EPA/600/R-13/170F I September 2014 I www.epa.gov/ncea
United States
Environmental Protection
Agency
Best Practices for Continuous
Monitoring of Temperature and Flow
in Wadeable Streams
National Center for Environmental Assessment
Office of Research and Development
-------
EPA/600/R-13/170F
September 2014
Best Practices for Continuous Monitoring of Temperature and
Flow in Wadeable Streams
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection Agency
and U.S. Geological Survey policy and approved for publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use, but is for
descriptive purposes only. This document does not supplant official published methods and does
not constitute an endorsement of a particular procedure or method.
ABSTRACT
The United States Environmental Protection Agency (U.S. EPA) is working with its regional
offices, states, tribes, river basin commissions and other entities to establish Regional Monitoring
Networks (RMNs) for freshwater wadeable streams. To the extent possible, uninterrupted,
biological, temperature and hydrologic data will be collected on an ongoing basis at RMN sites,
which are primarily located on smaller, minimally disturbed forested streams. The primary
purpose of this document is to provide guidance on how to collect accurate, year-round
temperature and hydrologic data at ungaged wadeable stream sites. It addresses questions related
to equipment needs, sensor configuration, sensor placement, installation techniques, data
retrieval, and data processing. This guidance is intended to increase comparability of continuous
temperature and hydrologic data collection at RMN sites and to ensure that the data are of
sufficient quality to be used in future analyses. It also addresses challenges posed by year-round
deployments. These data will be used for detecting temporal trends; providing information that
will allow for a better understanding of relationships between biological, thermal, and hydrologic
data; predicting and analyzing climate change impacts and quantifying natural variability.
Preferred citation: U.S. Environmental Protection Agency (EPA). (2014) Best Practices for Continuous
Monitoring of Temperature and Flow in Wadeable Streams. Global Change Research Program, National
Center for Environmental Assessment, Washington, DC; EPA/600/R-13/170F. Available from the
National Technical Information Service, Springfield, VA, and online at http://www.epa.gov/ncea.
11
-------
CONTENTS
TABLES vi
FIGURES vii
ABBREVIATIONS AND ACRONYMS x
PREFACE xi
AUTHORS, CONTRIBUTORS, AND REVIEWERS xii
EXECUTIVE SUMMARY xiii
1. PURPOSE AND SCOPE 1
2. TEMPERATURE 3
2.1. Equipment 3
2.1.1. Basic components 3
2.1.2. Considerations when choosing temperature sensors 3
2.1.3. Considerations when choosing portable data offload devices 5
2.1.4. Considerations when choosing radiation shields 7
2.2. Pre-Deployment 9
2.2.1. Accuracy check 9
2.2.2. Sensor configuration and launch 12
2.3. Deployment of Water Temperature Sensors 14
2.3.1. Guidelines for placement within the stream 14
2.3.2. Installation 15
2.3.3. Documentation 22
2.3.4. Common problems 25
2.4. Deployment of Air Temperature Sensors 26
2.5. Maintenance/Mid-Deployment Checks and Data Offload 28
2.6. Quality Assurance and Control 31
2.6.1. Mid- and post-deployment accuracy checks 32
2.6.2. Error screening 32
2.6.3. Record keeping 35
in
-------
CONTENTS (continued)
3. HYDROLOGY 35
3.1. Equipment 35
3.1.1. Basic components 35
3.1.2. Considerations when choosing pressure transducers 36
3.1.3. Staff gage 38
3.1.4. Protective housing 40
3.2. Pre-Deployment 41
3.2.1. Accuracy Check 41
3.2.2. Sensor configuration and launch 41
3.3. Deployment of Instream Pressure Transducers 43
3.3.1. Selecting a location 43
3.3.2. Staff gage installation 44
3.3.3. Pressure transducer installations 46
3.4. Deployment of On-Land Components 51
3.5. Elevation Surveys and Documentation 53
3.6. Maintenance/Mid-Deployment Checks and Data Offload 55
3.7. Quality Assurance and Control 58
3.7.1. Accuracy checks 58
3.7.2. Error screening 59
3.8. Developing Stage-Discharge Rating Curves 61
3.8.1. When to measure discharge 61
3.8.2. Equipment 62
3.8.3. Site selection 62
3.8.4. Measurements 63
3.8.5. Documentation 64
3.8.6. Quality assurance and control 65
3.8.7. Making flow rating curves 65
4. LITERATURE CITED 66
IV
-------
CONTENTS (continued)
APPENDICES
A. How to construct PVC housings for water temperature sensors with cable
installations A-l
B. How to make and install homemade radiation shields for air temperature
sensors B-l
C. Temperature sensor calibration forms C-l
D. Temperature sensor deployment & tracking forms D-l
E. Equipment lists for temperature sensor procedures E-l
F. Examples of alternate temperature sensor installation techniques F-1
G. Temperature sensor mid-deployment check forms G-l
H. QA/QC checklist for temperature sensor data H-l
I. Equipment lists for pressure transducer procedures 1-1
J. Field forms for water level and flow measurements J-l
K. QA/QC checklist for pressure transducer data K-l
-------
TABLES
Table 1. Specifications for temperature sensors used in theRMNs 4
Table 2. Examples of commercially available temperature sensors 6
Table 3. The "ice bucket" method can be used for single-point pre-deployment
accuracy checks (MDDNR, no date) 10
Table 4. Multi-point pre-deployment accuracy check 11
Table 5. Equipment needs for the pre-deployment accuracy check 12
Table 6. Quick guide for doing underwater epoxy installations (Isaak etal. 2013) 16
Table 7. Equipment list for doing underwater epoxy installations (Isaak et al. 2013) 18
Table 8. Guidelines for documenting the installation 23
Table 9. Equipment list for documenting sites 25
Table 10. Tips for minimizing the chance of sensors being lost or damaged 26
Table 11. Checklists for performing maintenance/mid-deployment checks and data
downloads 29
Table 12. Equipment list for conducting maintenance/mid-deployment checks and
data downloads 31
Table 13. Error screening procedure (Dunham et al. 2005, Sowder and Steel 2012,
Personal communication, Michael Kashiwagi (MDDNR), April 17, 2014,
Personal communication, Dustin Shull (PADEP), February 2, 2014) 33
Table 14. General summary of different types of problems that can occur with
continuous temperature data and recommended actions for addressing them 34
Table 15. Examples of commercially available pressure transducers 39
Table 16. Equipment list for staff gage installation 45
Table 17. Equipment list for transducer installation 48
Table 18. Quick guide for elevation surveys of staff gages and transducers using an
auto level 53
Table 19. Equipment list for elevation surveys 54
Table 20. General summary of different types of problems that can occur with
pressure transducer data and recommended actions for addressing them 60
VI
-------
FIGURES
Figure 1. Numerous temperature sensors are commercially available 5
Figure 2. Additional equipment is needed to offload data from the temperature
sensors onto a computer 7
Figure 3. Stream temperature measurements from four sensors at the same site during
eight days in July 2010
Figure 4. Unassembled components of an inexpensive PVC canister radiation shield
for water temperature sensors (described in more detail in Isaak et al.
2013) 9
Figure 5. Examples of radiation shields for air temperature sensors include A) the
Gill-style Onset RSI solar radiation shield (www.onsetcomp.com) and B)
custom design by Zachary Holden (Appendix B) 10
Figure 6. Probability of underestimating the maximum daily temperature at least 1°C
in relation to daily range of temperature and sampling interval (Dunham et
al. 2005) 13
Figure 7. Examples of large rocks (a, b) and cement bridge pilings (c, d) that provide
good sensor attachment sites 17
Figure 8. Close-up of a PVC solar shield with a temperature sensor glued to a rock in
a stream 18
Figure 9. Equipment needed to permanently install temperature sensors in streams
using underwater epoxy 19
Figure 10. If conditions permit, attach sensors in two places, using heavy-duty (e.g.,
120-lb tensile strength) cable ties 20
Figure 11. Using bent rebar for rebar installations has several advantages. It is more
secure because the metal bar is anchored in two places 21
Figure 12. Example of how a cable and rebar installation is configured for large
streams in Maryland 21
Figure 13. Example of a rebar installation in a small stream in Shenandoah National
Park, VA 22
Figure 14. Example of a hand-drawn map from a field form used by the Washington
State Department of Ecology (taken from Ward 2011) 24
Figure 15. Metal forestry tags can be attached to the downstream side of large rocks
to monument sites and air in relocation of sensors (Isaak et al. 2013) 24
Figure 16. Numerous pressure transducers are commercially available 38
vn
-------
FIGURES (continued)
Figure 17. Example of a USGS style staff gage (Type A) marked in 0.02-foot
increments 40
Figure 18. Example of a protective housing made of PVC pipe. This one is for a
vented pressure transducer, which is secured inside the pipe with zip ties 40
Figure 19. Some transducers come in a protective metal case and do not need to be
encased in an additional housing 41
Figure 20. Staff gage readings provide a quality check of transducer data 42
Figure 21. Examples of controls downstream of staff gages include A. riffle and B.
culvert 44
Figure 22. Examples of gage installation techniques include A. streambed, B. fixed
object-bridge wing wall, and C. fixed object-boulder 44
Figure 23. Example of a 0.5-inch by 3.75-inch wedge anchor with bolt and washer 46
Figure 24. Non-vented pressure transducer installation 47
Figure 25. Non-vented (left) and vented (right) pressure transducers attached to staff
gage board using conduit hangars 48
Figure 26. Examples of streambed installations of pressure transducers using rebar 50
Figure 27. Vented transducer data logger installation. The transducer is attached to
the tree with stainless steel conduit straps 51
Figure 28. Barometric pressure sensor installation for a non-vented transducer, A.
without the PVC cap andB. with the PVC cap 52
Figure 29. Examples of auto level and tripod used for elevation survey (A) and
permanent structure used as a benchmark (B) 54
Figure 30. An example of a completed elevation survey form 56
Figure 31. Data download from vented transducer with external data logger on land
using computer and cable (left) and non-vented transducer with internal
data logger using computer and data shuttle (right) 58
Figure 32. Example of a transducer download field data sheet 58
Figure 33. A culvert replacement downstream of a stream gage on Gulf Brook in
Pepperell, MA caused enough of a channel change to necessitate a new
rating curve 62
Figure 34. Examples of good cross-sections for making discharge measurements 63
Vlll
-------
FIGURES (continued)
Figure 35. Layout of a channel cross-section for obtaining discharge data, using the
velocity-area procedure 64
Figure 36. Example of regular (left) and log-log scale (right) rating curves created
using Aquatic Informatics' AQUARIUS software 66
IX
-------
ABBREVIATIONS AND ACRONYMS
DRBC Delaware River Basin Commission
DTS distributed temperature sensing
FMU Freshwater Monitoring Unit
GPS global positioning system
GZF gage height of zero flow
NIST National Institute of Standards and Technology
PVC polyvinyl chloride
PZF point of zero flow
RIFLS River Instream Flow Stewards Program
RMN Regional Monitoring Network
QAPP quality assurance project plan
QA/QC quality assurance/quality control
SOP standard operating procedures
TIR thermal infrared
U.S. EPA United States Environmental Protection Agency
USGS United States Geological Survey
-------
PREFACE
This guidance document was prepared by Tetra Tech, Inc., the Massachusetts Department of
Fish and Game/Division of Ecological Restoration, the U.S. Forest Service, the Massachusetts
Cooperative Fish and Wildlife Research Unit of the U.S. Geological Survey, the U.S.
