by several contract laboratories.
This report describes all field operations,
beginning with a brief description for each
of the three 1986 NSS surveys. Survey
results are summarized, problems encountered
during the survey are outlined, and solutions
to problems are recommended for future
work. The quality assurance plan is discussed
in Drous<§ et al. (1986). Processing laboratory
operations are discussed in L. J. Arent et
al. (in prep.) and Hillman et al. (1987). A
compilation of survey results will be available
through EPA-Corvallis (P. R. Kaufmann et
al., in prep.). A list of stream reaches
targeted for sampling, along with their
locations, will be given by P. R. Kaufmann
et al. (in prep.).
Phase I Survey
The 1986 NSS Phase I effort was
conducted primarily in the mid-Atlantic
region. It included the area bounded appro-
ximately by the Catskill and Pocono Moun-
tains to the north, the northern margin
of North Carolina to the south, the western
boundaries of Pennsylvania and West Virginia
to the west, and the Atlantic Ocean to the
east. This region was expected to contain
many areas of low ANC and was thought
to have relatively high levels of acidic
deposition.
Subregions targeted for sampling (see
Figure 1) included the Pocono and Catskill
Mountains (Region 1D) the Pine Barrens
and Chesapeake Bay (3B); the northern
portion of the Valley and Ridge Province
(2B) and the northern portion of the Ap-
palachian Plateau (2C). Results from Phase
I of the NSS will be used to determine the
percentage, extent, and location of streams
1
-------
£J PHASE I PILOT SURVEY
D MIDDLE ATLANTIC SURVEY AND
EPISODES PILOT SURVEY
VA SCREENING SURVEY
Figure 1. Regions and subregions sampled In the National Stream Survey, 1986.
that are presently acidic and of streams
that may be susceptible to acidification.
Each of the 276 stream reaches selected
for mid-Atlantic Phase I sampling was sched-
uled to be sampled twice during spring
baseflow conditions to quantify, to some
extent, the degree of temporal variance
within the spring sampling season.
Screening Area Survey
Several areas having lower acidic
deposition than the mid-Atlantic region were
selected for a screening survey designed
to assess the need for future Phase I
efforts.
The Screening survey area comprised
the southern Appalachian Mountains, includ-
ing parts of regions 2A, 2B, and 2C not
sampled in the Phase I or NSS-PS surveys;
the Piedmont (3A); the Ozark and Ouachita
Mountains (2D); and parts of the Florida
panhandle and peninsula (3C) (see Figure 1).
The statistical design of the Screening
survey allows regional characterization, just
as in the Phase I study area. Because each
Screening survey stream was sampled only
once, no temporal variance estimate is
possible. The single sample is not expected
to provide enough information to allow
thorough classification of the streams for
Phase II.
-------
Sampling in the Screening survey areas
was conducted concurrently with sampling
in Phase I areas. With the exception of
stream channel and flow measurements,
identical protocols were followed. For the
Phase I survey, stream discharge was mea-
sured; for Screening survey streams, dis-
charge was estimated (see Section 3).
Episodes Pilot Survey
Recent research on surface water
acidification has suggested that significant
changes in stream and lake chemistry can
occur during hydrologic events such as
snowmelt and rainfall (e.g., Eshleman and
Hemond, 1985; Shaffer and Galloway, 1982).
These changes can include decreases in pH
and alkalinity and increases in potentially
toxic aluminum species and may be sufficient
to cause harm to aquatic biota (e.g., Scho-
field and Trojnar, 1980; Gunn and Keller,
1984).
A pilot survey was conducted on a sub-
set of Phase I streams to provide a prelim-
inary assessment of the frequency, duration,
and causes of storm episodes in the mid-At-
lantic states. The pilot survey also evalu-
ated possible sampling designs and logistical
protocols. The Episodes Pilot survey used
Phase I sampling teams in the mid-Atlantic
region. Episodes sampling replaced normal
Phase I sampling during rain events.
Streams to be sampled for the Episodes
Pilot were selected according to ANC class
and watershed size. Those with high ANC
or large watersheds were excluded, because
streams with these characteristics are
unlikely to experience episodes. The model-
based sampling design required a similar
number of samples from each of four cells
in the design:
A limited number of streams were preselected
for possible episodes sampling from these
four combinations of ANC class and watershed
size.
Acid Neutralizing
Capacity
Watershed
Size
1. Low (< 50 /jeq/L)
2. Low
3. Moderate (50-200
4. Moderate
2
Small (< 5 mi
Moderate (5-15
Small
Moderate
-------
Section 2
Preparation for Field Operations
Overview
Experience gained from preceding
surveys, particularly the NSS-Pilot Survey
(NSS-PS) conducted in 1985 on a subset of
Phase I streams, provided the foundation
for planning and implementation of the 1986
NSS effort. Results of NSS-PS evaluations
of the research plan and of data quality
objectives and data management plans are
included in Messer et al. (1986). Details of
the NSS-PS field operations plan can be
found in Knapp et al., 1987.
Planning for field operations was
influenced primarily by the research plan
requirements for each of the component
surveys, the number of streams to be sam-
pled, the length and timing of the sampling
period, and the size of the survey area.
Survey Planning
The subregion boundaries were drawn
around areas expected, on the basis of water
quality data, to be predominantly below
an ANC of400 fieq/L (in Florida, below 200
/jeq/L). In the mid-Atlantic subregions,
streams in the lowest ANC strata were
chosen with a higher probability. Streams
were selected for the NSS data base (without
regard to accessibility), by using procedures
and criteria described in Messer et al. (1986)
and summarized below.
To accurately and confidently charac-
terize stream chemistry and associated
physiographic attributes, a statistically based
scheme was developed to ensure that the
streams sampled would be representative of
the target population (i.e., those streams of
interest, based on theprimary objectives of
the Aquatic Effects Research Program).
The selection and subsequent sampling of
streams during Phase I operations was
achieved by means of a three-step process.
The initial phase of the selection
process identified the potential target pop-
ulation of streams from which a statistical
sample could be drawn. Three regions of
the eastern United States, where surface
water acidification was most likely to occur
in the near future or where it had already
occurred, were identified using physiographic
boundaries and maps of surface water alk-
alinity. Each region was subdivided further
into subregions based on physiographic
similarity, vegetation, and land use patterns.
Delineation of subregions allowed for use
of a stratified sampling design to ensure
adequate spatial and physiographical repre-
sentation in the statistical sample.
Within the subregions, a statistical
sample of stream reacheswas selected using
U.S. Geological Survey (USGS) 1:250,000-scale
topographic maps and a point frame (grid
size = 64 scale mi2 or 165 km2). A stream
reach was defined as the length of stream
on the map between two tributary confluences
or between the headwater and the first
tributary confluence. This initial (or "first-
stage") sample served to estimate the total
target population in terms of the number,
length, and other geographic characteristics.
The first-stage sample was then screened,
using map characteristics, to eliminate reaches
that were not of interest (e.g., reservoirs,
urbanized areas, areas outside the subregion,
or areas with too large a drainage area).
A second probability sample was then
selected from the first-stage pool of stream
reaches. This "second-stage" sample repre-
sented those reaches that were scheduled
to be sampled during Phase I operations.
This second stage sample was a systematic
sample that was stratified based on sub-
region and the surface water alkalinity as
indicated by alkalinity maps. Additional
streams having historical water quality and
-------
hydrology data were included for sampling
as "special interest" streams.
The research plan allowed for a
2-month sampling period from mid-March
through mid-May 1986. The Phase I plan
included two visits to each of the 250
regular and 26 special-interest streams.
The first and second visits on a particular
stream had to be at least 2 weeks apart.
The 200 routine and 3 special-interest
Screening survey streams were visited one
time only.
Division of Study Area
The extensive geographic area and
the large number of streams to be sampled
for the NSS necessitated subdividing the
sampling regions so that survey schedules
could be met. The northern Phase I/Ep-
isodes Pilot region and the southern Screen-
ing region each were divided into two parts.
Each contained approximately the same
number of streams. The Phase I/Episodes
Pilot mid-Atlantic area was divided along
an approximate east-west line. The upper
mid-Atlantic (DMA) portion included the
states of Pennsylvania, New York, Delaware,
and a small part of Maryland. The lower
mid-Atlantic (LMA) portion included West
Virginia, Virginia, a large part of Maryland,
and parts of North Carolina and Pennsyl-
vania. The DMA area contained 127 routine
streams and 16 special-interest streams.
The LMA area contained 123 routine and
10 special-interest streams. At the request
of the investigating scientist, one special-in-
terest stream was added during the survey
when a nearby Phase I stream proved to
be dry. The Screening survey also was
divided into two regions: the Southern
Appalachian region and an area encompassing
parts of Florida and Arkansas. Each region
contained 100 regular streams, and they
contained 1 and 2 special-interest streams,
respectively.
Each of the four survey areas was
divided further into 8 to 15 "base sites."
Base sites, which were determined before
the survey began, served as temporary
headquarters for field operations. Base
site selection was based on availability of
services. These included: express courier,
alternate shipping carriers, motel accom-
modations, K- king, and proximity to major
roads. Base sites were located within stream
clusters. The suitability of sites and all
arrangements were confirmed during recon-
naissance trips prior to sampling.
Sample Site Inforrr^iion
U.S. EPA-Corvallis provided USGS
topographic maps marked to show stream
reaches to be sampled, 1:250,000-scale maps
from which the reaches were chosen, and
1:24,000-scale maps.
For each NSS stream, representatives
("local cooperators") from numerous agencies
were contacted for site access information.
The primary agencies contacted were the Soil
Conservation Service (SCS), state forestry
departments, the U.S. Forest Service, the
Tennessee Valley Authority, the National Park
Service, and state departments of fish and
wildlife. Over 220 individuals from the SCS
were contacted. The local cooperators were
asked to provide assistance by identifying
and contacting landowners, describing the
project, and obtaining access permission.
Additional information requested included
descriptions of travel routes with estimated
driving and hiking times, descriptions of
land use, and difficulty of access to stream
sites. All information, along with the
1:24,000-scale maps and other appropriate
county and regional maps, was included in
a packet assembled for each stream. These
packets were used by the sampling teams to
locate streams.
Sampling Sequence and Scheduling
A scheduling priority based on date of
spring leafout (phenology) was superimposed
on the overall sampling time frame. This
scheduling consideration was intended (1) to
provide sampling consistency among sub-
regions and (2) to minimize the influence of
external factors affecting stream chemistry
during the season of maximum plant growth.
-------
An exception to the phenological scheduling
occurred in Florida, where leafout was
almost completed by the time the survey
began.
To maintain consistency with phenologi-
cal requirements, sampling generally followed
a movement from coastal to inland areas,
from south to north, and from lower to
higher elevations. If a base site was central
to streams located in different leafout zones,
sampling was scheduled accordingly. The
scheduling of sample collection within each
survey area was planned with consideration
given to logistics and to the overall sample
collection goals as well as to phenology.
For each Phase I or Screening stream
visit, both an upstream and a downstream
sample were required. These were collected
on the same day by the same team. Each
of the special-interest streams was sampled
at a single location only, usually at the
site where stream data had been collected
by other agencies. The sequence in which
the upstream and downstream sites were
sampled was randomized and predetermined
for all NSS streams. During episodes,
streams were sampled at the downstream
sites only.