Environmental Protection Agency Region 3 office, and the Global Change Assessment Staff in
the Air, Climate, and Energy Program at the U.S. Environmental Protection Agency. The
document facilitates more uniform and effective collection of continuous temperature and stage
(depth) data at ungaged sites in wadeable streams, and addresses questions related to equipment
needs, sensor configuration, sensor placement, installation techniques, data retrieval, and data
processing.
XI
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
The Global Change Assessment Staff (GCAS) in the Air, Climate, and Energy (ACE) Program
of the National Center for Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency (EPA) is responsible for publishing this report. Tetra
Tech, Inc. prepared this document under Contract No. EP-C-12-060; EPA Work Assignment No.
0-01. Dr. Britta Bierwagen served as the Technical Project Officer.
AUTHORS
Center for Ecological Sciences, Tetra Tech, Inc., Owings Mills, MD
Jen Stamp, Anna Hamilton
Massachusetts Department of Fish and Game, Division of Ecological Restoration, Boston, MA
Michelle Craddock, Laila Parker
U.S. Geological Survey, Massachusetts Cooperative Fish and Wildlife Research Unit, Amherst
MA
Allison H. Roy
U.S. Forest Service, Rocky Mountain Research Station, Boise, ID
Daniel J. Isaak
U.S. Forest Service, Missoula, MT
Zachary Holden
U.S. EPA Region 3. Wheeling. WV
Margaret Passmore
U.S. EPA, Office of Research and Development Washington DC
Britta G. Bierwagen
REVIEWERS
U.S. EPA Reviewers
Richard Mitchell, Joe Flotemersch, and Lil Herger
External Reviewers
Hunter Carrick (Central Michigan University), Mark Rains (University of South Florida), John
Yearsley (University of Washington)
ACKNOWLEDGMENTS
The authors would like to thank the many partners who reviewed early versions of this draft for
clarity and usefulness. Their comments and input have substantially improved this document.
xn
-------
EXECUTIVE SUMMARY
The United States Environmental Protection Agency (U.S. EPA) is working with its regional
offices, states, tribes, river basin commissions, and other entities to establish Regional
Monitoring Networks (RMNs) for freshwater wadeable streams. To date, RMNs have been
established in the Northeast, Mid-Atlantic, and Southeast. Approximately 30 RMN sites are
being monitored in each region. Many of these sites are located on smaller, minimally disturbed
streams in forested watersheds with drainage areas less than 100 km . To the extent possible,
uninterrupted, biological, temperature, and hydrologic data will be collected on an ongoing basis
at RMN sites. Several entities already conduct annual monitoring at targeted sites through
existing programs. The goal is to integrate the RMNs into such existing programs. By
coordinating efforts regionally, resources can be pooled efficiently. Detailed information about
the development of the RMNs, including site selection criteria, biological sampling protocols,
and data usage, is compiled in a separate report.
The primary purpose of this document is to provide guidance on how to collect accurate, year-
round temperature and hydrologic data at ungaged wadeable stream sites. It addresses questions
related to equipment needs, sensor configuration, sensor placement, installation techniques, data
retrieval, and data processing. The protocols in this report describe simple methods and reference
inexpensive tools for collecting data sustainably over the long term. This guidance is intended to
increase comparability of continuous temperature and hydrologic data collection at RMN sites
and to ensure that the data are of sufficient quality to be used in future analyses. It also addresses
challenges posed by year-round deployments. These data will be used for detecting temporal
trends; providing information that will allow for a better understanding of relationships between
biological, thermal, and hydrologic data; predicting and analyzing climate change impacts; and
quantifying natural variability.
The section of this document on temperature describes protocols for measuring continuous water
and air temperature. At least one water and air temperature sensor should be deployed at each
RMN site. The air temperature data will be used to help determine if water temperature sensors
are dewatered during their deployment and will also provide important insights about the
responsiveness of stream temperatures to air temperatures and the differing vulnerabilities of
streams to thermal change. Two installation techniques for water temperature sensors are
described in this document: the underwater epoxy method and a method in which cabling is used
to attach the sensor to rebar or stable instream structures.
The hydrology section describes how to install and maintain vented and unvented pressure
transducers that measure continuous stage (depth). Staff gages should also be installed at RMN
sites because they provide a means for checking the accuracy of the transducer data. This report
also includes a brief section on how to take discharge measurements and develop stage-discharge
curves. Taken alone, stage measurements yield some information about streamflow patterns,
including the timing, frequency, and duration of high flows, but to better assess patterns and
changes in stream hydrology, converting stage measurements into streamflow is most useful.
This document also describes how to accurately geo-reference and document the temperature
sensors and pressure transducers in a way that enables field personnel to relocate them during
Xlll
-------
subsequent visits. Efforts should be made to revisit the sites within the first month of installation
to confirm that the installations are holding properly. After these initial deployment checks, sites
should be visited as frequently as schedules allow to check the condition of the equipment,
gather data for mid-deployment accuracy checks, and offload data.
After data are offloaded, quality assurance and quality control (QA/QC) procedures are
performed to verify the quality of the data and to check for potential errors. In this document, we
describe a series of checks to perform on the data, including identifying and removing
observations recorded before and after the equipment is correctly positioned and plotting
measurements to check visually for missing data, outliers, and other abnormalities. Large
amounts of data will accumulate quickly so a central database should be developed and
maintained from the initial stages of monitoring. Any changes made to the data should be
carefully documented, and all data forms should be organized, easily accessible, and archived in
a way that allows for safe, long-term storage.
xiv
-------
1. PURPOSE AND SCOPE
The United States Environmental Protection Agency (U.S. EPA) is working with its regional
offices, states, tribes, river basin commissions and other entities to establish Regional Monitoring
Networks (RMNs) at which long-term biological, thermal, and hydrologic data will be collected
from freshwater wadeable streams. At this time, RMNs have been established in the Northeast,
Mid-Atlantic, and Southeast, and these efforts are being expanded into other regions.
When selecting sites for the RMNs, attempts have been made to utilize and build on data already
being collected by participating entities. Site selection considerations include low level of
anthropogenic disturbance; long historical sampling record for biological, thermal, or
hydrological data; sustainability (e.g., accessibility, partners); classification; level of protection;
and expected vulnerability to climate change. Approximately 30 RMN sites are being monitored
in each region. Many of these sites are located on smaller, minimally disturbed streams in
forested watersheds with drainage areas less than 100 km2.
The need for the RMNs stems from the lack of long-term, contemporaneous biological,
temperature, and flow data, particularly at minimally disturbed, freshwater wadeable stream sites
(Jackson and Fureder 2006, Mazor et al. 2009, Kennen et al. 2011, Isaak et al. 2012). To the
extent possible, uninterrupted, biological, temperature, and hydrologic data will be collected on
an ongoing basis at RMN sites. These data will be used for detecting temporal trends; providing
information that will allow for a better understanding of relationships between biological,
thermal, and hydrologic data; providing information about response and recovery of organisms
to extreme weather events; analyzing climate change impacts (e.g., the National Climate
Assessment Indicator System); testing hypotheses and predictive models related to climate
change vulnerability; and quantifying natural variability. Several entities already conduct annual
monitoring at targeted sites through existing programs. The goal is to integrate the RMNs into
such existing programs. By coordinating efforts regionally, resources can be pooled to gain
efficiencies. More detailed information about the RMNs, including site selection criteria,
biological sampling protocols, and data usage, can be found in a separate report (U.S. EPA
2014).
The primary purpose of this document is to provide guidance on how to collect accurate, year-
round temperature and hydrologic data at ungaged wadeable stream sites. It addresses questions
related to equipment needs, sensor configuration, sensor placement, installation techniques, data
retrieval, and data processing. This guidance is intended to increase comparability of continuous
temperature and hydrologic data collection at RMN sites and to ensure that the data are of
sufficient quality to be used in future analyses. It also addresses challenges posed by year-round
deployments. Year-round monitoring of temperature and hydrology is fundamental to
understanding aquatic ecology. Although more and more continuous data are being collected,
many of these data are not being collected year-round. For example, considerable amounts of
stream temperature data are now routinely collected using inexpensive temperature sensors, but
many of the temperature measurements are made only during summer months due to logistical
constraints associated with stream access, concerns that large annual floods will destroy sensor
installations, and an intentional focus on the summer season because it captures a critical time
period for most aquatic species' survival. Summer-only data provide a narrow view of thermal
-------
regimes in streams and misses ecologically relevant information about the date of spring onset,
growing season length, overall variability, and total annual thermal units.
Flow regime (magnitude, frequency, duration, timing, and rate of change) also strongly
influences stream ecology (Poff et al. 1997). The U.S. Geological Survey (USGS) has been
collecting continuous hydrologic data in streams since 1889, and currently maintains over 7,000
continuous gages. The USGS stream monitoring network provides critical, long-term, high
quality information about our nation's streams and rivers that can be used for planning and trend
analysis (e.g., flood forecasting, water allocation, wastewater treatment, and recreation). Many
USGS stream gages are located in large rivers that have multiple human uses. Most RMN sites
are located on smaller, minimally disturbed forested streams with drainage areas less than 100
km2, so only a limited number can be co-located with USGS gages. This necessitates that
hydrologic data be collected independently. The collection of year-round hydrologic data from
headwater and mid-order streams is important. Streams of this size are instrumental in
connecting upland and riparian systems with river systems (Vannote et al. 1980). Moreover,
small upland streams, which are inhabited by temperature-sensitive organisms, are likely to
experience substantial impacts from climate change (Durance and Ormerod 2007) and could also
serve as refugia from the extremes in temperature and flow that are projected to occur (Meyer et
al. 2007).
This report is divided into two main sections: one on temperature and one on hydrology. The
temperature section describes protocols for measuring continuous water and air temperature. The
hydrology section discusses how to install and maintain pressure transducers and staff gages to
measure continuous stage (depth), and also includes a brief section on how to make discharge
measurements to develop stage-discharge curves. The protocols in this report describe simple
methods and reference inexpensive tools for collecting data sustainably over the long term.
Although more advanced equipment is available, such as distributed temperature sensing (DTS)
(Selker et al. 2006) and thermal infrared (TIR) imagery (Torgerson et al. 2001), most RMN
members lack the resources to purchase and deploy these types of advanced equipment at
numerous sites for long periods of time. Thus, they are not covered in this report. Temperature
and hydrologic modeling are also considered to be outside the scope of this report.
Although this document was written for the RMNs, much of the information and detail (e.g., data
forms) is useful to entities that are developing quality assurance project plans (QAPPs) and
standard operating procedures documents (SOPs) for measuring stream temperature and
hydrologic data with pressure transducers. We recognize that entities like USGS and some
monitoring programs have been early adopters of continuous sensor technology and have written
their own protocols for deploying temperature sensors and pressure transducers (e.g., MDDNR
no date, Danielson 2006, Wagner et al. 2006, Wilde 2006). The information in this document is
not meant to replace those SOPs, but it might supplement them. This document differs from
existing SOPs in that both temperature and stage (depth) information are compiled into one
place, and deployment techniques specifically address challenges posed by year-round
deployment. Although we attempted to make this document as comprehensive as possible, we
acknowledge that certain situations might warrant an alternative methodology, procedure, or
process. Future updates and improvements will be made as we receive feedback from partners
who are field-testing these methodologies.
-------
2. TEMPERATURE
This section describes protocols for measuring continuous water and air temperature. At least
one water and air temperature sensor should be deployed at each RMN site. The air temperature
data will be used to gain a better understanding of the relationship between air and water
temperature at each RMN site. This will provide important insights about the responsiveness of
stream temperatures to air temperatures and the differing vulnerabilities of streams to thermal
change. The data will also be used for quality assurance and quality control (QA/QC) procedures
(e.g., to help determine whether water temperature sensors are dewatered during their
deployment).
The protocols described in this section are based on a review of existing temperature protocols
from the Washington State Department of Ecology (Ward 2003, Bilhimer and Stohr 2009, Ward
2011), the U.S. Forest Service (Dunham et al. 2005, Isaak and Koran 2011, Sowder and Steel
2012, Isaak et al. 2013, Holden et al. 2013), Alaska (Mauger 2008), and Maryland (MDDNR, no
date).
2.1. Equipment
This section describes equipment needed for collecting air and water temperature data and
discusses considerations that influence equipment selection.
2.1.1. Basic components
The following basic components are needed to collect and access continuous temperature
measurements:
• A temperature sensor
• A data offload device that is compatible with the model of the sensor
• A computer with software that is compatible with the data offload device
• A radiation shield to prevent direct solar radiation from hitting the sensor (this can also
serve as a protective housing)
Additional equipment is needed to install the temperature sensors. This is discussed in more
detail in Section 2.3.2.
2.1.2. Considerations when choosing temperature sensors
When selecting temperature sensors, consider factors such as durability, accuracy, resolution,
measurement range, memory, and battery life. For RMN sites, select sensors that are durable
(able to withstand years of use in challenging conditions), have a minimum accuracy of ±0.5°C,
capture the full range of expected temperatures, have a memory that is sufficient to record
measurements at 30-minute intervals during the deployment period, and have adequate battery
life (Table 1). Water temperature sensors should be waterproof (e.g., IP code level 8). Sensors
are available that can measure water temperature with accuracies of ±0.2°C. If resources permit,
consider using these more accurate water temperature sensors at RMN sites.
-------
Definitions of accuracy, precision,
resolution, and bias
Accuracy refers to how close the temperature
measurement is to its "true" value, or its
"correctness." Precision is the degree to which
repeated measurements produce the same
results, assuming conditions are unchanged
(also referred to as variance or "tightness").
Resolution is the smallest change that the sensor
can detect, or the "fineness" to which the sensor
can be read. Bias refers to whether a systematic
offset occurs between the measured value and
the "true" value.
The lifespan of the sensor should be another
consideration. Some sensors are made to last
five years or longer before their batteries
expire or their cases start to leak. If sensors
with non-replaceable batteries are used,
document the sensor's use so you know when
to remove them from circulation and
budget/plan for their replacement. Perform
accuracy checks (Sections 2.2.1 and 2.6.1) to
ensure that sensors meet the specifications
quoted by the manufacturers. If these steps are
taken, any errors associated with equipment
differences will be extremely small compared
to the temperature variability within and
among sites.
Table 1. Specifications for temperature sensors used in the RMNs
Characteristic
Submersible/waterproof
Programmable start time and date
Minimum accuracy2
Resolution4
Measurement range - able to capture
the full range of expected temperatures
Memory
Battery life
Water sensor
yes1
yes
±0.5°C3
<0.5°C
-5 to 37° C will
typically work
Air sensor
optional
yes
±0.5°C
<0.5°C
depending on the location,
-20 to SOT might be necessary
(a typically available range)
Sufficient to record measurements at 30-minute intervals
during deployment period
Sufficient to remain active during deployment period
Sometimes sensors that are not waterproof are used to measure water temperature. This is done by housing them in
waterproof, non-drilled PVC canisters. However, laboratory trials suggest a time lag between changes in water
temperature and air temperature within a canister (Dunham et al. 2005).
2Accuracy varies depending on temperature range; make sure the sensor can accurately record measurements over
the temperature range that you expect the sensor to commonly experience.
3Water temperature sensors with accuracies of ±0.2°C are currently available. Their use at RMN sites is encouraged.
Resolution is the smallest detectable increment that the sensor can measure; it needs to be less than the accuracy.
If the pressure transducer that I am purchasing can measure water temperature, do I need to
purchase a separate temperature sensor?
It depends. If the temperature sensor in the transducer meets the minimum accuracy of ±0.5°C and the
sensor placement conditions described in Section 2.3.1 are met, deploying a separate temperature
sensor is unnecessary. Having a separate temperature sensor, however, does have some advantages.
For one, it can serve as a back-up if the transducer unexpectedly fails. Data from the temperature
sensor can also be used for quality control checks on the transducer data.
-------
The same sensors that are used for measuring water temperature can be used to measure air
temperature provided they capture the full range of air temperatures that are expected to occur at
a site (depending on the location, a range of -20°C to 50°C might be necessary). Less expensive,
non-waterproof sensors can also be used to measure air temperature, provided the sensors are
protected from the elements (see Table 2). Radiation shields, which are discussed in Section
2.1.4, can serve this purpose.
Examples of some commercially available temperature sensors are found in Table 2, and pictures
of these sensors are shown in Figure 1.
A)
B)
C)
F)
Figure 1. Numerous temperature sensors are commercially available. Examples include: A)
Onset Hobo© Water Temp Pro v2; B) Onset Tidbit© v2; C) Gemini Tinytag Aquatic 2; D)
Thermoworks LogTag; E) MadgeTech Temp 101A; and F) Maxim Integrated Products
Thermochron ibutton.
If different sites have different makes and models of temperature sensors, how much variability
will this introduce?
Not much, provided the proper accuracy checks are performed to ensure that sensors meet the
specifications quoted by the manufacturers. Any errors associated with equipment differences are
extremely small compared to the temperature variability within and among sites.
2.1.3. Considerations when choosing portable data offload devices
A data offload device is typically a pocket-sized device called a base station that is sold by the
manufacturer of a sensor. For example, the Onset sensors use a coupler specific to the model of
the sensor, which must be attached to the sensor before the sensor can be connected to the base
station. The base station is then connected to a computer via a cable (Figure 2). To view the data,
the appropriate software must be installed on the computer.
-------
Table 2. Examples of commercially available temperature sensors
Manufacturer (website)
Onset (onsetcomp.com)
Gemini
(geminidataloggers.com)
Maxim Integrated
Products
(maximintegrated.com)
Thermoworks
(thermo works . com)
Madge Tech
(madgetech.com)
Sensor model
Hobo© Water
Temp Pro v2
(U22-001)
TidbiT© v2 Temp
Sensor (UTBI-
001)
Hobo© U20 Water
Level Logger
(U20-001-04)
Tinytag Aquatic 2
(TG-4100)
Thermochroni-
button(DS1922L)
Log Tag
Temp 101 A
Water-
proof
yes
yes
yes
yes
no
no
no
Temperature range
-40° to 70°C (air);
maximum sustained
water = 50°C
-20° to 70°C (air);
maximum sustained
water = 30°C
-20° to 50°C (air)
-40° to 70°C
-40° to 85°C
-40° to 85°C
-40° to 80°C
Accuracy1
0.2°C from
0° to 50°C
0.2°C from
0° to 50°C
0.44 from 0°
to 50°C
0.5°C from
0°C to 50°C
0.5°C from -
10°Cto
+65 °C (with
software
correction)
0.5°C from -
20 to 40°C
0.5°C
Resolution1
0.02°C at
25°C
0.02°C at
25°C
0.10°C at
25°C
0.01°C
0.5°C
(8-Bit) or
0.06°C (11-
Bit)
<0.1°C
0.01°C
Battery life
(typical use)
and
replaceability
6 years,
factory-
replaceable
5 years, non-
replaceable
5 years,
factory-
replaceable
1 year, user-
replaceable
4 years, non-
replaceable
2 to 3 years,
technician-
replaceable
10 years, user-
replaceable
Approximate
price2 ($)
123
133
495
170
35
35
89
Accuracy and resolution over additional temperature ranges can be found on the manufacturer specification sheets.