In the Phase I areas, the second set
of samples from a given stream had to be
collected before leafout, but at least 2 weeks
after the first visit. To accomplish this,
crews in each of the Phase I areas collected
the two sets of samples for a subset of
base sites before moving on to the next
subset (Tables 1 and 2).
Routine sampling in the Phase I areas
was conducted 5 days per week, Monday
through Friday. The Phase I effort shifted
to the Episodes Pilot survey whenever a
storm event with significant precipitation
was expected; this included weekends.
Approximately 30 sets of episode samples
were anticipated in the Phase I region
during the course of the survey. The 5-day-
per-week schedule in Phase I areas was
planned to accommodate this level of episode
sampling without interfering with the overall
sampling schedule.
Table 1. Summary of Base Site Movement and
Number of Streams
Sampled: Upper Mid-Atlantic
SITE DATE OF DATE OF
NO. BASE SITE ARRIVAL DEPARTURE
1 Milford, DE 17 MAR 19 MAR
2 Mount Holly, NJ 19 MAR 24 MAR
3 Rockaway, NJ 24 MAR 26 MAR
4 York, PA 26 MAR 29 MAR
1 Milford, DE 29 MAR 2 APR
2 Mount Laurel, NJ 2 APR 5 APR
3 Rockaway, NJ 5 APR 9 APR
4 York, PA 9 APR 12 APR
5 Altoona, PA 12 APR 17 APR
6 DuBois, PA 17 APR 23 APR
7 Williamsport, PA 23 APR 27 APR
8 Wilkes Barre, PA 27 APR 30 APR
9 Kingston, NY 30 APR 2 MAY
5 Altoona, PA 2 MAY 5 MAY
6 DuBois, PA 5 MAY 9 MAY
7 Williamsport, PA 9 MAY 11 MAY
8 Wilkes Barre, PA 11 MAY 14 MAY
9 Kingston, NY 14 MAY 16 MAY
NO. OF
STREAMS
11
16
16
18
9
16
12
21
14
15
20
20
17
10
23
9
23
17
Table 2. Summary of Base Site Movement and
Number of Streams
Sampled: Lower Mid-Atlantic
SITE DATE OF DATE OF
NO. BASE SITE ARRIVAL DEPARTURE
1 Williamsburg, VA 17 MAR 20 MAR
2 Fredericksburg, VA 20 MAR 25 MAR
3 Sheperdstown, WV 25 MAR 28 MAR
4 Charlottesville, VA 28 MAR 1 APR
1 Williamsburg, VA 1 APR 4 APR
2 Fredericksburg, VA 4 APR 8 APR
3 Sheperdstown, WV 8 APR 11 APR
4 Charlottesville, VA 11 APR 17 APR
5 Oak Hill, WV 17 APR 25 APR
6 Elkins, WV 25 APR 1 MAY
7 Oakland, MD 1 MAY 4 MAY
5 Oak Hill, WV 4 MAY 9 MAY
6 Elkins, WV 9 MAY 13 MAY
7 Oakland, MD 13 MAY 16 MAY
NO. OF
STREAMS
16
15
23
13
17
9
22
22
32
17
13
23
23
15
Screening area sampling was usually
conducted 6 days per week, Sunday through
Friday. Because the study plan specified
that samples be taken only under baseflow
conditions, no samples were collected during
or immediately following storm events.
Streams were allowed to return to base-
flow. Tables 3 and 4 summarize base site
movement and number of streams sampled
for the Screening area.
-------
Table 3. Summary of Base Site Movement and
Number of Streams
Sampled: Southeast Screening
SITE
NO. BASE SITE
DATE OF DATE OF NO. OF
ARRIVAL DEPARTURE STREAMS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Tupelo, MS
Tuscaloosa, AL
Montgomery, AL
Columbus, GA
Madison, GA
Greenville, SC
Bremen, GA
Cartersville, GA
Atlanta, GA
Fort Payne, AL
Crossville, TN
Wilkesboro, SC
Blowing Rock, NC
Salem, VA
17
19
23
24
28
2
7
10
11
21
25
4
5
8
MAR
MAR
MAR
MAR
MAR
APR
APR
APR
APR
APR
APR
MAY
MAY
MAY
19
23
24
28
2
7
10
11
21
25
4
5
8
13
MAR
MAR
MAR
MAR
APR
APR
APR
APR
APR
APR
MAY
MAY
MAY
MAY
4
5
2
10
6
6
5
3
13
10
16
2
5
11
Table 4. Summary of Base Site Movement and
Number of Streams
Sampled: Arkansas/Florida
SITE
NO. BASE SITE
DATE OF DATE OF NO. OF
ARRIVAL DEPARTURE STREAMS
1
2
3
4
5
6
7
8
9
Hugo, OK
DeQueen, AR
Hot Springs, AR
Con way, AR
Ozark, AR
Kissimmee, FL
Starke, FL
Marianna, FL
Niceville, FL
17 MAR
20 MAR
25 MAR
31 MAR
11 APR
21 APR
23 APR
28 APR
5 MAY
20 MAR
25 MAR
31 MAR
11 APR
21 APR
23 APR
28 APR
5 MAY
9 MAY
6
12
8
17
9
4
14
19
13
Laboratory Location
Field experiments conducted during
the NSS-PS (Messer et al., 1986) indicated
that samples from a diverse group of surface
waters could be held for several days under
appropriately controlled conditions without
undergoing major changes in water chem-
istry. The result of this finding was that
the mobile laboratories could be centralized
in Las Vegas rather than being located at
base sites in the field as they were in
previous NSWS surveys. This finding was
particularly advantageous for survey plan-
ning. The large geographic area covered by
the NSS made it unlikely that a single
location within each of the four base areas
would have been adequately central to
sampling sites for the entire survey period.
The laboratories would have had to be moved
several times. This would have been very
time-consuming in terms of the (I) initial
reconnaissance required to find suitable
locations with all the facilities necessary
to support the laboratories and (2) in terms
of the disruption that would have resulted
from moving during sampling operations.
Not all sampling sites would have been within
reasonable driving distance to the labora-
tories, and alternative shipping arrangements
would have been necessary on a frequent
basis.
Protocol Development
Laboratory Protocol
Most of the laboratory protocols devel-
oped during the NSWS Eastern Lake Survey
(ELS) were applicable to processing of NSS
samples. A complete description of the
laboratory methods used for the NSS can
be found in Hillman et al. (1987) and in L
J. Arent et al. (in prep).
Sampling Protocol
Protocols for sample collection and in
situ measurement techniques developed for
the ELS were unsuitable for streams. Con-
sequently, the NSS-PS goals focused on
identifying field equipment and testing
techniques specific to stream sample collec-
tion. All field techniques used during the
NSS were developed in the NSS-PS. Modi-
fications were related primarily to field pH
and hydrologic measurements. A comparison
of the open- and closed-atmosphere field
pH measurement techniques developed during
the NSS-PS indicated no significant differ-
ence between values obtained with either
method (Knapp et al., 1987). The open vessel
method required less equipment and was
more time-efficient; therefore, it was chosen
for the NSS. Physical and chemical para-
meters measured in the NSS (Table 5) were
similar to those measured in other NSWS
surveys. A detailed description of NSS
protocols is given in Section 3.
-------
Table 5. Physical and Chemical Parameters Measured In the National Stream Survey
FIELD
MEASUREMENTS
Oxygen
pH, Closed System8
pH, Open System
Conductivity
Stream Stage Height
Stream Velocity"
Temperature
ANALYTICAL
LABORATORY
MEASUREMENTS
Al, Organic Extractable3
Al, Total Extractable
Al, Total
Acid Neutralizing Capacity (ANC)
Base Neutralizing Capacity (BNC)
Ca
Cl
DIG, Equilibrated
DIC, Initial
uuo
F", Total Dissolved
Fe
i(+
UNITS
mg/L
pH units
pH units
/^S/cm
0.001 ft
m/sec
*C
UNITS
mg/L
mg/L
mg/L
^eq/L
peq/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
Phase I Only
PROCESSING
LABORATORY
MEASUREMENTS
Al, Total Monomeric
Al, Non-exchangeable Monomeric
True Color
DIC, Closed System
pH, Closed System
Conductivity
Turbidity
ANALYTICAL
LABORATORY
MEASUREMENTS
Mg2+
Mn
Na+
NH4+
NO,'
P, Total3
P, Total Dissolved5
pH, Equilibrated
pH, Initial ANC
pH, Initial BNC
Silica (SiO2)
SO42'
Conductivity
Total Non-filterable Solids3
UNITS
mg/L
mg/L
APHA units
mg/L
pH units
A/S/cm
NTU
UNITS
mg/L
mg/L
mg/L
mg/L
mg/L
ma /I
mg/L
pH units
pH units
pH units
mg/L
mg/L
pS/cm
mg/L
Guidelines for Sampling
The NSS sampling area emcompassed
diverse conditions not encountered in the
small geographic region sampled during the
NSS-PS. To some extent these conditions
were anticipated beforehand, and guidelines
were developed to assist samplers in judging
whether a questionable stream reach should
be sampled. Streams having water chemistry
dominated by some factor not germane to
the objectives of the NSWS were labeled
as "non-interest" streams and were not
sampled. Non-interest streams were charac-
terized by:
• pH less than 3.3 (e.g., streams polluted
by acidic mine drainage)
• conductivity greater than 500 /jS/cm (e.g.,
those contaminated by industrial pollution)
• coastal areas with water chemistry af-
fected by tidal influence (i.e., conductivity
greater than 250 ;uS/cm)
• 90 percent of the stream reach dry or
stagnant (e.g., beaver ponds and swamps)
• a large reservoir inundating the stream
reach.
Personnel
Staffing Requirements
Experience gained during the NSS-PS
indicated that one 2-person sampling team
could collect samples from an average of
seven streams per week. It was determined
that 14 sampling teams were needed to
maintain the survey schedule. Five teams
were required for each of the Phase I areas;
two teams for each of the Screening areas.
To accommodate the increased work load
of Episodes sampling, the Phase I teams
converted to three, 3-person teams during
rain events.
-------
Personnel Duties
The field crews in each study area
also included a base coordinator and a
logistics coordinator. The base coordin-
ator had responsibility for all field opera-
tions, including scheduling the sampling
itineraries, supervising the field crews,
maintaining communications with Las Vegas,
assuring timely shipment of samples, and
performing other administrative and super-
visory duties. All presampling field opera-
tions planning, including reconnaissance,
was conducted by the base coordinators.
The logistics coordinator provided
assistance to the base coordinator and
attended to the details associated with
moving, contacting local cooperators, obtain-
ing access permission, and maintaining supply
inventories. In addition, the logistics
coordinators were trained in all aspects of
sampling and were available as backup
samplers.
Personnel Selection and Training
Field personnel were selected on the
basis of knowledge of water sampling tech-
niques and on field experience. Preference
was given to those having experience from
previous NSWS surveys. New employees
were hired and equipment was issued in
Las Vegas. From February 24 to March
16, 1986, all field personnel were trained
in NSS project design and purpose and in
field safety procedures. Samplers drove
survey vehicles to Oak Ridge National
Laboratory (ORNL) in Oak Ridge, Tennessee.
Training at ORNL covered NSS logistics
and operations, instrumentation, stream
sample collection and measurement techni-
ques, and proper data recording. The final
week of training was held at Nantahala
Outdoor Center in Bryson City, North Caro-
lina, where orienteering, outdoor skills,
and safety were stressed. Trainees also
continued to practice sample collection and
stream measurement techniques. Lectures,
small group sessions, and streamside practice
were used to teach trainees these skills.