2As of January 2013 and subject to change; reduced prices might be available for bulk orders.
-------
Figure 2. Additional equipment is needed to offload data from the temperature sensors onto
a computer. This figure depicts the coupler, base station, and USB cable needed to download
data from the Onset Tidbit© v2 sensor (www.onsetcomp.com). A waterproof shuttle can also
be used in place of the base station for on-site data offload and temporary data storage.
These devices must be compatible with the make and model of the sensor.
Some manufacturers make small, portable waterproof devices (often referred to as shuttles) that
can offload data while the sensor remains in the stream. These devices, which can be used to
temporarily store the data and can also serve as base stations, are more expensive than the non-
waterproof base stations. At RMN sites, instream data offloads are generally not possible
because sensors are housed in radiation shields (Section 2.1.4) and the sensors must be removed
from these shields before they can be attached to the data offload device. Thus, the main benefits
of the waterproof shuttles are that, after sensors are removed from the stream, field personnel can
work with them in inclement weather, and they are easy to carry in and out of remote sites where
bringing a laptop is impractical.
Consider compatibility when purchasing equipment. Ensure that the data offload device and
software are compatible with the model of the temperature sensor. If purchasing multiple
sensors, buying the same model is often the most cost-effective because reduced prices might be
available for bulk orders and only one data offload device and one software package are
necessary for that particular model of sensor. The non-waterproof data offload device/base
station for the Onset Tidbit© v2 Temp Sensor costs approximately $118, software costs $89-99,
and the waterproof shuttle costs $237 (these represent approximate prices on the Onset website
[http://www.onsetcomp.com/] as of June 2013).
2.1.4. Considerations when choosing radiation shields
Temperature sensors should be outfitted with radiation shields so that sunlight striking the
sensors will not bias temperature readings (Figure 3) (Dunham et al. 2005, Isaak and Horan
2011). These shields can also serve as protective housings and provide secure attachment points
for the temperature sensors. Radiation shields can be purchased from a manufacturer or
constructed less expensively from materials purchased at a local hardware store.
-------
50 100 150 200 250
30 minute interval
300
350
400
Figure 3. Stream temperature measurements from four sensors at the same site during eight
days in July 2010. All sensors had solar shields during the first four days when temperature
measurements overlap. The solar shield was removed from one sensor on day 5 (black arrow)
and temperature spikes became apparent when sunlight struck the sensor (reproduced from
Isaakand Horan 2011).
As an example, a polyvinyl chloride (PVC)
canister, which is simple and inexpensive to
construct, can be used to shield the water
temperature sensors. The one shown in Figure 4
consists of two pieces, including a flat, solid
bottom and a screw-top cap. Several holes are
drilled into the canister to make it neutrally
buoyant and to facilitate water circulation. The
threads are wrapped with Teflon tape, and then
a small piece of neoprene is zip-tied to the
temperature sensor to hold it inside the cap
(Isaak et al. 2013). If using the underwater
epoxy installation method (described in Section
2.3.2), the PVC canister must have a lip at the
bottom around which the epoxy can be wrapped
or else the epoxy will not hold the PVC to the
rock. More details on how to construct PVC
canister shields, including specific part
numbers, can be found in Isaak et al. (2013). An
alternative method for constructing PVC
canisters is described in Appendix A (Mauger
2008).
Does the color of the radiation shield
matter?
Clear shields have been shown to cause
erroneously high readings (Dunham et al.
2005). For opaque shields, there is no
evidence that color affects temperature
readings, likely because the shield comes in
minimal contact with the sensor and
flowing water removes effects of heat
conduction.
Color does, however, affect visibility.
White shields facilitate sensor retrieval
because they are easily seen but their
increased visibility could also increase
vandalism rates, especially in heavily
trafficked areas. If vandalism is a concern,
use radiation shields that are neutral in color
or paint them to reduce their visibility.
-------
Figure 4. Unassembled components of an inexpensive PVC canister radiation shield for water
temperature sensors (described in more detail in Isaak et al. 2013).
For air temperature sensors, the most effective radiation shields are mechanically aspirated, with
a small fan located within the shield that maintains air flow through the shield in low wind
conditions. Because these devices require power, passive/non-aspirated designs are more suitable
for remote deployment.
A wide range of passive radiation shields are commercially available, and their effectiveness
varies with the type or brand of funnel. The standard shield used in most weather/climate stations
is called a Gill radiation shield (Gill 1979). This shield is a series of plates that reflect incoming
radiation and passively radiate accumulated heat via conduction (see Figure 5A). In areas with
persistent winter and spring snowpack, the radiation shield should block incident radiation not
just from above, but also from below, because radiation reflecting off the snow can directly
strike sensors and bias temperature readings (Holden et al. 2013).
Commercially available, passive shields (e.g., Decagon Devices Inc., Campell Scientific, Onset
Inc.) can be purchased for approximately $50-80 each. The custom-made version developed by
Zachary Holden (Appendix B; Figure 5B) costs approximately $2.50-3.00 per shield. Holden
tested his design and found that it performed well compared to commercially available shields
(Holden et al. 2013). A YouTube video (accessed 15 June 2013) with instructions on how to
construct the custom-made shield can be found at: www.youtube.com/watch?v=LkVmJRsw5vs.
Other designs for air temperature shields are also available. For example, some entities use a
modified version of the PVC canister shown in Figure 4. The modified version differs in that it
has larger, more numerous holes. When selecting a design, consider whether the performance of
the shield has been tested and documented (e.g., compared against "known" validation
temperature records to determine error rates).
2.2. Pre-Deployment
2.2.1. Accuracy check
Before deploying the temperature sensors in the field, perform an accuracy check to verify that
the sensors meet the accuracy quoted by the manufacturer. Also check the battery life and make
sure the sensors are launching and downloading data properly. Temperature sensors with low
battery levels should be removed from circulation.
-------
A)
B)
Figure 5. Examples of radiation shields for air temperature sensors include A) the Gill-style
Onset RSI solar radiation shield (www.onsetcomp.com) and B) custom design by Zachary
Holden (Appendix B).
Perform either a single- or multi-point accuracy check. A simple and effective procedure for
performing single-point accuracy checks on water temperature sensors is the "ice bucket"
method, in which sensors are placed in an ice bath for several hours and sensor readings are
checked against readings from a National Institute of Standards and Technology (NIST)-certified
thermometer. Table 3 contains step-by-step procedures for the "ice bucket" method that the
Maryland Department of Natural Resources (MDDNR, no date) uses. Similar techniques are
described in Onset Computer Corporation (2014) and Dunham et al. (2005).
Table 3. The "ice bucket" method can be used for single-point pre-deployment accuracy
checks (MDDNR, no date)
Task
Procedure
Setup
1. Prepare an ice bath in a large cooler (e.g., 48 quart). Fill the cooler half full with water
and add 10 Ibs. of ice. Be sure to use fresh water (dissolved minerals can alter the
thermal properties of water). Place an air stone in the bottom of the ice bath to ensure
mixing of the water.
2. Allow the ice bath to sit for 2 hours to allow water temperature to stabilize.
3. Measure the temperature of the ice bath using a NIST-certified thermometer. The
temperature should read 0°C.
4. Prepare the sensors by connecting them to a computer and programming them to record
at 2-minute intervals.
Test
5. Put sensors in the cooler (sensor side down). The sensors should be fully submerged.
6. Leave the sensors in the ice bath for a minimum of 1 hour.
7. Remove the sensors from the cooler.
Accuracy
check
8. Download the data from each sensor and plot the measurements
9. The final sensor measurements should be within ±0.5°C of the NIST-certified
thermometer readings. If the sensors are outside that range, separate them out for further
testing or return them to the manufacturer.
10. Record the data on an accuracy check datasheet (see example in Appendix C).
10
-------
The multi-point check, which is described in Table 4, is more rigorous. First, sensors are set up
to record at the same sampling intervals that will be used in the field (e.g., every 30 minutes at
RMN sites). Next, the sensors are exposed to alternating warm and cold cycles that approximate
the temperatures and duration of diurnal fluxes to which the sensors will be exposed in the field.
At various points during this process, temperature measurements are taken with a NIST-certified
thermometer. After the sensors complete these cycles, the data are downloaded onto a computer
and values from the sensors are checked against: 1) readings that are taken with the NIST-
certified thermometer; and 2) mean values obtained from the other sensors (any one sensor
recording a value far from the other sensors is likely inaccurate, according to Sowder and Steel
2012). The mean is used for the calibration accuracy check because it is based on multiple
measurements and is generally more stable than single-point measurements like the minimum
and maximum. The same procedures can be used for waterproof and non-waterproof sensors,
except the non-waterproof sensors should not be placed in water baths.
Table 4. Multi-point pre-deployment accuracy check
Task
Procedure
Setup
1. Connect sensor to computer.
2. Check battery health.
3. Program sensor to record at 30-minute intervals.
Test bath
4. Put sensors in an open container/s; for the water temperature sensors, fill container with
enough water to fully submerge the sensors.
5. Put the container/s in a room that is at room temperature (near 20°C) for at least 4 hours
(the objective is to have the temperatures to equilibrate).
6. As close as possible to a time when the sensor is recording a measurement, gently mix
the water in the container and measure the water temperature with a NIST-certified
thermometer; record the value on an accuracy check datasheet (see sample form in
Appendix C).
7. Put the container with the sensors in the refrigerator (near 0°C) for at least 4 hours.
8. Remove the container from the refrigerator as close as possible to a time when the sensor
is recording a measurement; gently mix the water and measure the water temperature
with a NIST thermometer; record the value on a datasheet (Appendix C).
9. Repeat steps 5 through 8 several times to create multiple warming/cooling cycles.
10. Remove the sensors from the container/s.
Accuracy
check
11. Download the data from the sensors onto a computer as soon as possible so that the
sensor can be shut off to conserve battery life.
12. Calculate the overall average temperature of each individual sensor for the entire
calibration period as well as the maximums and minimums of each temperature cycle.
Compare the mean temperature value to the group average. For the dates and times
when measurements overlapped, compare the sensor temperature values to the NIST
thermometer values. Calculate the average difference between these values. It should not
exceed the accuracy quoted by the manufacturer of the temperature sensor, which should
be ±0.5°C or, in some cases, ±0.2°C for sensors deployed at RMN sites. Sensors that
have anomalous readings should be set aside and returned to the manufacturer for
replacement.
13. Record all accuracy check information on a data sheet (see sample form in Appendix C).
11
-------
Whichever technique is used, differences in readings from the sensors and NIST-certified
thermometers should not exceed the accuracy quoted by the manufacturer. For sensors deployed
at RMN sites, that number is ±0.5°C or, in some cases, ±0.2°C. Sensors that have anomalous
readings should be set aside for further testing or returned to the manufacturer for replacement.
All accuracy check data should be recorded on a datasheet like the one shown in Appendix C.
These records should be properly archived and stored in a place that can be easily accessed for
future reference (these data will eventually be used for QA/QC procedures, as discussed in
Section 2.6.1). Equipment needs for this procedure are listed in Table 5.
Table 5. Equipment needs for the pre-deployment accuracy check
Task
Performing the
calibration
procedure
Recording
measurements
Supply list
• Temperature sensor/s
• National Institute of Standards and Technology (NIST)
reference thermometer with an accuracy of ±0.2°C
• Field (e.g., red liquid) thermometers (optional)
• Containers to hold the sensors
• Water
• Refrigerator
• Clock or watch
• Accuracy check data sheet (Appendix C)
• Computer that has the appropriate software for reading
traceable or calibrated
the temperature sensor
2.2.2. Sensor configuration and launch
Sensors can be programmed to start before or after the planned deployment time, or can be
launched on site. Whatever the launch time, the exact time the sensor is correctly positioned
should be recorded so that observations recorded before and after that time can later be removed
during data processing (Section 2.6.2). Configure the sensors to record point temperature
measurements in degrees Celsius at intervals of 30 minutes. If the sensor has insufficient
memory capacity to record at 30-minute intervals for the deployment period, program the sensors
to record point measurements every 60 or 90 minutes. Figure 6 suggests that 60-minute intervals
are likely sufficient, although 30-minute intervals might more closely match other continuous
data collection efforts, such as the hydrologic data being collected by the pressure transducers.
The following practices will make data processing and screening easier and more efficient:
• Set the sensors up so that they start recording on the hour (xx:00) or half hour (xx:30)
• Set the units to degrees Celsius
• Set the air and water temperature sensors up so that they record at the same time
• Consider using military time (if this is an option) to avoid potential confusion with
a.m./p.m.
12
-------
Consider using local standard time (e.g., UTC-5 for sites in the Eastern Time zone)
instead of daylight savings time. Regardless of which you choose, when doing accuracy
checks, make sure any discrete measurements taken are consistent with this setting.
Considerations when choosing a sampling interval
Choosing a sampling interval is a balancing act. If long intervals are used, you could miss the
maximum and minimum daily temperatures because they might occur only briefly within a day.
Missing them would underestimate the warmest temperatures and overestimate the coldest
temperatures. If too short an interval is used, too much memory could be used, requiring more site
visits, data management, storage, and cleaning.
Dunham et al. (2005) compared the amount of bias in daily summary metrics that is incurred by using
different recording intervals. As shown in Figure 6, until intervals longer than 2 hours are used, bias
generally is not an issue. Sites with larger diel fluctuations have a greater probability of missing the
true maximum than those with smaller diel fluctuations.
At RMN sites, the 30-minute interval will be reevaluated after the first or second year of data collection
to determine whether a longer interval (i.e., 60 or 90 minutes) can adequately capture the thermal
regimes at sites.
0.20
8 10 12 14 16 18 20
Daily Range of Temperature (°C)
Figure 6. Probability of underestimating the maximum daily temperature at least 1°C in
relation to daily range of temperature and sampling interval (Dunham et al. 2005).
13
-------
Record the configuration and launch information for each sensor on a datasheet, similar to the
one shown in Appendix D. Include serial number, programmed deployment time, and recording
interval. If the sensor has "sensor high, low, and multiple sampling" features and "wrap-around-
when-full, overwrite oldest data" functions, turn these functions off.
2.3. Deployment of Water Temperature Sensors
Deploy at least one water temperature sensor at each RMN site. Place sensors in locations that
can be safely accessed and waded by crew members during different flow conditions. If possible,
avoid reaches with very high gradients (>7%) because sensor retention rates are inversely related
with slope (Isaak et al. 2013). Prior to installing the sensors, obtain permission from all relevant
parties (e.g., property owners if site access will involve crossing private property, the local
Department of Public Works if the water temperature sensor will be attached to a bridge
abutment). Try to minimize site impacts.
This document describes techniques that can be used to attach sensors to natural objects in
streambeds at remote sites, and also covers installations in which sensors are attached to built
structures like bridges. Provided the sensor is collecting data that are representative of the
characteristics of the reach from which the biological data are being collected, the sensor need
not be sited in the exact location from which biota are being collected.
2.3.1. Guidelines for placement within the stream
No simple, straightforward formula for selecting a location exists because each site is different.
In general, however, areas of well-mixed water moving through runs and pools are preferable
over riffles. Deploy sensors in locations with as many of the following characteristics as
possible, prioritized in this order:
Representative of the characteristics of the
reach from which the biological data were
collected (e.g., not below additional warm
or cold water sources)
Well-mixed horizontally and vertically
Of sufficient depth to keep the sensor
submerged year round
Stable, accessible, and easy to relocate
Protected from physical impacts
associated with high flow events (e.g., the
downstream side of a large landmark rock
or log)
Low human activity to reduce vandalism
and accidental snagging
How do I know if the water is well-
mixed?
In general, any moving water will be well
mixed. To verify this, take numerous
instantaneous temperature measurements
near the deployment location. If the stream
can be easily waded, a simple cross-
sectional temperature survey, consisting of
at least 10 measurements, can be done. If
crews have access to a multi-probe meter, it
is helpful to measure dissolved oxygen and
conductivity as well, because variability in
these measures could indicate sources of
thermal variation (Dunham 2005). If these
measures exhibit a high degree of
variability, consider moving to a different
deployment location.
14
-------
Do not select locations that:
• Are areas of high use, visibility, or fishing access
• Have heavy beaver activity
• Have backwater pools, eddies, or standing water that might stratify during low flow
conditions
• Are influenced by localized warm or cool water sources, such as
o a tributary confluence
o an impoundment (including beaver ponds)
o a lake outlet
o point-source discharges
o stream side wetland areas
o hot springs
o groundwater seeps
When possible, deploy sensors approximately 6 inches (<0.15 meters) above the stream bottom
(per Schuett-Hames et al. 1999). In some small, shallow streams you may have no choice but to
place a sensor near the stream bottom to ensure that it remains submerged during low flows. If
this happens, noting this on the field form is important because: 1) the temperature readings
could be influenced by groundwater and subsurface flow (Stanford and Ward 1988, Baxter and
Hauer 2000, Edwards 2001); and 2) sensors on or near the streambed are more susceptible to
burial by moving substrates (sensors should never be intentionally buried).
If conditions permit, install sensors on the downstream side of the structure to which the sensor
is being attached (e.g., a large rock or log). Doing so will help protect the sensor from high water
velocities and associated substrate movement and transport of debris that commonly damage or
dislodge sensors (Dunham et al. 2005). Ideally, the structure will also hide the sensor from
potential vandals.
2.3.2. Installation
If possible, install the sensor during low flow conditions to enable field personnel to check
whether the water is well-mixed and of sufficient depth year round. Flow conditions can be
determined by studying the hydrographs of local streams. In the next two sections, we briefly
describe two installation methods: the underwater epoxy method (Isaak and Horan 2011, Isaak et
al. 2013) and a method in which sensors are cabled to rebar or stable instream structures such as
large rocks or boulders, woody debris, or roots (Mauger 2008, Ward 2011). Equipment needs for
both techniques are listed in Appendix E. Site-specific conditions will dictate which installation
technique is most appropriate. The best technique ultimately will be the one that is easiest to use
and has the highest sensor retention rates over time. The pros and cons of each method are
discussed in Sections 2.3.2.1 and 2.3.2.2.
15
-------
2.3.2.1. Underwaterepoxy
The underwater epoxy method can be used in multiple environments provided a suitable anchor
point is available, such as a large rock or cement structure (e.g., bridge support). Table 6
describes the step-by-step procedure for completing underwater epoxy installations. The first
step is to select an appropriate rock or cement structure. The structure must have a relatively flat
downstream attachment surface and must be in water that is moving and deep enough to remain
submerged for the entire year. The best structures are those that not only remain immobile during
floods, but also are wide and protrude well above the low flow water surface to provide an
effective shield against moving rocks and debris. (Note: do not move rocks into the channel to
serve as attachment sites—if you can lift the rock, high flows will certainly dislodge it, not to
mention the safety risk associated with lifting large boulders). Ideally, on the downstream side of
the attachment point, pockets of relatively calm water with smaller substrate sizes will be present
(if large rocks and cobbles occur on the downstream side, similarly large substrates likely will
move there again during the next flood, and these could dislodge or break the sensor). Cement
bridge pilings at road crossings are also good attachment points for this protocol. Photos of good
attachment points are shown in Figure 7. Research has shown that heat conduction through
attachment structures like these does not bias temperature measurements if installations are done
in keeping with the protocols described below (Isaak and Horan 2011).