-------
Section 3
Field Operations
Overview
The Phase I and Screening surveys fol-
lowed different schedules and required
different sampling frequencies. However,
daily operations, stream sampling protocols!
shipping, and sample processing were essen-
tially identical. The Episodes Pilot had
different daily operations, but sample collec-
tion, streamside measurements, and sample
processing remained virtually identical to
Phase I work.
A routine sample for any of the three
portions of the survey consisted of one
Cubitainer and four syringes of streamwater.
Streamside measurements always included pH,
conductivity, temperature, and dissolved
oxygen. Duplicate and blank samples were
collected daily in each survey area as part
of the quality assurance program. Stream-
side measurements differed only for hydro-
logy. Hydrologic parameters were estimated
in Screening areas and were measured for
Phase I and Episodes Pilot areas.
Daily Base Site Operations Summary
Daily base site operations began before
dawn. Instruments were calibrated, equip-
ment and supplies were loaded, and itin-
eraries were finalized. After the field teams
departed from the base site, the base coor-
dinators worked on communications, data
forms, administrative paperwork, and future
planning. The logistics coordinators worked
on supply inventories, access permission
problems, and site logistics. Either the
logistics coordinator or the base coordinator
remained available to receive calls from
samplers. When samplers returned to the
base site, the logistics coordinator repacked
samples into shipping coolers and shipped
them to the Las Vegas laboratory, where
the syringe samples were analyzed and the
Cubitainer samples were divided into ali-
quots, were preserved, and then were shipped
to the contract laboratories for analysis.
The base coordinator conveyed the number
and the type of samples collected to the
Las Vegas Communications Center. A de-
briefing session outlining problems and
suggestions completed the daily operations.
These activities are summarized in Figure
2 and are discussed in detail later in this
section.
Sampling
Daily Sampling Operations
Examples of all NSS data forms are in-
cluded in Appendix A.
An overview of daily sampling operations
is given in Figure 3. Samplers calibrated
the pH and dissolved oxygen meters and
checked the calibration of the conductivity
meter. In Phase I areas, the calibration
of the flow meter was checked also. Follow-
ing morning calibration, the filing of itinerary
forms, and the loading of supplies, samplers
traveled to the first site in two- or four-
wheel drive vehicles.
Samplers used maps, compasses, and
landmarks to determine and mark on the
1:24,000-scale USGS topographic maps the
exact location at which streams were sampled.
If a hike was required to reach the site or
if the site was difficult to find, samplers
in Phase I areas marked the location with
flagging tape to aid in locating the site on
the second visit.
On the first visit, each site was de-
scribed on the Watershed Characteristics
Form (NSWS Form 7). Watershed distur-
bances, bank vegetative cover type and per-
cent, and stream substrate were described
on the form. The exact site location was
marked on the topographic map, and photo-
10
-------
SAMPLING TEAM
INSTRUMENT CALIBRATION
AND QUALITY CONTROL CHECKS
PACKING OF
EQUIPMENT AND SUPPLIES
[DEPARTURE FOR SITE |
SAMPLE COLLECTION AND
FIELD MEASUREMENTS
RETURN TO BASE SITE
POST-CALIBRATION AND
COMPLETION OF DATA FORMS
CLEAN-UP AND
PREPARATION FOR
NEXT DAY
LOGISTICS
COORDINATOR
BASE
COORDINATOR
I
ASSISTANCE WITH
CALIBRATION AND
PACKING
CALIBRATION
SET-UP
WEATHER CHECK
ACCESS PERMISSION, DATA FORM CHECKS,
CHARTING OF PROGRESS, FUTURE
SCHEDULING, INVENTORY OF SUPPLIES,
ADMINISTRATIVE PAPERWORK, BASE SITE
MOVEMENT, COMMUNICATIONS, MAP
PACKET PREPARATION
ASSISTANCE TO SAMPLERS
ASSISTANCE WITH
SAMPLE PACKING
CHECKING OF STREAM
DATA FORMS
WEATHER CHECK
SAMPLE SHIPMENT
DAILY DEBRIEFING AND
SCHEDULING FOR NEXT
DAY
DAILY CALL TO
COMMUNICATIONS
CENTER
Figure 2. Summary of dally base site operations.
graphs of the site were taken. One sampler
calibrated (if needed) or checked calibrations
of instruments, took readings for pH, con-
ductivity, and dissolved oxygen, and recorded
data on the field logbook form. The second
sampler collected the routine water sample,
a duplicate or blank if such was required,
and samples of stream water for pH meas-
urements. The second sampler also took
measurements of channel dimensions, stage
height, and water velocity. Samplers alter-
nated duties within each team as desired,
but team composition remained consistent
throughout the survey.
Additional streamside data included
date, time, elevation of the site, cloud cover,
quality control check solution (QCCS) results,
team identification, instrument problems,
and any conditions which could affect water
quality.
Upon completion of sampling activities,
samplers packed equipment and samples,
checked data forms, and continued to the
next site. Samplers called base coordinators
while traveling between sites whenever
possible. Information from field logbook
forms was transferred to the four-part stream
11
-------
CALIBRATE,
DO QC CHECKS
_L
LOAD VEHICLES,
TRAVEL TO SITE
DESCRIBE SITE
1
_L
INSTALL STEEL ROD,18
READ STAGE B
Z
CONDUCT HYDROLOGY
MEASUREMENTS 2B
SAMPLER 1
Z
f
SET UP INSTRUMENTS,
DO ON-SITE CALIBRATIONS
AND CHECKS, PURGE PUMP
•
SAMPLERS
MEASURE pH
COLLECT BLANK
SAMPLE (IF NECESSARY)
Z
MEASURE CONDUCTIVITY
MEASURE DISSOLVED
OXYGEN
COLLECT ROUTINE SAMPLE
I
COLLECT DUPLICATE
SAMPLE (IF NECESSARY)
t
| PACK UP SAMPLES
READ STAGE, B
REMOVE STEEL ROD 2B
NO
1= FIRST SITE VISIT
2= SECOND SITE VISIT
A a UPSTREAM SITE
B= DOWNSTREAM SITE
CALL BASE COORDINATOR
RETURN TO BASE SITE
t
DO FINAL QC CHECKS,
PACK AND SHIP SAMPLES,
PREPARE FOR NEXT DAY
Figure 3. Summary of sampling operations.
12
-------
data forms (NSWS Form 4) after samplers
returned to the base sites. The oxygen
meter and any other instruments that did
not meet quality control criteria in the
field were rechecked, and final oxygen
calibration check data were recorded on
stream data forms.
Sampling Methods
Sample Collection--
Samples were collected by using the
techniques developed in the NSS-Pilot study,
which are described in detail in Knapp et
al., 1987. Routine water samples were
collected by pumping water through Tygon
tubing held in the center of the stream
cross-section by using a sampling arm.
The water was pumped by portable, battery-
-driven peristaltic pumps. Each sample
collected represented a time-integrated
sample of the stream flow. A bulk sample
was collected in a 4-L polyethylene Cubi-
tainer. Four 60-mL polypropylene syringes
equipped with gastight valves were filled
so that the samples were not exposed to
the atmosphere. These syringes were used
for dissolved inorganic carbon (DIG), pH,
PCV aluminum, and MIBK-extractable alum-
inum analyses. Each container was rinsed
three times with sample water prior to
filling, and new Tygon tubing was used at
each site. Collected samples were trans-
ferred immediately to coolers lined with
frozen gel packs.
Phase I Channel and Flow
Measurements--
All hydrologic information was recorded
on the hydrologic data form (NSWS Form
4 A).
On the first visit to each Phase I
downstream site, a steel rod (i.e., nonrecord-
ing staff gauge) was hammered into the
streambed at a protected location out of
the main flow. Whenever possible, the
height of the top of the steel rod, used as
the reference point, was represented as 3.0
feet. Stream stage measurements were made
relative to this value. If a gauging station
was already present at the site (e.g., at
many special interest streams), gauge mea-
surements were made at this location. Stage
height was measured immediately after gauge
installation, after samples were collected,
and just before departure from the stream.
The steel rod was left in place until the
stream was revisited.
On the second visit to the downstream
site, stage height was recorded in the same
manner. In addition, stream width, depth,
and flow were determined. Stream flow
was measured with a Marsh-McBirney Model
201-D electromagnetic current meter with
the probe mounted to a wading rod. Meter
calibration was checked daily during the
routine morning calibration and was rechecked
on site before the sampler entered the
stream. Once a week, the zero value was
checked in static water. The probe was
allowed to sit undisturbed for 30 minutes,
and the meter zero was adjusted if the
value was outside the allowed range.
To measure stream discharge at each
downstream Phase I site, a USGS procedure
(Carter and Davidian, 1968; Buchanan and
Somers, 1969) was adapted. The procedure
follows:
(1) Beginning at the right edge of
the water (REW) and facing downstream,
a tape measure was stretched taut
across the channel perpendicular to
the stream flow and approximately
20 cm above the surface of the water.
The reach was chosen so as to have
an approximately U-shaped, straight
channel with minimal eddies or tur-
bulence.
(2) The stream width was measured
and divided into 8 to 15 equidistant
intervals. This was done by dividing
the total stream width by an integer
value near ten and then by rounding
down to a convenient number. An
additional interval was added if this
procedure resulted in one end of the
13
-------
channel width having a section greater
than one interval width.
(3) After a calibration check was
performed, stream depth at the center
of the first interval from the REW
was measured with the marked wading
rod on the flow meter.
(4) The current meter probe was
placed at 60 percent of the depth
measured down from the surface or
at 40 percent of the depth measured
up from the bottom. The meter was
allowed to equilibrate for 20 seconds,
then the current velocity was recorded.
This procedure was repeated for all
intervals.
Screening Area Channel and Flow
MeasurementS"
In Screening areas, or in Phase I areas
if the stream was too dangerous to enter,
channel depth, width, and velocity were
estimated. In Screening areas, as in Phase
I areas, these measurements were taken at
downstream sites. Stream width was mea-
sured with a meter tape or was estimated
by standing at the stream edge and by
sighting down the length of one arm toward
the other shore. While maintaining the
extended arm at the same angle, the mea-
surer then pivoted to point at an object or
location at the same elevation as the water
surface. The distance from the measurer's
feet to the indicated point was recorded
as the estimated stream width. This method
was tested and was found to be repeatable
among samplers and to be reasonably ac-
curate.
The mean depth of the entire channel
area over which velocity estimates were
to be made was estimated. Current velocity
was estimated by measuring the time re-
quired for an object (usually an orange) to
float down a known length of the stream
channel. The average velocity of three
trials was recorded as meters per second.
Streamside Measurements-
pH--AII field pH determinations in
the NSS were made with Beckman pHI-21
portable meters with Orion-Ross Model 8104
combination electrodes and automatic ther-
mocompensator sensors.
The pH protocol was almost identical
to the open method which was developed
and refined in the NSS-PS, except that
following a 3-minute prereading equilibration
swirl in a 150-mL sample aliquot, the elec-
trode was allowed to equilibrate further,
unswirled, for 2 minutes before a reading
was taken on a new sample aliquot. Replicate
readings were taken on new sample aliquots
until two successive readings agreed within
0.03 pH unit. The final pH and temperature
were the values reported.