Table 6. Quick guide for doing underwater epoxy installations (Isaak et al. 2013)
Task
Procedure
Select
attachment
point
1. The attachment point should have the following features:
• Is protruding a foot or more above the water surface at low flows (rocks of this size
can be easily seen and relocated during subsequent field visits, and are large enough
to prevent other rocks from sliding over them and potentially hitting the sensors
during high flows)
• Is wide enough to protect sensor from moving rocks/debris during floods
• Has flat attachment site on downstream side and relatively deep water with flow
• Has small substrate on downstream side and 8 inches of space for shuttle attachment
Installation
2. Check sensor for blinking indicator light. Record sensor serial number and metal
forestry tag number on field form (Appendix D).
3. Put on gloves and use wire brush to clean surface of attachment site (approximately 6
inches above stream bed). Place a few cobbles or suitably sized rocks near the sensor
site to later lean against the attached sensor.
4. Moisten gloves and scoop out small amount (about quarter-size in diameter, 1/8 in
thick) of white and black epoxy from each container, and mix together for at least 1
minute. Apply epoxy to back of sensor (or PVC canister) and metal forestry tag.
5. Gently push and slightly twist sensor (or PVC solar shield holding the sensor) onto
attachment site.
6. Lean a rock against the face of PVC to hold it in place while the epoxy sets and check
attachment site with plastic viewing box.
Monument
7. Attach forestry tag on boulder directly above sensor (above water line).
8. Mark the site as a waypoint on GPS and record coordinates on data sheet
(Appendix D). Take several photos of site, and record photo numbers on data sheet.
16
-------
Figure 7. Examples of large rocks (a, b) and cement bridge pilings (c, d) that provide good
sensor attachment sites. Each site has a flat downstream attachment surface that is shielded
during floods from bedload and debris. Arrows point to the solar shield containing a sensor;
circles highlight metal forestry tags epoxied above the sensor to monument the site (photos
from Isaak et al. 2013).
After selecting an attachment point, use epoxy to attach the sensor to the structure, and lean a
rock against the face of the PVC canister to hold it in place while the epoxy sets (Figure 8). Use
of a specific underwater epoxy is critical to the success of this method. Isaak and Horan (2011)
tested several types of epoxy and found that only Fox FX-764 epoxy provided durable cement-
like attachments that worked well in field conditions. Installations with the Fox FX-764 epoxy
work well in a wide range of temperatures (2 to 20°C), but the epoxy becomes less cohesive in
water temperatures exceeding 20°C. Installations done in especially warm streams might need to
be done when water temperatures are relatively cool to allow the epoxy to set (Isaak et al. 2013).
Once the epoxy sets, it will hold the sensor across the full range of annual and daily water
temperatures.
If you are inexperienced with this technique, complete some practice runs in the laboratory first,
and then perform your first field installations at a few easily accessible locations that can be
checked after a few days. If a sensor needs to be reattached to a structure that has old Fox
bonding agent on it, using a different attachment point is best if possible.
17
-------
Figure 8. Close-up of a PVC solar shield with a
temperature sensor glued to a rock in a stream. The
rocks propped against the front of the solar shield
holds it in place while the epoxy sets during the first
24 hours.
Equipment needs for underwater epoxy
installations are summarized in Table 7
and Figure 9. Additional details on
underwater epoxy installations are
available in Isaak et al. (2013) and in a
YouTube training video (accessed 15
June 2013):
http ://www. youtube. com/watch?v=
vaYaycwfmXs&feature=youtu.be
The epoxy can also be used to attach a
metal forestry tag near the sensor
location. This can make the sensor
easier to relocate during subsequent
site visits (Figure 9). The numbers on
the tags can be used as unique site
identifiers. This is useful for data
organization, especially if different
sensors are used at a site over time
(Isaak and Koran 2013).
Table 7. Equipment list for doing underwater epoxy installations (Isaak et al. 2013)
Task
Supply list
Installation
Temperature sensor
Radiation shield (PVC canister, 1-1/2" with screw top, mid-section, base)
Underwater epoxy (FX-764 Splash Zone Epoxy)
lars for mixing the epoxy
Underwater viewing box
Lead weights, % oz
Neoprene, 3 mm
Rubber gloves
Plumber's tape
Wire brush
Zip ties, 4"
Metal mirror
• Metal forestry tags
• Spray paint
Monument
GPS
Camera
Data sheet (similar to Appendix D)
'Monumenting is discussed in Section 2.3.3.
18
-------
Fox Indus
FX-7&4 Sptasft Zone
A Cnrnponenl
Mur *quai parts 1A to
Caution: C
www Joxind co
Fox Indus
FX-764 Spla
B Component
Mix equal PHI"
C*otion: Contains
www foxirxj con
d
Figure 9. Equipment needed to permanently install temperature sensors in streams using
underwater epoxy. The equipment includes: (a) two-part FX-764 epoxy from Fox Industries,
(b) PVC solar shield, (c) temperature sensor, (d) cable ties, (e) plumber's tape, (f) rubber
gloves, (g) plastic viewing box, (h) wire brush, and (i) metal forestry tree tag (Isaak and Horan
2013).
Pros and cons of the underwater epoxy method are summarized below.
Pros:
• It requires minimal effort and materials.
• It provides durable installations that have been shown to withstand floods and associated
bedload movement (based on field trials, 80-90% of sensors installed correctly remained
in place after 1- and 2-year intervals [Isaak et al. 2013]).
• Retention rates have been formally tested and documented in a wide variety of streams
and are based on several years of field trials (Isaak et al. 2013).
• Once a sensor site is successfully established, it is easily maintained in years thereafter
simply by replacing sensors in PVC housings that remain in place.
19
-------
Cons
• Finding suitable attachment sites within appropriate macrohabitats (runs) might be
difficult, particularly in smaller streams that lack large rocky substrates.
• The technique is used primarily with Onset Tidbit© v2 sensors. The design would have
to be modified to accommodate larger sensors like the Onset Hobo© Water Temp Pro v2
sensors.
• Obtaining the epoxy might be difficult.
2.3.2.2. Cabling the sensor to rebar or stable instream structures
Sensors may be cabled to rebar or stable instream structures like large rocks or boulders, roots, or
woody debris. If site conditions permit, attach the sensors to the downstream side of the instream
structure, as this will shield the sensor from moving rocks or debris during floods. Use heavy-
duty (e.g., 120-lb tensile strength) cable ties or wire to attach the sensors to the structures. If
conditions permit, attach the sensor at two points, as shown in Figure 10. If you think the
structure might move during high flow events, consider cabling or chaining the structure to an
object on the nearest bank (or to another stable instream structure).
If a site lacks these types of stable
instream structures and the stream
bottom is such that a metal stake can
be driven into it (e.g., no near-surface
bedrock or consolidated sediments),
the rebar method can be used. An
appropriately sized piece of rebar
(generally 2-4 feet) is driven into the
streambed, deep enough to stay in
place during high stream flow events.
The sensor and its protective housing
are attached to the rebar via heavy-
duty cable ties or wire.
If possible, use rebar that is bent as
shown in Figure 11 (design provided
by Christopher M. Harbourt,
Principal Engineer, Waterborne
Environmental, Inc. Champaign, IL.
www.waterborne-env.com). If one
leg of the rebar is a little longer than
the other, the installations will be easier because you can start the longer leg first and then pound
the rest into place. Using bent rebar has several advantages. For one, it is more secure because
the metal bar is anchored in two places. A second benefit is that it poses less safety risk for
people wading through the stream because no sharp points protrude upward into the water
column (Personal communication, Christopher Harbourt, May 2, 2014).
Figure 10. If conditions permit, attach sensors in two
places, using heavy-duty (e.g., 120-lb tensile strength)
cable ties. This picture was provided by MDDNR
Monitoring and Non-Tidal Assessment.
20
-------
Figure 11. Using bent rebar for
rebar installations has several
advantages. It is more secure
because the metal bar is anchored
in two places. It also poses less
safety risk for people who are
wading through the stream because
no sharp points protrude upward
into the water column. This design
was provided by Christopher M.
Harbourt, Principal Engineer,
Waterborne Environmental, Inc.
Champaign, IL. www.waterborne-
env.com. Installations are easier if
one leg is a little longer than the
other because you can start the
longer leg first and then pound the
rest into place.
Various techniques can be used to attach the sensor and
its protective housing to the rebar. Some examples are
shown in Figure 12 and Figure 13. Appendix F contains
detailed descriptions of two additional installation
techniques. One describes the cable/rebar method that is
used in Cook Inlet salmon streams in Alaska (Mauger
2008), and the second is a technique the Delaware River
Basin Commission (DRBC) developed that uses a low
profile concrete base.
Figure 12. Example of how a cable and rebar
installation is configured for large streams in
Maryland. During the field installation, the cable is
tightened so that the sensor sits within the PVC
housing. Heavy-duty cable ties are then run
through two of the side holes in the PVC housing to
secure the sensor to the side of the PVC. This
picture was provided by MDDNR Monitoring and
Non-Tidal Assessment.
21
-------
Rebarand
metal hose
clamp
Figure 13. Example of a rebar installation
in a small stream in Shenandoah National
Park, VA. The sensor is located inside the
PVC housing, which has several holes
drilled into it to facilitate water
circulation. Two additional holes are
drilled into the PVC housing to serve as
attachment points for the rebar. The
housing slides over the rebar as shown in
the picture. The housing is then secured
to the rebar with a metal clamp. Hex
head screws are preferred over flat heads
because using a socket-end screwdriver is
easier if you are working under water,
especially if you are working by feel. In
larger streams where flooding is more of
a risk, consider using two metal clamps:
one on the top and one on the bottom.
That way, if the sensor becomes displaced, the housing will stay with the rebar, and an
underwater metal detector can be used to help find the sensor. In larger streams, the rebar is
driven deeper into the substrate than in the photo shown here. This lessens the chances of its
catching on logs and other debris during high flow events and also poses less safety risk for
people who are wading through the stream. This picture was provided by Craig Snyder, USGS
Survey (csnvder@usgs.gov).