Conducf/vfty--Conductivity was measured
in situ with YSI Model 33 SCT meters.
Measurements were made with the probe
immersed in the stream at mid-depth, and
conductivity and water temperature were
recorded when the conductivity reading
changed less than 5 juS/cm over a 1-minute
period.
Dissolved Oxygen-Dissolved oxygen was
measured with a YSI Model 54A dissolved
oxygen meter combined with a YSI Model
5739 pressure-compensating oxygen-tempera-
ture probe. The stream dissolved oxygen
concentration was measured by immersing
the probe in flowing stream water at mid-
depth. Dissolved oxygen and water tempera-
ture were recorded when the oxygen reading
changed less than 0.5 mg/L over a 1-minute
period.
Quality Assurance of Field Operations
Quality Assurance Samples-
Two types of quality assurance samples,
duplicates and blanks, were collected at
streamside daily. Two of the 14 teams
collected a "field blank" sample at the first
site visited each day. Reagent-grade deion-
ized water (ASTM Type I) was prepared
14
-------
with a Millipore Milli-RO/Super Q System
in the Las Vegas processing laboratory and
was shipped to the field three times per
week This water was carried to streamside
h 4 L CubitaTners and was pumped through
he mSne apparatus into clean sample
containeT by "using standard technique.
DIG and pH analyses were not performed
CubaTne? and °Lr syringes) was fi..ed
with stream water from the pump immediate-
Iv after the routine sample was collected
by using identical techniques. A second
set of pH readings was obtained by using
anazed by the laboratory for
e-batch of sarnies pressed.
Standard Preparation-
Standards for instrument calibration
and calibration checks were prepared three
times weeklv in the Las Vegas laboratory
anTweTshipped at 4 'C to'the field. This"
frequency of "shipment was p.anned to guar-
antee that standards would not be more than
against a National Bureau of Standards JNBS)
thermometer each ™7* "^^J
did not agree within 0.5 C, the thermo
compensator was replaced.
Conducf,V/fy-Three solut,ons were
used to check the factory cal.brat.on of
the conductivity meters [ ^ ^^
sampling: ™ A/S/cm (5 x 10 N KO), ,w
before and Rafter each strearr fading with
the 74 ^S/cm solution as a QCCS^ FaNu^e
to meet acceptable values (64 to 84 /i5/cmj
for these checks requ.red clean.ng or replace-
ment of the probe.
Oisso,^ Oxyoe.-The dissolved oxygen
meters were calibrated at the base site each
J£££L? S* Te^ «
were checked aga^st .an NBS^ thermometer
calibrated at each site by using the theoret,-
cal partial pressure o_ oxygen at amb«rrt
temperature and elevation. The disso vea
oxygen probe was inserted into a water-fght
chamber C^ai^
was .mmersed ,n the
A rigorous calibration and calibration
check protocol for all field instruments
was followed on each sampling day. Equip-
ment maintenance was conducted on a week-
ly basis
pH-Field pH meters were calibrated
at thebase site each morning by using
commercially available, high-ionic-strength
buffer solutions (pH 7.00 and pH 4.01). A
nH 4 (1 x 10'4 N HoSOJ QCCS, prepared
by the laboratory in Las Vegas, was used
to check the calibration of the meter after
morning calibration and at each stream site.
The meter was recalibrated if it failed to
read between pH 3.90 and 4.10. The auto-
matic thermocompensator was checked
Acceptable values of these checks
were within ± 0.5 mg O2/L of the calibration
value.
LOQISTICS
Communications
»/«„-«
The commun.cat.ons center in i Las Vegas
monitored all field samphng ^^«» "^
ing sample sh.pment, number of streams
sampled, weather, sampling projections, supply
requests, and miscellaneous problems^ It
served as a point of contact for all technical
and logistical quesfons, provide d a backup
contact for samplers when base and logistics
coordinators were not available coordinated
the assignment of duplicate and blank samp-
15
-------
3Cted 3S a liaison and Administrative matters. Base coordina-
* the us '
The base coordinator was the main
contact with the communications center.
Two calls per day were required from each
base site. In the morning call, information
was provided on team itineraries, number
nH,
es, and any problems. In the evening,
-
.dentification
v
weather and sampling schedules. Relevant
information was passed on by the commun-
drectna
directing
t *
,11 ™ J t!amS' By
all commun,cat,ons through the
prevnted
was
A tinht r,Q+ L- 4 •
A tight network of communication
s anr"6 ^ ^ '°9iStiCS C°°r-
dinators and the sampling teams. Before
fiibri dnntartUre v the m°;ning> the Samp'erS
died ou an mnerary form that detailed
and olanL ^ '• ^"^ ^ ^^
and planned return times. If samplers did
not cal, m by the shipping deadline, search
and rescue operations were initiated. Samp-
n^-Nr-fT^ t0,Cal1 in 3S S°°n as
trave channpH pS3hmP '^ " r°Ute °f
travel changed. Each sampling team carried
rovlSn th emer?e,ncy telePh°ne numbers
covering the whole samphng region by
y'
D ..
Base coordinators often communicated
? tO exchan9e ideas and
, to arrange to share supplies,
and to ensure that the proper number of
duplicates and blanks were collected each
day. Conference calls were held weekly
for all base coordinators, supervisory per-
sonnel, members of the EPA management
team, and representatives from the Las
Vegas communications center, laboratory,
and quality assurance group. These calls
covered protocol changes, sampling problems,
forms.
Shipping
Sample Shipment--
SamP'es were shiPPed °n the day of
collection to ensure arrival at the Las Vegas
labo^ory within 24 hours. Overnight courfer
service was used in most cases. In some
rem°te areaS' ShJPment deadlines we£™
the early afternoon, and samplers either
began their day earlie or shipped the
fr°m 3 IOCati°" Close to ?hffstm
Rem°te Shippin9 Was arra"9ed j" advance
by the base co°rdinator when possible, but
all samplers carried shipping forms and a
direc<°rV °f '".pplng locations In som^
cases, alternatives to the courier service
were necessary, and counter-to-counter
service on major airlines was used. Accounts
were established in advance with several
airNneS Servin9 the Campling area"
Immediately after collection, samples
were placed in portable, soft coolers whh
reMem Packs^nd were transported to
the vehicles. Samples were then
into insu.ated coolers for snpping
data forms were placed in Ziploc bags
t0p of the samP|e^- SyringesPwere place
in P'aStJC ""^ners to" prevent breakage.
The number of coolant packs enclosed with
the samples was chosen to maintain sample
temperatures near 4 'C. Samples and forms
f°r a particular stream were contained in
tne same coolers. The identification numbers
of the samples for each stream sampled
that daY- including duplicates and blanks,
were recorded and reported to the Las Vegas
communications center
Supply Shipment-
Supplies for field operations were
shipped three times weekly from the Las
Vegas laboratory and warehouse. Standard
shipments were arranged before the survey
began. They included routinely used items
16
-------
such as shipping coolers, calibration stan-
dards, refrigerant packs, valves and cases
for syringes, and deionized water. Special
orders of nonroutine items or changes in
routine orders were reported to the com-
munications center. A schedule and list
of addresses for base sites was prepared
by each base coordinator before the survey
began. These schedules were modified
throughout the survey. Warehouse personnel
called the communications center to obtain
up-to-date addresses for each base coor-
dinator in order to prevent misrouted ship-
ments.
The supplies needed to conduct the
survey filled all available vehicle space
during moves between sites. Careful coor-
dination of lodging and careful shipping
arrangements were made to prevent large
shipments from arriving at one base area
prior to a move.
Form Completion and Transfer--
The Stream Data Form (NSWS Form
4) documented sample identification and
ensured proper sample processing by labor-
atory personnel. The base coordinator
checked the forms, signed them, and re-
moved the back copy from each four-part
form before enclosing them with the samples.
Upon sample arrival in Las Vegas, the
laboratory coordinator recorded batch infor-
mation and cooler temperatures on the forms
and then removed the third copy from each.
The remaining two copies were forwarded
to the quality assurance division.
Other data forms (i.e., the Watershed
Characteristics [NSWS Form 7], Hydrologic
Data [NSWS Form 4A], and Episodic Data
[NSWS Form 6]) were retained until the
base coordinator had checked them thorou-
ghly. The base coordinator signed them and
then removed the back copy of these three-
part forms. They were shipped to the Las
Vegas communications center by overnight
courier service along with other adminis-
trative paperwork and exposed film. Forms
were transferred to the quality assurance
division by the communications center twice
a week.
The base coordinator compared the
retained copy of the Stream Data Form
(NSWS Form 4) with its associated field
logbook form to look for errors or omissions.
Inaccuracies were reported to the QA repre-
sentative who corrected the original form.
Copies of the original were made and were
returned to the base coordinator to check
against the original change. After the QA
division had completed the form check, the
original of each was sent to ORNL for entry
into the NSS data base.
On-Going Scheduling Considerations
Despite tentative itineraries for sampling
prepared before the survey had begun and
despite a rigid time frame in which to sample
all sites, scheduling required flexibility.
Weather provided a source of uncertainty
for Phase I and Screening scheduling. The
base or logistics coordinator called a local
weather service (National Weather Service,
airport weather stations, radio stations)
morning and evening to obtain a forecast.
In Phase I areas, rain forced a rapid shift
to episode sampling which disrupted Phase
I scheduling. In the Screening areas, samp-
ling was stopped until rainfall ended and
until streams returned to baseflow.
The size and dimension of each base
area were chosen to include the majority
of sites within easy driving distance. How-
ever, some outliers required additional driving
or hiking time and special shipping arrange-
ments. In some cases, one or two teams
stayed overnight at remote locations in
order to reach outlying sites. The logis-
tics coordinator accompanied teams in these
cases.
Estimating the time necessary to sample
each stream was difficult. Often the time
necessary for particular sites could not be
estimated until the sites had been visited
once. On the first visit to each site, addi-
tional time was allotted for finding the
site. Scheduling was adjusted for sites of
varying difficulty and distance. The 2-week
17
-------
minimum interval required between the first
and second visits to each sampling site in
Phase I areas also influenced the scheduling
and sequencing of streams to be sampled.
Favorable weather conditions and familiarity
with the sites caused the second sampling
cycle to progress faster than the first, and
this created a potential for reducing the
time between first and second visits to a
site below the 14-day minimum time require-
ment.
To minimize the time required for
moving between base sites, samples were
collected during the move in most cases.
On the morning of the move, base and
logistics coordinators helped samplers pack
equipment and belongings, calibrate instru-
ments, and depart. Sampling teams returned
from the field in time to ship samples from
the new base site. The base and logistics
coordinators packed the calibration equip-
ment and supplies, all spare sampling sup-
plies, and any remaining personal gear.
The logistics coordinator drove to the new
site, and the base coordinator remained at
the old site to provide a communications
link for the samplers. Once the logistics
coordinator arrived, the base coordinator
moved to the new base site. In cases where
both coordinators had to drive at the same
time, communications responsibilities were
assumed by the Las Vegas communications
center.
Episodes Pilot Operations
Initiation of Episodes Sampling
Base coordinators chose streams to
be sampled during episodes from a list of
potential episodes streams. No stream could
be sampled for episodes more than once.
If it was impossible to reach any of the
target streams in an area in time to catch
an approaching front, additional target
streams were selected based on available
ANC and watershed-size information.
The forecast of an approaching storm
front caused a switch to episodes sampling.
Whenever possible, the predetermined and
randomized sequence for sampling upstream
versus downstream sites was maintained
for streams selected for the Episodes Pilot,
so that base stage samples could be used
for the Phase I and the Episodes Pilot
surveys. In some cases, rapidly approaching
fronts did not allow time to sample both
nodes, following the preassigned sequence,
before rain began to fall. In these cases,
the sampling team went directly to the
downstream site and collected the base flow
sample for the Episodes Pilot study before
the storm began. If the proper sequence
could not be followed, protocol dictated
that the stream be rescheduled for Phase I
sampling.
All possible Phase I samples were
collected before it became necessary to
switch to episode sampling. If a storm
was forecast to arrive late in the day, base
coordinators sent teams out early in the
day to do routine Phase I sampling at streams
suitable for subsequent episode sampling.
If the weather system developed into a
suitable storm, teams reorganized into episode
teams, returned to these streams for the
episode, and collected an additional baseflow
sample if time allowed. Teams were never
sent to a new, unfamiliar stream for episode
sampling late in the day or evening. If
the storm was forecast to arrive early or
mid-day, no attempt was made to collect
Phase I samples. Samplers organized into
episode teams and proceeded directly to
streamside to collect baseflow episode
samples.
Episodes Logistics
Personnel Duties-
During episodes, the five two-person
Phase I sampling teams regrouped into three
episode teams. Increased sampling activities
during an episode indicated the need for
three persons per team. These groups of
three remained at the episode sampling sites
for the duration of the event in most cases.
The remaining sampler, the logistics coor-
dinator, and the base coordinator acted as
"runners" (contact persons). The runners
18
-------
joined the sampling teams on the upstream
site if necessary, carried supplies to the
episode site, established the base camp,
and collected the base stage sample. The
runners carried these samples back to the
base site or to a shipping location. Detailed
site location descriptions were shared bet-
ween the runners so that each person was
familiar with the location of all three epi-
sode teams. Once the initial sampling period
was over, the base coordinator remained at
the base site to coordinate activities and
to provide a safety network. The logistics
coordinator and remaining runners were
available to return to the episode sites and
to pick up more samples, to deliver supplies,
or to replace fatigued samplers.
Communications-
Prior to departure for each episode,
the base coordinator relayed the sample
schedule, stream identifications, and logistics
plan to the Las Vegas communications cen-
ter. The communications center remained
on alert during the episode sampling period
and took over communications completely
during the time the base coordinator was
in the field. The base coordinator regained
communications responsibility after returning
from the field; the communications center
provided a backup. Sampling teams arranged
call-in schedules in advance or through
their runners. Protocol required that the
runner check on teams if call-in times were
not met. If a problem had arisen or if
teams could not be located, the communica-
tions center and appropriate authorities
would have been contacted. No such prob-
lems arose during the survey. All samples
were required to be shipped to arrive at
the Las Vegas laboratory and be divided
into aliquots within a 24-hour period.
Episodes Sampling
Routine Measurements-
Figure 4 gives a detailed flow chart
of sampling procedures. On arrival at the
episode site and once for every 30 minutes
during the rising stage of the episode, stage
height, cumulative rainfall, pH, temperature,
conductivity, and dissolved oxygen were
measured. Techniques for these measure-
ments were almost identical to those used
for Phase I sampling.
A minor difference in pH technique
was instituted during the survey. Rapid
pH fluctuations occur during the rising stage
of episodes. It was very difficult to obtain
readings for subsequent pH aliquots (fresh
beakers filled with stream water) that fall
within 0.03 pH unit of one another, the
required range for a "stable" pH reading.
To correct this problem, a full Cubitainer
of stream water was collected for each
30-minute pH reading. All aliquots for that
reading were taken from that Cubitainer.
A change in episodes protocol was
initiated during the course of the survey.
During intermittent rainstorms when samplers
remained on site in expectation of renewed
rainfall, the frequency of pH, conductivity,
dissolved oxygen, and stage measurements
was reduced to 2-hour intervals.
Sample Collection and Flow
Measurement-
During the ideal event, stream water
samples were collected at base stage, rising
stage, peak stage, and falling stage. The
base stage sample was collected immediately
after arrival and before rainfall had begun.
The rising stage sample was taken when
the pH had fallen to its lowest level (a
decrease of at least 0.3 unit) below the
base stage pH. The peak stage sample was
taken after increases in stage height, but
not until the reading on the staff gauge
did not rise between two successive monitor-
ing intervals. The falling stage sample was
taken when the stream dropped to one-third
of its total peak stage rise. A rising stage
sample was not collected if pH depression
was not observed. An irregular storm of
long duration could require that several
samples be collected before the greatest
pH depression or before the highest stage
height occurred. Only one sample was
shipped and analyzed for each sampling
19
-------
WAIT FOR
RAIN OR
STAGE RISE
NO
ARRIVE AT SITE, CALIBRATE
COLLECT BASE FLOW SAMPLE
STAG
RISING OR
PRECIPITATION
TARTE
PEAK
SAMPLE
COLLECTED.
7
NO
,YES
CONDUCT QCC,
COLLECT PEAK
STAGE SAMPLE
MONIT
INTER
30 Mir
ORING
1UTES
MONITOR
STREAM
PH
DROP>0.
SINCE LAST
SAMPLE
HAS
STAGE FALLEN
BY 1/3
pH
ROP>0.3
SINCE LAST
SAMPLE
CONDUCT QCC, COLLECT
RISING STAGE SAMPLE
COLLECT FALLING
STAGE SAMPLES
PACK EQUIPMENT,
RECORD ALL DATA,
RETURN TO BASE
>2
SING AND
PEAK STAGE
SAMPLES 0
HAND
DISCARD PREVIOUS RISING
STAGE SAMPLE WITH
HIGHER pH
Figure 4. Flow chart for episodes sampling.
20
-------
interval. Stream flow was measured four
times during the event, as close to the time
of collection of the four water samples as
possible.
Blank and duplicate samples were
assigned to events, when appropriate, as
part of the regular quality assurance pro-
gram. Blank samples were collected during
any of the four event sampling periods,
but duplicate samples were collected only
during the base stage or falling stage sam-
pling periods. It can take up to five min-
utes to collect one sample. If rapid changes
in chemical and physical conditions occurred
during the rising and peak stages, sequential
duplicate samples might not be true dupli-
cates, but might have separate chemical
and physical characteristics. Quality control
checks were made for each instrument each
time an event sample was collected or at
3-hour intervals if no samples were being
collected.
Return to Routine Sampling-
Phase I sampling could be resumed as
early as 12 hours after the end of an event
of short duration (8 hours or less), or at
least 24 hours after a long duration event
(more than 8 hours). When samplers
returned to Phase I sampling, they checked
each stream for high turbidity or flow or
other signs of continued impact from the
event. If the stream remained impacted,
it was rescheduled for a later date.
21
-------
Section 4
Summary of Results
Results and discussions of field
activities are given below. Laboratory
operations are summarized in L. J. Arent
et al. (in prep.). Quality assurance opera-
tions and data are summarized in K. A.
Cougan et al. (in prep.).
Few unpredicted difficulties were
encountered during the survey, and sampling
was completed on schedule. Less than 1
percent of the samples arrived at the pro-
cessing laboratory and were divided into
aliquots past the 24 hour time limit. The
fact that 1986 was an unusually dry year
had several effects on sampling in the Phase
I/Episodes and the Screening survey areas.
In the more southerly regions, streams that
might normally be flowing during the spring
were completely dry or stagnant. A number
of streams were sampled at only one node
because more than 90 percent of the reach
could not be sampled. Although a total of
30 sets of episodes samples had been
anticipated, the dry weather allowed only
2 complete sets and 7 partial sets of samp-
les to be collected. Tables 6 and 7 sum-
marize the Phase I and Screening survey
results.
Table 6. Summary of Streams Visited
Phase I
Upper Mid-Atlantic
Of a total of 143 Phase I and special-
interest streams, 140 were sampled at up-
stream and downstream sites on each of
two visits. Two unsampled streams were
in tidal marshlands; the remaining stream
had a conductivity greater than 500
The second visits to each site occurred
between 9 and 21 days after the first visits.
When they were sampled for a second time,
11 streams were sampled earlier than the
recommended 14 days after the first visit.
Lower Mid-Atlantic
Of a total of 133 Phase I and special-
interest streams, 127 were sampled. Six
streams were not sampled; three were dry,
and three had conductivities greater than
500 pS/cm. Five streams were sampled
partially: two because of access permission
problems at the upper or lower site, two
because the lower site was tidal, and one
because of high conductivity at the lower
REGION
UPPER
MID-ATLANTIC
LOWER
MID-ATLANTIC
SOUTHEASTERN
SCREENING
ARKANSAS/
FLORIDA
TOTAL
STREAMS3
143
133
101
102
TOTAL
SPECIAL-INTEREST
STREAMS
16
10
1
2
TOTAL STREAMS
SAMPLED"
140
127
89
91
NUMBER OF STREAMS
PARTIALLY SAMPLED0
0
5
0
13
NUMBER OF
STREAMS
NOT SAMPLED
3
6
12
11
* Includes special-interest streams.
Includes partially sampled streams.
0 Missing upper or lower sites on one or both visits
22
-------
Table 7. Incompletely Sampled Streams wltti Explanations
EXPLANATIONS
REGION
CONDUCTIVITY
TIDAL > 500 /jS/cm DRY
NO FLOWING
NO ACCESS WATER
INACCESSIBLE
UPPER
MID-ATLANTIC
LOWER
MID-ATLANTIC
Streams
not sampled 2 1
Streams - "
partially
sampled
Streams -33-
not sampled
Streams 21-2
partially
sampled
SOUTHEASTERN Streams
SCREENING not sampled
Streams -
partially
sampled
ARKANSAS/
FLORIDA
Streams - - 9 1 1
not sampled
Streams - - 12 1
partially
sampled
site. One stream reach was in a coastal
area, but samplers did not realize that it
was tidal, because the conductivity was
less than 250 /jS/cm at the time of their
first visit. On the second visit, the tide
was in, and the conductivity was far over
500 juS/cm.
The second stream visits occurred 8
to 21 days after the first visits. When they
were sampled for a second time, 34 streams
were sampled earlier than the recommended
14 days after the first visit, but only 4 of
these visits were within a timespan of less
than 12 days.
Screening
Southern Appalachians
Of the 101 Screening and special inter-
est streams, 89 were sampled. Nine streams
were dry, one was inaccessible because of
hazardous conditions, and one was inundated
by a major water project that did not appear
on the topographic maps from which the
streams were chosen. Access permission
could not be gained for one stream.
Arkansas/Florida
Of a total of 102 Screening and special-
interest streams, 91 were sampled. Of the
11 streams not sampled, 9 were dry, 1 was
stagnant, and 1 had no access permission.
Thirteen routine streams were sampled at
one location only. Twelve of these streams
were dry or too shallow for over 90 percent
of their length, and access permission had
been denied at the lower site of the other
stream.
Episodes Pilot
Upper Mid-Atlantic
Only three defined rain events (precipi-
tation > 0.20 inches at the base site within
a 24-hour period) occurred in the region
during the sampling period. Two streams
23
-------
were sampled during two events. The third
rainstorm was missed because of indefinite
weather forecasts and sampler fatique
(Table 8).