Pros and cons of the cabling method are summarized below.
Pros:
• Provides more flexibility in placement location than the underwater epoxy method.
• If the rebar method is used and a high flow or other event occurs that causes the sensor to
become displaced, you could use a metal detector to relocate the sensor.
Cons
• For year-round deployments in streams that experience annual high flow events, sensors
might be more prone to being buried or dislodged by moving substrates than sensors
installed using the underwater epoxy technique.
• Retention rates are not as well documented as for the underwater epoxy method.
2.3.3. Documentation
One of the most common reasons for the loss of temperature sensors is failure to relocate the
sensor after initial field deployment (Dunham et al. 2005). Thus, accurately georeferencing each
sensor is critical as is documenting sensor placement in a way that enables field personnel to
22
-------
relocate the sensor during subsequent visits. Table 8 contains a list of guidelines for documenting
sensor locations. First, record global positioning system (GPS) coordinates (latitude and
longitude) for the exact site at which each sensor is deployed, as well as the datum of the GPS. If
you are unable to obtain GPS coordinates in the field, note the sensor location on an accurate
map and determine the coordinates later. Although GPS coordinates are useful for getting close
to the site, they are insufficient alone, so digital photos should also be taken at each site and
archived in a centralized database for future use. Photographs are important for relocating the
instruments, documenting any changes to the monitoring location during the course of the study,
and showing the near-stream habitat of the location where the sensor is deployed.
Table 8. Guidelines for documenting the installation
Task
Procedure
Initial deployment
• Use GPS to georeference the site.
• Take photographs from different perspectives; at least one photo should have
a visual marker pointing to the temperature sensor. Archive photos in a
central database for future use.
• Make detailed hand-drawn maps with landmark references (e.g., unique rock,
log, root, flagging, or tree), sensor locations, direction of stream flow, places
to park, paths to the stream, and any other characteristics that might be
appropriate, annotate photographs when viewing them later on a computer,
or both.
• Complete the field form.
• Take an instantaneous stream temperature at the location of the sensor with a
NIST-calibrated field thermometer. Note: this measurement should be taken
as close as possible to the time when the sensor will be recording a reading.
Take the photos from different perspectives (e.g., upstream and downstream). Include at least
one shot with a visual marker (e.g., someone pointing to the underwater location of the sensor).
Later, when viewing these photos on a computer, annotate them with notes about landmark
references (e.g., unique rock, log, root, flagging, tree), sensor locations, direction of stream flow,
places to park, paths to the stream, and whatever else might be appropriate. Detailed hand-drawn
maps like the one shown in Figure 14 (Ward 2011) are also helpful. Relocation success can also
be improved by using metal forestry tags to enhance the visibility of the structures to which the
sensors are attached (where suitable) (Figure 15). The use of signage at RMN sites is optional. If
your agency has specific policies regarding signage, follow those protocols. Be aware, however,
that using signs could increase the chance of vandalism.
23
-------
Continuous Temperature Survey Form
Station*: QQC IfD Station Name: (2^**- *£ /.4frttS&«£6- Samplers:
Interval Frequency 00:30
Water Temperature Logger
La* l£52jLZlL_
Wjtor Depth ) t S ft Deployment Depth I -^ ft
Height
-------
If the stream can be easily waded, take an instantaneous stream temperature measurement with a
NIST-calibrated thermometer at the location of the sensor. This measurement will be used for an
accuracy check (Section 2.6.1). To minimize the chance of a faulty measurement, take the
measurement as close as possible to the sensor and as close as possible to the time when the
sensor is recording a measurement. Make sure sufficient time has passed to allow the NIST-
calibrated field thermometer to stabilize before recording the measurement and note whether the
time is standard or daylight savings time. Equipment needs for taking these measurements and
for documenting sensors are summarized in Table 9.
Table 9. Equipment list for documenting sites
Task
Georeferencing and
monumenting
Measuring
temperature
Supply list
• GPS
• Camera
• Map or gazetteer, or both
• Metal forestry tags
• Field form (similar to Appendix D, Figure 3.3-4)
• NIST-calibrated field thermometer or multi-probe meter, or both
2.3.4. Common problems
As summarized in Dunham et al. (2005), the three most common problems that cause sensors to
be damaged or lost are:
• failure to relocate the temperature sensor after initial field deployment
• human tampering or vandalism; and
• natural disturbances, such as flooding, substrate movement, and animal influences (e.g.,
trampling by livestock or wildlife, beaver pond construction)
Tips for minimizing the chances of losing sensors are summarized in Table 10. If there are sites
where problems such as vandalism are expected to occur, consider deploying more than one
sensor so that one can serve as a back-up. If you are using metal cables, clamps or rebar for
installations and the sensor becomes displaced, consider using an underwater metal detector to
help relocate the sensor.
Does ice affect the sensors?
Not if the sensors are properly installed and remain submerged during winter. Ice in a stream is 0°C,
so it will not bias winter water temperature readings, which are also near 0°C when ice is present. If
the sensor is securely attached to the downstream side of a large boulder or other protective structure,
chunks of ice moving downstream will not affect it.
25
-------
Table 10. Tips for minimizing the chance of sensors being lost or damaged
Tips for minimizing chance of damage or loss
Problem
Failure to relocate sensor
after initial deployment
Follow the documentation steps described in Section 2.3.3, which include:
• Georeferencing all sensors
• Monumenting sites with metal forestry tags
• Taking photographs
• Creating detailed maps and notes
Disruption or vandalism
from humans or livestock
Do not put sensors in areas of high use, visibility, or fishing access
Camouflage the sensors (this is generally the least expensive option,
but could also make the data sensor more difficult to relocate)
Secure the sensor in locked and signed housing that is relatively
impervious to physical vandalism or disruption
Actively coordinate with ongoing research, monitoring, or management
efforts in the study area, not only to minimize duplication of
temperature sampling efforts, but also to minimize problems with
unintentional interference
Natural disturbances
Where possible, install sensors on the downstream side of a large
landmark rock or log, as this will protect the sensor from moving rocks
or debris during floods
Do not site sensors in areas with known beaver activity
House the sensor in a protective case
2.4. Deployment of Air Temperature Sensors
Air temperature plays an important role in controlling stream temperature (Hill et al. 2013). As
discussed in Section 2.1.4, unbiased air temperature measurements can be difficult to obtain. If
the sensor is located in an open area with full sun, unless a mechanically aspirated radiation
shield is used (which requires power to run a small fan), bias will be associated with the radiation
shield. If the sensor is located in a shady area near a stream, air temperature will be affected by
the distance from the water (the closer the sensor is to the stream, the more the water buffers the
air temperature because energy goes into evaporating stream water rather than warming the
ground and air).
The objective at RMN sites is to measure streamside temperature. The streamside air temperature
data will be used to gain a better understanding of the relationship between air and water
temperature at each RMN site. These data will provide important insights about the
responsiveness of stream temperatures to air temperatures and the differing vulnerabilities of
streams to thermal change. To capture this relationship effectively, the location of the air
temperature sensor in relation to the water temperature sensor must remain constant throughout
the period of data collection. Although air temperature has a strong effect on stream temperature,
many other controls, such as groundwater discharge, affect stream temperature. Although these
other controlling factors are of interest, due to resource limitations, only air temperature can be
consistently collected at RMN sites at this time.
26
-------
At least one air temperature sensor should be deployed at each RMN site. If the riparian zone is
forested (which is likely—most of the RMN sites are located in forested watersheds), mount the
air temperature sensor to the tree that is: 1) nearest the water temperature sensor, and 2) large
enough to support the radiation shield and sensor (ideally >12 inches (0.305 meters) in diameter).
Trees generally make the best attachment points because they provide stability, some degree of
shade, and can help hide the sensor from potential vandals. If a suitable tree is present, attach the
radiation shield and sensor to its north side, out of direct sunlight, using the simple four-step
process described in Appendix B. Outfitting the temperature sensors with radiation shields is
critical so that sunlight does not strike the sensor and bias the temperature readings (Dunham et
al. 2005, Isaak and Horan 2011) (Section 2.1.4). Do not rely on vegetation alone as the primary
radiation shield.
To be consistent with typical meteorological observations, air sensors should be placed at a
height of 2 meters, or approximately 6 feet, off the ground. Because trees, vegetation, and the
ground create radiation microenvironments, try to minimize the amount of other vegetation near
the sensor. If the air temperature sensor is not waterproof, ensure that the radiation shield
provides sufficient protection from the elements.
If trees or other suitable, stable existing structures (e.g., fence posts) are absent, mount the
radiation shield and air temperature sensor to a 10-foot piece of 1/2-inch diameter PVC pipe,
which can be purchased at a local building supply store. Using a drill, create a 1/8-inch hole in
the PVC pipe at a height of 2 meters above the ground. Next, insert a 12-inch piece of heavy-
gauge steel wire (a metal coat hanger works well) through the hole. This metal wire can be
inserted through one of the small corrugated tubes at the top of the solar radiation shield. Plastic
zip ties can then be used to stabilize and secure the solar radiation shield to the PVC pipe. Then,
pound a 3- to 4-foot piece of metal rebar into the ground to a depth of approximately 1-foot, and
slide the PVC pipe onto the rebar. This method has limitations. For one, it has only been used
experimentally and has not been extensively tested. Another is that sensors installed in the open
are more likely to be vandalized, and may also be more prone to destruction by animals.
Installations should be documented per the procedures outlined in Section 2.3.3. After
installation and during subsequent site visits, instantaneous air temperature measurements should
be taken with a NIST-calibrated thermometer at the location of the sensor, as close as possible to
the time of the expected sensor recording. These data will later be used to check the accuracy of
the sensor (Section 2.6.1).