Table 8. NSWS Episodes Pilot Summary
STREAM NUMBER
AND NAME
TYPE OF NO. OF
DATE SAMPLE SAMPLES
Lower Mid-Atlantic
3B059034*
Turner Run, VA
3B059038
Courthouse Creek, VA
3B048101
Marshall Creek, VA
2B058017*
Muddy Bridge Creek, PA
20 MAR 86 B1, R 2
20 MAR 86 B1, R 2
06 APR 86 B1, P, F 3
16 APR 86 B1, R, P, F 4
-,1
2B058024* 16 APR 86 B1 P F 3
Little Irish Creek, PA
2C046018* 21 APR 86 B
Blue Knob Creek, WV
1
2C046050* 21 APR 86 B, R P F 4
Hedricks Creek, WV
1D030093
No Name, NJ
2C028043*
No Name, PA
Upper Mid-Atlantic
07 APR 86 B, R, P
21 APR 86 B, P
seven streams were sampled during four of
these events (see Table 8). Sampling was
precluded for two of the remaining three
precipitation events because of severe and
localized thunderstorms which caused extr-
emely localized rainfall and unsafe condi-
tions. One rain event was missed entirely
because of logistical constraints that arose
when samplers were moving to a new base
site. Some watersheds did not receive enough
rainfall for observable changes in water
chemistry and hydrology to occur.
B = Base stage
R = Rising stage
P = Peak stage
F = Falling stage
* = Target episode stream
B = Baseflow measured as a Phase I sample on previous
day or earlier on same day
Only one of three stream sites sampled
for each event received enough precipitation
to result in observable changes in water
chemistry and hydrology. For this reason,
only two sets of episode samples for the
period between March 17 and May 15, 1986,
were sent to the Las Vegas processing
laboratory.
Lower Mid-Atlantic
In the lower mid-Atlantic region, seven
defined rain events occurred. A total of
24
-------
Section 5
Observations and Recommendations
Table 9 summarizes problems encoun-
tered during the survey, solutions imple-
mented in the field, and recommendations
for future work. In the following section,
observations associated with all operations
of the NSS are discussed and summarized.
Scheduling
The upper mid-Atlantic, lower mid-
Atlantic, and Arkansas samples were col-
lected before full spring leafout occurred.
In the Southern Appalachian Screening area,
leafout was accelerated by the unusually
warm, dry conditions, and it was not pos-
sible to collect all samples before substantial
leafout had occurred.
Many of the areas in which sampling
took place were heavily populated or were
popular vacation locations. Accommodations
were often difficult to find. The constantly
changing sampling schedule made it difficult
to reserve lodging in advance. It was
necessary to have alternative choices avail-
able at all times.
The time span between first and second
visits to Phase I streams was less in several
cases than the recommended 14 days. Delays
in gaining access permission for the first
site visit, limited flexibility in the schedule
of moves between base locations, and more
rapid progress on the second cycle of samp-
ling in each base area all contributed to
this situation.
Local cooperators provided essential
help in locating sites before and during the
survey. Samplers were often accompanied
to sites by the cooperators or by landown-
ers. Sampling schedules often had to be
altered in order to accommodate their sche-
dules. Additional time and schedule flexi-
bility should be allowed in future surveys
for important cooperative efforts such as
these. More use should be made of local
expertise for identifying reaches not suitable
for sampling prior to the beginning of the
sampling effort, for example reaches with
tidal influence.
Site access
Some delays were encountered prior to
the beginning of the survey in gaining stream
site access permission and stream location
Information. Logistics coordinators spent
a large amount of time during the survey
contacting landowners and agency personnel.
For the most part, landowners were
cooperative and agreeable in allowing access
to streams. Of more than 600 stream visits,
only seven streams could not be sampled
because access had been denied. In the
lower mid-Atlantic, sampling at several sites
owned by mining companies was delayed
until nearly the end of the survey; this
forced the timespan between the first and
second visits to drop to 8 days, which was
well below the 14-day recommended minimum.
Stream Site Location
Physical access to the majority of
sites was straightforward; most streams
were within short or easy walking distance
from a roadway. Although very few problems
were encountered in locating sites, exceptions
are discussed below.
The process of verifying the stream
as the correct site and of finding the optimal
sampling point often proved time-consuming.
In two cases only, samplers realized on
their second visit to a site that they had
sampled the wrong stream on their first
visit. Many of the coastal areas had few
roads, and traversing marshy, unfamiliar
territory on foot was difficult. Stream
sites in coastal areas were difficult to locate
25
-------
Guidelines for how long to delay
before sampling following a rain-
storm not comprehensive.
Weather information sketchy and
unreliable.
Assignment of upstream and
downstream site sampling order
could not always be random.
Impractical to assign duplicate and
blank samples to sites requiring
lengthy or especially difficult
hikes. Samplers became over-
fatigued.
Time between first and second
visits to site in Phase I areas less
than 14 days in some cases.
Avoided sampling if water
appeared turbid or water level was
up-
Made decisions based on calls to
many different sources.
Exceptions to random sampling
order were made when risks for
sample contamination or safety
were involved.
None employed during survey.
None employed during survey.
Set conservative, more definitive
guidelines for baseflow sampling
after rain events.
Subscribe to a weather service
that gives information by county.
Take this possibility into consider-
ation during planning stages.
Take this possibility into consider-
ation during planning stages.
Plan schedules to allow more time
for the first sampling cycle than
for the second cycle.
Problems with hotels: Phone
messages lost; incorrect informa-
tion given out.
Length of workdays at times
caused excessive fatigue.
Drivers' fatigue occurred
frequently.
Samplers all arrived back at base
site at different times. Difficult
to hold group "debriefing."
LOGISTICS
Warned callers of problem.
Requested cooperation from hotel
management.
None employed during survey.
Alternated long drives between
teams.
Scheduled evening group meetings
or met with each team separately.
Spend more time during reconnais-
sance discussing project needs
with hotel management.
Reduce work load: Keep episodes
work separate from NSWS survey
work.
Reduce overall work load. Choose
base sites carefully to limit length
of drives.
Lessen work load so that samplers
return earlier in the day so that
meetings can be held during
regular work days.
Tidal streams often not recognized
as tidal by samplers.
Stream dry, stagnant, or flowing
underground for a large proportion
of its length.
SITE ACCESS AND INFORMATION
Stream site data were placed in a
"non-interest category" of streams
in the data base.
Moved sampling site to wherever
feasible to sample. If necessary
to move 70 to 90% of stream
length, sample taken at only one
site. If > 90% of reach could not
be sampled, site was eliminated.
Use other indicators in addition to
conductivity to identify tidal
streams, such as appearance of
stream banks and vegetation type.
Obtain this information from
cooperators, if possible.
Presampling reconnaissance would
eliminate these streams.
(continued)
26
-------
Table 9. (continued)
—
PROBLEMS
•
On second visit, originally sampled
site was dry.
Sampling location at special-inter-
est sites not always readily
apparent.
More time than anticipated spent
accessing sites and verifying
stream ID.
SOLUTIONS EMPLOYED
Moved upstream or downstream
until an acceptable site was found.
Filled out new watershed charac-
teristics form and made notes on
data form.
Sampled at gauging station if
present; at most representa- live
spot; or at most downstream of
several possible sites.
Revised schedules to accommodate
additional time.
RECOMMENDATIONS
Pick sampling sites more carefully,
considering lower water conditions
that might occur later on.
Clarify site information with
cooperators, or have cooperators
accompany sampling teams.
Do presampling reconnaissance for
all streams; increase time for
training in orienteering.
Unacceptably early shipping dead-
lines in smaller towns.
Samples near 0 "C or partially
frozen on arrival at processing
laboratory.
Samples greater than 4 °C on
arrival at laboratory.
Shipments of supplies and stan-
dards inadvertently sent to
incorrect field location by ware-
house personnel.
Styrofoam coolers with heavy
Cubitainers in them broke up
during shipment.
Information sent by courier to the
Las Vegas communications center
sometimes lost.
A few shipments were lost or
misrouted by the express courier
service.
Nearly ran out of certain supplies.
Cubitainers leaked during ship-
ment.
SHIPPING
Moved to larger towns with later
shipping times, made special
arrangements with couriers, drove
long distances to places with later
shipping times, or had samplers
ship from remote locations.
Packed samples with fewer gel
packs. Qualified data.
Used only fully frozen gel packs,
higher percentage of nonplastic
gel packs. Qualified data.
Base coordinators contacted com-
munications center on morning of
each shipment and gave shipping
destination for that particular
shipment.
Shifted to hard, plastic coolers
whenever available and reinforced
styrofoam coolers with strapping
tape.
Better communication on shipments
to and from the field.
Tracked and recovered shipments,
but incidents inconvenienced field
operations.
Conserved supplies and shifted
excess supplies between sampling
regions.
Checked that caps were fully
tightened before shipping, checked
Cubitainers for holes.
Choose sites with late shipping
times, whenever possible.
Same as solution.
Use only hard, plastic coolers with
a combination of hard, plastic and
soft-sided gel packs.
Same as solution.
Use only hard, plastic coolers.
Require tracking system for all
shipments to and from field.
Require tracking system for all
shipments to and from field so
missing shipments can be traced
quickly.
Set up computerized inventory
system and make better predic-
tions of supplies needed.
Same as solution.
(continued)
27
-------
Assignment of blanks and dupli-
cates often unclear.
Difficult for base coordinators to
limit the number of calls to the
communications center.
Feedback from laboratory to base
coordinators on sample condition
upon arrival from field not always
received from communications
center.
The laboratory had difficulty
separating Phase I, visit 1 samples
from Phase I, visit 2 samples.
Samplers could not anticipate
when or where they would find
telephones to call in and would
often miss scheduled calls.
COMMUNICATIONS
Took more blanks and duplicates
than were needed for quality as-
surance program.
Called as needed. Communications
center provided a staff which was
sufficient to handle all calls
None employed during survey.
Samples were clearly labeled "Visit
1" or "Visit 2."
Samplers maintained call-in sched-
ule whenever possible.
Schedule more specific call-in
times to improve communications.
Allow morning and evening calls
Provide staff to handle peak
times. Provide "call-in" check list
to base coordinators so they have
reminders of needed information.
Provide more direct communication
from laboratory supervisor to field
base coordinators.
Incorporate a space on the label
for this information.
Presampling reconnaissance should
include information on locations of
telephones in remote areas.
No definition for "slow" stabiliza-
tion for pH measurements.
Lower range for YSI conductivity
meter had poor resolution.
Difficult to prevent pH probe from
touching beaker walls.
Conductivity QCC solutions were
not always accurate at the begin-
ning of the survey.
Large differences occurred bet-
ween stream temperature and QCC
solutions used at streamside.
EQUIPMENT AND PROTOCOLS
No consistent guidelines followed.
None.
Required constant attention.
Improved the preparation protocols
for QCC solutions.
Kept standards protected from sun
and wind as much as possible.
Define "slow" stabilization (e.g., >
5 trials or > 2 min. per trial).
Consider other meters for future
work.
Use a pH stand to hold electrode
during measurements.
Continue to follow protocol
designed during NSS survey.
Keep QCC solutions in insulated,
opaque containers.
Braided stream channel made
hydrologic measurements difficult.
Downstream site not always
suitable for hydrology.
Lower water levels on second visit
caused many steel rods to be out
of water.
HYDROLOGY
Moved upstream or downstream to
an unbraided channel or sampled
on largest of channels.
Did hydrology further upstream or
at upstream site. Filled out new
watershed characteristics form for
hydrology site and marked map.
Made best estimate possible of
drop in stream level.
Take samples and measurements at
same site as hydrology.
Have criteria pre-established for
choosing alternative hydrologic
locations.
Choose location for steel rod more
carefully; use some kind of per-
manent marker in addition to steel
rod.
(continued)
28
-------
Table 9. (continued)
In many Screening area streams,
10 meters was too long a distance
to measure flow velocity.
Many channels not suitable for
estimating hydrology.
Very few target episodes streams
suitable for episodes sampling.
The 0.20-inches-of-rain- within-
-24hours rule was not a consis-
tently viable criterion for switch-
ing to episodes sampling.
The pH changed significantly
between replicates during events,
preventing stable reading.
If episodes sampling done after
second site visit, steel rod for
stage measurement already re-
moved.
Weather at base site not indicative
of weather at stream sites.
Survey suffered under severe time
constraints because the Phase I
and Episodes Pilot surveys were
combined.
Episodes sites were difficult to
find and access at night.
Insufficient directions to replace-
ment teams or runners caused
delays in reaching site during
episodes.
Used shorter measured distance of
variable length for estimate.
Best estimate possible was made.
EPISODES PILOT
Selected additional streams which
were not on original target list
and that had low or moderate
ANC.
Used best judgment.
Took one Cubitainer of stream
water for each reading, and took
replicates from it.
Stage measurements from episode
do not relate to stage measure-
ments fore Phase I in data base.
Decision to attempt episodes
sampling was based on weather at
base site combined with forecasts.
It was often wrong.
The Phase I survey was success-
fully completed at the expense of
the Episodes Pilot survey.
Samplers allowed more time for
setting up episodes sampling
stations.
Maps and careful directions were
given to replacement teams and
runners if they could not accom-
pany teams to site.
Use flow meter to make measure-
ments at all streams.
Use flow meter to make measure-
ments at all streams.
Do reconnaissance of all potential
episodes streams to verify that
they are suitable for sampling.
Allow time for full-scale episodes
project.
Use same technique used during
Episodes Pilot or use continuous
monitoring equipment.
Use permanent marker for stage
measurement in addition to steel
rod.
Subscribe to county-by-county
weather forecasting service. In-
clude time for several "false
alarms" in overall schedule.
Conduct projects separately with a
separate set of personnel.
Mark sites well, or require that
samplers arrive before nightfall.
Require that runners and replace-
ment teams have visited site at
some time previously.
because the actual channels seldom corres-
ponded to what was shown on the maps.
Stream channels influenced by tides, but
with conductivities lower than 250 ^S/cm,
were sometimes not recognized as tidal on
the first visit. These streams were subse-
quently excluded from the target population.
Several sampling situations were en-
countered during the survey which had not
been anticipated before the survey began.
These included:
• Streams having poorly
multiple parallel channels
defined or
29
-------
bite reconnaissance will reduce time spent
a '
Stream Channel and Flow
Measurements
ihese variable stream characteristics re-
were used successfully to estimate flow in
S°me °f these cases
TK +. . .
me method of estimating stream depth
drop a great deal during the time between
•*• «« - rodsmjrhs noTngar par" y
could di9 a cha""el
lower
area was
sirs
loo deep to
sa
rods were removed or oherwise re m
30
-------
pered with between first and second visits,
no clear way existed to determine the
change in stage between the first and second
visits to the site. In future work, more
training should be given on the optimal
placement of staff gauges.
Shipping
Several minor problems with sample
shipping were noted and solved during the
survey. Shipping cooler temperatures some-
times deviated significantly from the recom-
mended 4 'C upon arrival at the processing
laboratory. Either coolant packs were not
sufficiently frozen to keep samples cold
or the use of too many coolant packs caused
the cooler to drop below freezing. Numbers
and types of coolant packs were adjusted
to overcome this problem. Styrofoam cool-
ers loaded with heavy samples and refriger-
ant' packs, frequently cracked during trans-
port This problem was solved by using
hard plastic coolers or by taping the more
fragile, styrofoam containers for reinforce-
ment.
Some samples were damaged in ship-
ment The actual incidence of this problem
was very low, but problems that did occur
included leaks in Cubitainers, which were
caused by imperfections in the container
or by improperly tightened caps; also, syr-
inge tips occasionally broke off in transit.
Samples occasionally arrived late be-
cause of problems with the overnight courier
service. The early shipping deadlines for
smaller towns used as base sites frequently
required making special arrangements with
the couriers, driving long distances to cen-
tral package drop-off stations in larger
cities, shipping by counter-to-counter airline
service, or having samplers ship from loc-
ations closer to the stream sites. The early
courier service deadlines in some areas often
limited sample collection to one stream per
day per team. A mobile field laboratory
located nearby, rather than in Las Vegas,
would in many cases have allowed the later-
return of the sampling teams and thus would
have extended the effective sampling day.
Advantages of centrally locating the labor-
atory (see page 12) far outweighed the
disadvantages, however.
In the beginning of the survey, some
supply shipments were misrouted and some
field supplies became depleted. These prob-
lems were resolved and did not cause delays
in the project.
Equipment
The incidence of equipment failure
was very low. In those few cases when
failures did occur, alternative equipment or
repairs solved the problem. Experience
gained from the NSS-PS proved to be inval-
uable in providing guidelines for meter care
and troubleshooting.
The pH meters performed well, but
the responsivity of many of the pH electrodes
declined significantly near the end of the
survey Poor functioning was evidenced by
failure to meet quality control requirements
and by slow electrode response times. As
soon as symptoms were noted, the failing
electrodes were replaced and were returned
to Las Vegas for re-etching, according to
the instructions provided by the manufacturer.
As in the NSS-Pilot, the pH meters were
prone to moisture problems. Some pH
measurements were lost because meters
malfunctioned in wet weather. In a few
cases, samplers dropped meters into streams
and did not have backup meters with them.
In these situations samplers collected an
additional Cubitainer of stream water and
measured pH with a spare meter at the
base site.
An additional problem with the pH
meters was related to the practice of twirling
the probe in the air by the leads in order
to remove air bubbles from the coils. Probes
occasionally had to be replaced because of
faltering leads.
As in the NSS-Pilot, it was noted that
the YSI conductivity meter had poor reso-
lution at the low ranges which are most
important for the survey. This meter was
31
-------
chosen despite this limitation because of
its durability and reliability in comparison
to other more sensitive meters. There was
a shortage of spare conductivity meters as
well as of Marsh-McBirney current meters-
this shortage could not be alleviated during
he survey. Very few measurements were
lost because of this shortage.
Safety
+K MO w "'~ '*"'yni en iu ouuue of
tne NSS, the number of people involved
in the field, and the number of miles trav-
eled the safety record was excellent No
work-related injuries occurred, although
one of the field vehicles suffered severe
damage in a driving accident.
The communications link between field
samplers and coordinators was maintained
at almost all times, despite the difficulty
of locatmg telephones in some remote areas
Only ,n one instance did a sampling team
miss the required call before the deadline-
they were able to call just before initiation
of rescue operations In this case the
stream site was so remote and difficult to
reach that the samplers underestimated the
time needed to complete their work.
Episodes Pilot
Two major and unforeseen factors
constrained the Episodes Pilot Study m
usual weather and a shortage of available
samphng time. Because of schedule pres-
sures, base coordinators continuously had
o dec.de between (1) remaining in the Phase
I sampling mode and maintaining schedules
and (2) preparing for episode sampling
whenever rain was predicted.
The spring of 1986 was particularly
dry. Moreover, the few precipitation events
that did occur within the sampling window
never consisted of major rain fronts but
were small systems with localized showers
This greatly hampered the forecasting ability
of the weather services, which were called
at least twice a day. Table 10 illustrates
the unreliability of weather predictions for
the lower mid-Atlantic region. Coordinators
were faced with the dilemma of losing a
day of Phase I sampling while waiting for
an episode that might never materialize or
continuing with Phase I sampling but taking
a chance on missing the beginning of an
ep.sode. The latter was considered the
most practical and explains why a complete
sample set from storm episode hydrographs
could not be obtained in some cases
Table 10. Weather Predictions for the
Lower Mid-Atlantic
'PREDICTION MEASURED (inches)
March 17
March 18
March 19
March 20
March 26
March 27
April 4
April 5
April 6
April 7
April 8
April 9
April 10
April 11
April 14
April 15
April 16
April 20
April 21
April 22
April 28
April 29
April 30
May 1
May 6
May 7
May 11
May 13
May 14
May 15
chance
60-100%
90%
70%
40%
40%
chance
70%
30-60%
partly sunny
30%
60%
30-50%-
20-40%
chance
50-60%
90%
80%
70%
chance
50%
40%
50%
chance
30%
30-60%
40%
50-70%
40-60%
30-40%
• •
o
\J
0.17-0.22a
0
0.03-0.14
0.20-0.82a
0.20-0.36
Q
1.25*
0.25
0.28a
0.14
0.54
0
Q
0.21-0.78
Q
0.12-1.25
0.21
0
fcpisoaes samples collected.
Even if a significant amount of rain
fell at the base site, the scattered nature
of most precipitation events precluded any
judgement about rainfall at streamsites 20
to 50 miles away. Almost all precipitation
events occurred in the evenings or at night
when sampling teams had returned from a
full day's work; therefore, sampler fatigue
32
-------
was a constant consideration. For safety
reasons, the coordinators sometimes decided
to send only two episode teams instead of
three into the field. The remaining samplers
represented the relief crew for the second
shift and were able to rest for a few hours
before relieving some of the samplers still
in the field.
Frequently the episode streams were
unsuitable for prolonged sampling during
a precipitation event for the following
reasons:
• Some streams were not within easy
driving distance from the base site.
must be available to wait for storms that
may or may not materialize. The experience
gained during the Episodes Pilot has demon-
strated that the probability design employed
in the Pilot is not logistically practical.
Despite uncooperative weather conditions
and schedule limitations that reduced the
number of episodes samples to a total which
was well below the anticipated number, the
information gained from the Episodes Pilot
survey will be useful in planning future
episodes work.
Summary
The NSS involved approximately 1400
a residence, and all-night sampling
would have disturbed the landowners.
• Some streams were not easily acces-
sible from a road, required long hikes,
and were difficult to find, particularly
in the dark.
The existing protocol for when to
return to routine Phase I sampling after a
rainstorm was not comprensive. The amount
and duration of rainfall at the base area
often were very different from that at
actual stream sites, so a protocol based on
the duration of rainfall did not work well
in practice. It was difficult to judge by
visual inspection whether streams had re-
turned to baseflow. Each area sampled was
very different in its response to rain, i.e.,
how much rain it took to visibly alter the
streams. In some areas a small amount of
rainfall would appear to affect the streams
much more than a larger amount of rainfall
in a different area that had different
geology.
Snowstorms occurred in the upper and
the lower mid-Atlantic. During snowmelt
from these storms, streams possibly could
have been sampled as events if schedules
had been less rigid.
Future episodes work should not be
combined with other survey work. Time
lent safety record. Minor problems that
were encountered during the survey were
all solved in a timely manner. A high degree
of advance planning, organization, cooper-
ation, and good communication among all
groups involved in this large-scale survey
was vitally important to its success.
33
-------
1.
3
References
o,,,,.,,,, T«~*, • ^ ,.V~" "' K|W<-0<-1U1 c ror gaging streams
.__. Purvey, Techniques of Water Resources Investiqations of the Unit**
Geological Survey, Book 3, Chap. A6. 13 pp. ^^anons of the Un.ted
man; ," Td H' R Hem0nd 1985' ™e role of or9anic acids in the acid-
Res 2l!l503-1°510SU " W3terS * ^^ Wat6rShed' Massartusett.. ^ ler Hour.
5 chemistry
^
r CVA/M;' °o L Mayer' D" V" PeCk' J" R- Baker- and G- J- Filbin. 1987 National
Surface Water Survey, National Stream Survey (Phase I-Pilot Survey) Field
Nevada P°rt ^ 6°0/8-87/019' U'S' Environmental Protection Agency, S
t' °' H- Landers' J' M- Eilers- D- F- Brakke, W. S. Overton E P
Vo Pnr', t- Cr°nWe- -1986- Characteristics °f Lakes in the Eastern United States!
US Environml t'Tp !SCr'Ptl0rS a"d Pnysic°-Chemical Relationships. EPA600/4-86/007a
U.S. Environmental Protection Agency. Washington, D.C.
9' SJ! ! 6r< cJ' J" nu N- Eshleman- s- Stambough, and P. Kaufmann. 1986. National
meSpfoSo A'56 ' -/il0VSUrVey °ata Rep0rt EPA 600/4-86/026, U.S. EnvET
mental Protection Agency, Corvallis, Oregon.
t0 br°°k trout
n H^ m nWaterS- /n; P°//Uted Ra/n' T- Toribara- M. Mille, and P
Morrow (eds.), Plenum Press. New York, pp. 341-362.
-J- OK' Ga"°Way- 1982' Acid P^cipitation: The impact of two
'" Shenandoah National Pa*. Virginia. Int. Symp. on Hydromete-
ology P
34
-------
Appendix A
Data Forms Used in the National Stream Survey
NATIONAL SURFACE WATER SURVEY
STREAM DATA
FORM 4
STREAM ID:
STREAM NAME:
D D M
PROGRAM:
D PHASE 1
D SCREENING
D EPISODE PILOT
TIME
START
-INISH ;
U/L
Fl FVATIDN-
PHASF 1 VISIT «-
M M Y Y
SAMPLES COLLECTED
DHOUTINE
D DUPLICATE
d BLANK
GAUGE HEIGHT (II)
O
- o
(FIELD RECALIBRATION?) O O
QCCS -pH 4. 00 1
OCCS INITIAL: . 'Q
ROUTINE
SAMPLE TEMP.:
DUPLICATE
SAMPLE TEMP.
QCCS FINAL:
Oi
-c O
O
-c O
O
EPISODE SAMPLE TYPE /">
D BASE FLOW - EPISODE ONLY ^^
D BASE FLOW - EPISODE AND PHASE 1
D RISING STAGE
D PEAK STAGE
D FALLING STAGE
RAIN £\
(CHECK ONE ONLY)
QNO
D PREV D MOD
D LIGHT D HEAVY
CLOUD COVER
%o
UNCOMPENSATED
CONDUCTIVITY uS cm-1
QCCS INITIAL: (~)
OCCS TEMP: _
IN SITU:
STREAM TEMP.: _
QCCS FINAL:
QCCS T.EMP : _
-c O
o
-c 0
o
-c O
SHIPPING INFORMATION
D D M M M Y Y
SHlpppn fanu
Tn
D FED. EX D SATURDAY DELIVERY
d COMMFRP.IAI
» OF nooi FHS
TOTAL « OF SAMPI FS
» OF SAMPLES THIS
r.nra FR
DISSOLVED OXYGEN mg / 1
QCC - Theoretical — Measured
INITIAL: 1 1 . (^)
IN SITU: O
FINAL. 1 H O
COOLER TEMPERATURE
AT SHIPMENT ON RECEIPT
°0 ',-
BATCH in
3 DUPLICATE RAMPI F in
D Rl ANK SAMPI f in
n RAKE
PI RISC
D PFAK
d FAI 1
NOT SAMPLED
D INACCESSIBLE
O NO ACCESS PERMIT
D TOO SHALLOW
n
FIELD CflEW DATA
-RFW in
SAMPI FR 1
SAMPI FR .1
CHFCKFn HY
DATA QUALIFIERS
A INSTRUMENT UNSTABLE
D SLOW STABILIZATION
Q DID NOT MEET QCC
*
Y
7
FORM DISTRIBUTION
WHITE COPY - ORNL
PINK COPY EMSL-LV
YELLOW COPY — FIELD
ORANGE COPY — MOBILE LAB
Revised 1-6-86
GILL'S (702) 362-2100
Figure A-1. NSWS Form 4.
35
-------
NATIONAL SURFACE WATER SURVEY
HYDROLOGIC DATA
FORM 4A
SHEET.
D D M M M
DATE:
FLOW METER ID: .
STREAM NAMF'
ESTIMATED HYDROLOGY:
DEPTH (max.-fl.)
TIME START: : W|DTH (me,ers)
TIME END: : VELOCITY (m sec -1
SAMPLE TYPE:
D PHASE I
D SCREENING
D EPISODE PILOT
._c
._c
. c
EPISODE TYPE: CHECK ONE
D BASE - EPISODE ONLY
D BASE - EPISODE AND PHASE I
PRISING
D PEAK
D FALLING
)EST. MEAS. f-^.
n a U
) D D O
) D D O
MEASURED HYDROLOGY:
TIME:
START : _
FINISH :
STAGE
INTERVAL CENTER (m)
1.
2.
3.
4.
5.
6.
7.
8. (min)
9.
10.
11.
12.
13.
14.
15. _
O
O
O
. O
. _o
O
O
O
. _o
O
. _o
O
. _o
O
O
(ft) STEEL ROD STAGE (ft.) WIDTH (m)
_o
_ _o
Intorual Width
DEPTH AT
CENTER (It)
. _o
. O
. _o
. O
0
. O
. O
0
. O
. O
. O
. O
. O
O
. _o
0 0
0 0
(cm)
VELOCITY AT
CENTER (m sec"1 )
. O
O
. O
. O
. O
0
. O
O
. _o
O
O
. O
. _o
O
. O
COMMENTS:
FIELD CREW DATA:
SAMPLER 2:
SAMPLER 3:
CHECKED BY:
DATA QUALIFIERS
(K) INSTRUMENT UNSTABLE
(b) SLOW STABILIZATION
(5) DID NOT MEET OCC
® •
—
C2)
FORM DISTRIBUTION
WHITE COPY — ORNL
YELLOW COPY- FIELD
Revised 1-86
GILL'S (702] 362-2100
Figure A-2. NSWS Form 4A.
36
-------
NATIONAL SURFACE WATER SURVEY
STREAM EPISODE DATA
FORM 6
D D
DATE BEGIN:
DATE END:
M M M Y Y
TIME:
ARHIVAI
U/L
L
STREAM NAME:
BASE FLOW SAMPLE
RISING SAMPLE
PEAK SAMPLE
FALLING SAMPLE
DEPARTURE
INCREMENTS)
PRECIP. (in)
.0
.o
-O
o
UNCOHHECTED
COND. (uS cm 1)
DISS Oj
(mg/l)
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o .
o .
o
0 .
o
o
o
o .
o .
o
0 .
0 .
0 .
o
o—
_ 0
0 .
0 .
0 .
0 .
o .
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
_o
o .
o
o
o
o
o
o ._
COMMENTS:
o o
o o
- O O
o o
- O O
_ o o
- O O
_ o o
o o
o o
o o
- O O
- O O
o o
- O O
__o o _
o o
__o_ o
o o
_ o o
- O O
o o
- 0 O
DATA QUALIFIERS
fA)INSTRUMENT UNSTABLE
(D)SLOW STABILIZATION
(QIDID NOT MEET QCC
rift
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
I!~o
o
o
o
o
o
(j)Base Flow
(2)Rising
(3)Peak
(4)Falling
n
o
n
n
n
n
n
n
n
D
n
n
n
O
n
)OOOOC
I Ml
O
n
0
(2)
FIELD CREW DATA
HRFW in
RAUPI FH 1
RAMPI FR 7
RAMPI FR n
r.HFrKFn RY-
FORM DISTRIBUTION
WHITE — ORNL
PINK — E
ilSL-LV
- FIELD
-6-86
362-2100
Figure A-3. NSWS Form 6.
37
-------
NATIONAL SURFACE WATER SURVEY
WATERSHED CHARACTERISTICS
FORM 7
D D M M M Y Y
DATE.
STREAM ID U/L STREAM NAME LATITUDE:
LONGITUDE:
COUNTY STATE 1:250.000 MAP NAME MAP DATE ULLVAUUN.
STREAM WIDTH (ml
1:24.000 MAP NAME MAP DATE
STREAM DEPTH (m)
WATERSHED ACTIVITIES/DISTURBANCES
(Check all that apply)
Distance From
Stream (meters)
D Roadways Along Stream:
n Pav^H
D Crossings Above Stream:
H Culvert
n Bri'lgfMl
D Dwellings:
n Ring IP
n MniiipiP
D Agriculture:
n FonrpH
PHOTOGRAPHS COMMENTS:
FRAME ID AZIMUTH
Q LAP CARD
o -
o •
0 ' //
MEAS. EST.
n n
n n
BANK COVERAGE WITHIN 100 METERS OF O
STREAM BED {Check a I that apply)
Type Abtent Sparse Moderate Heavy
<25% 25-75% >75%
Deciduous Trees: Q D D D
Coniferous Trees: D D D D
Shrubs: Q Q D Q
Wetland Areas: D D D D
Grasses and Forbs: D D Q D
Moss: D D Q D
Rocky/Bare Slopes: D d D D
STREAM SUBSTRATE O
{Check all that apply)
Type Ab*ent Spar«« Modvrale Heavy
< 25V. 25-75% > 75%
Boulders: > 25 cm CD D d D
Cobble: 6-25 cm D D D D
Gravel: 0.2-6 cm D D D D
Sand: < 0.2 cm D D D D
Silt and Clay: D D D D
Aufwuchs: D D D D
FIELD CREW DATA DATA QUALIFIERS
ORFW in fy\
QAMPI F« ' fVi
SAMPI FR ? &>
£4MPI FO 3
rHPr.kFn RY
FORM DISTRIBUTION
White Copy — ORNL
Pink Copy — EMSL-LV
Yellow Copy — FIELD
Revised 1-86
GILL'S (702) 362-2100
Figure A-4. NSWS Form 7.
38
*U.S. Government Printing Office : 198
-------
SUBREGIONS OF THE NATIONAL STREAM SURVEY-PHASE I
o M is! cA
t? t'-', til «.
H. <"., F) 03
o ;-* '
I ' ^ § f
- W O'i '--1
(U - H-
HI & H
fcl H i— O
? -:: §
C '- 3
4 -: C3
0)
O H-
& O
3
-3 TO
O CD
Northern
Appalachians (2Cn)
Valley and Ridge (2Bn)
Poconos/Catskills (ID)
Southern Blue Ridge (2As)
(Pilot Study)
Mid-Atlantic
Coastal Plain (3B)
Ozarks/Ouachitas (2D)
••11
Southern-Appalachians (2X)
------- |