If an air temperature sensor cannot be installed at a site, compile daily air temperature
observations from the nearest active weather station. Online resources like Utah State
University's Climate Database Server (http://climate.usurf.usu.edu/mapGUI/mapGUI.php) can
be used to locate the nearest weather station and to obtain data. How well the weather station
data approximate on-site conditions depends on factors such as distance between the site and
weather station and differences in topography and weather patterns. Local surface air
temperatures can vary substantially (5-10°C) from nearby weather stations in mountainous
terrain. This is particularly true in valleys, where temperature can become decoupled from the
free atmosphere as a result of cold air drainage and pooling (Holden et al. 2011).
27
-------
Obtaining data from the nearest weather station is of value even if an air temperature sensor is
installed at a site. The data could be used for QA/QC purposes (e.g., a comparison of the two
datasets that reveals differences in patterns could indicate that the on-site sensor is
malfunctioning). Or, if the comparison shows little difference between data sets, the weather
station data could be used in place of the data from the on-site sensor. This would free up a
temperature sensor for use elsewhere, which could be important if resources are limited.
2.5. Maintenance/Mid-Deployment Checks and Data Offload
After equipment is installed at a RMN site, try to revisit the site within the first month to confirm
that the installation is holding properly and that the sensor remains fully submerged in flowing
water. After these initial deployment checks, visit the sites as frequently as your schedule permits
to check the condition of the sensors, gather data for mid-deployment accuracy checks and
offload data. More frequent site visits will help ensure the longevity of the sensors as well as data
quality. At a minimum, try to visit the RMN sites annually, ideally during low flow conditions.
To gather data for the mid-deployment accuracy checks, collect instantaneous stream
temperature measurements near the sensor with a NIST-calibrated field thermometer. To
minimize the chance of a faulty measurement, take the measurement as close as possible to the
sensor and as close as possible to the time when the sensor is recording a measurement, make
sure sufficient time has passed to allow the NIST-calibrated field thermometer to stabilize before
recording the measurement and note on your field form whether the time is standard or daylight
savings time.
A biofouling check can also be performed but is optional at RMN sites. To perform this check,
remove the sensor and gently clean it (per manufacturer's instructions) to remove any biofilm or
sediment, then return the cleaned sensor to the stream. Note on your field form the time at which
the "pre-cleaning" measurement was made and the time of the first "post cleaning"
measurement. Compare the readings.
If possible, bring a laptop (with the appropriate software) and base station or waterproof shuttle
into the field so that the data can be offloaded onto the computer on-site. This allows field
personnel to:
• quickly screen the data for atypical results (if any readings are unusual, consider
replacing the sensor or moving it to a different location, e.g., perhaps the sensor is
coming out of the water during baseflow conditions)
• check the battery life on certain types of sensors (batteries in some sensors should be
replaced each year, while others, like the Onset TidbiT v2, last five years or longer under
normal use)
• back up the data (e.g., onto a flash drive) before clearing the sensor memory
Table 11 and Appendix G contain checklists for performing these maintenance/mid-deployment
checks and data downloads. Note things that could affect the quality of your data, such as signs
of physical damage, vandalism, or disturbance. Also ensure that the sensor is not buried by
sediment and remove anything that could foul the sensor and bias the temperature readings (e.g.,
28
-------
debris, aquatic vegetation, algae). Take photographs to document any changes to the monitoring
location during the course of the study.
Table 11. Checklists for performing maintenance/mid-deployment checks and data
downloads
Task
Maintenance/
mid-
deployment
checks
Data offload
Sensor retrieval
Procedure
• Check the security of the housing and deployment equipment and adjust if
necessary.
• Look for signs of physical damage, vandalism, or disturbance.
• Ensure that the sensor is submerged. If it is not, move it to a location where it is
covered by water and will remain so during periods of base-flow.
• Ensure the sensor is not buried in sediment. If it is, remove the sediment and
reinstall the sensor where it will not be buried during future high flow events.
Note on the datasheet that the sensor was buried because temperature recordings
are likely to be significantly biased toward cooler temperatures by hyporheic
flows. In many cases, temperature recordings from buried sensors should be
discarded because of high bias and because accurate adjustments are difficult to
apply.
• Remove anything that could bias the temperature readings (e.g., debris, aquatic
vegetation, algae). Note: if sensors have protective housings with fine screens or
small flow-through holes, they can be easily fouled in eutrophic systems with
abundant periphyton or algal growth.
• Take photos to document any changes to the monitoring location (particularly
those that could influence readings).
• Take instantaneous stream temperature measurements at the location of the sensor
with a NIST-calibrated field thermometer. Note: this measurement should be
taken as close as possible to the time when the sensor will be recording a reading.
• Record observations on the field form (like the one shown in Appendix G).
• OPTIONAL: Biofouling check. Remove the sensor and gently clean it (per
manufacturer's instructions) to remove any biofilm or sediment, then return the
cleaned sensor to the stream. Note on your field form the time the 'pre-cleaning'
measurement was made and the time of the first 'post cleaning' measurement.
Compare the readings.
• Before connecting the sensor to the data offload device, gently wipe the sensor
with a soft wet cloth or soft bristled brush to remove any biofilm or sediment that
could affect its ability to connect.
• Attach the sensor to a base station or shuttle and then connect the data offload
device to a computer with the appropriate software.
• Once the connection is established, follow manufacturer's downloading
procedures.
• Clear the sensor memory as necessary to ensure sufficient capacity for continued
deployment.
• If a sensor is removed from a site, before leaving the site, mark it with a
temporary tag identifying the site, date, and time of retrieval.
• Be sure to record the exact times of deployment (in proper position) and
recovery. This information is needed for trimming data after retrieval.
29
-------
Table 11. Checklists for performing maintenance/mid-deployment checks and data
downloads (continued)
Task
Procedure
Data offload
Before connecting the sensor to the data offload device, gently wipe the sensor
with a soft wet cloth or soft bristled brush to remove any biofilm or sediment that
could affect its ability to connect.
Attach the sensor to a base station or shuttle and then connect the data offload
device to a computer with the appropriate software.
Once the connection is established, follow manufacturer's downloading
procedures.
Clear the sensor memory as necessary to ensure sufficient capacity for continued
deployment.
Sensor retrieval
• If a sensor is removed from a site, before leaving the site, mark it with a temporary
tag identifying the site, date, and time of retrieval.
• Be sure to record the exact times of deployment (in proper position) and recovery.
This information is needed for trimming data after retrieval.
To gather data for the mid-deployment accuracy checks, collect instantaneous stream
temperature measurements near the sensor with a NIST-calibrated field thermometer. To
minimize the chance of a faulty measurement, take the measurement as close as possible to the
sensor and as close as possible to the time when the sensor is recording a measurement, make
sure sufficient time has passed to allow the NIST-calibrated field thermometer to stabilize before
recording the measurement and note on your field form whether the time is standard or daylight
savings time.
A biofouling check can also be performed but is optional at RMN sites. To perform this check,
remove the sensor and gently clean it (per manufacturer's instructions) to remove any biofilm or
sediment, then return the cleaned sensor to the stream. Note on your field form the time at which
the "pre-cleaning" measurement was made and the time of the first "post cleaning"
measurement. Compare the readings.
If possible, bring a laptop (with the appropriate software) and base station or waterproof shuttle
into the field so that the data can be offloaded onto the computer on-site. This allows field
personnel to:
• quickly screen the data for atypical results (if any readings are unusual, consider
replacing the sensor or moving it to a different location, e.g., perhaps the sensor is
coming out of the water during baseflow conditions)
• check the battery life on certain types of sensors (batteries in some sensors should be
replaced each year, while others, like the Onset TidbiT v2, last five years or longer under
normal use)
• back up the data (e.g., onto a flash drive) before clearing the sensor memory
30
-------
If bringing a laptop into the field is impractical, offload the data onto a portable data storage
device (like the Onset waterproof shuttle) and view it later when you have access to a computer.
Data should be transferred from the data offload device to a computer as soon as possible.
If the sensor is functioning properly and has sufficient battery life, it can be redeployed at the
site. If the sensor is removed from the site, before leaving the site, mark the sensor with a
temporary tag identifying the site, date, and time of retrieval, and replace it with another sensor
that has passed a pre-deployment accuracy check (Section 2.2.1). Accurately recording the time
and date of retrieval is critical, as this information will be used in the error screening described in
Section 2.7.2. When you replace a sensor, if the original one was not permanently attached to a
structure, put the sensor as close as possible to the original location to minimize potential sources
of variability in the long-term record. That comprehensive field notes be taken during every site
visit is essential. Sample field forms can be found in Appendix G. Equipment needs for
maintenance, mid-deployment checks, and data downloads are summarized in Table 12.
2.6. Quality Assurance and Control
Perform QA/QC procedures after data are offloaded to verify the quality of the data and to check
for potential errors. Document these procedures (e.g., on a form like the one shown in Appendix
H). Any forms and other documentation should be organized, easily accessible, and archived in a
way that allows for safe, long-term storage. The series of data cleaning steps described in Section
2.6.2 can improve data quality, reduce the time and effort of data processing, and increase
collaboration and comparability across projects, streams, and regions (Sowder and Steele 2012).
Table 12. Equipment list for conducting maintenance/mid-deployment checks and data
downloads
Task
Relocating sensor
Documenting on-site
conditions
Data offloads
Measuring temperature
Back-up equipment (in
case a sensor needs to be
replaced)
Supply list
• GPS
• Map, gazetteer, or both
• Annotated photos or hand-drawn map with landmark references (see
Section 2.3. 3)
• Camera
• Field form (similar to Appendix E)
• Base station or portable shuttle
• Laptop (if practical) and data back-up device (e.g., flash drive)
• NIST-calibrated field thermometer, multi-probe meter, or both
• Calibrated replacement sensors and other necessary deployment
equipment
-------
2.6.1. Mid- and post-deployment accuracy checks
After sensor data are downloaded, compare the instantaneous NIST-calibrated field thermometer
measurements from the mid-deployment checks to the corresponding sensor readings. The sensor
readings should not exceed the accuracy quoted by the manufacturer. If a sensor fails this check,
repeat the procedure. If it fails a second time, flag the data with an appropriate data qualifier, and
review the field notes to look for comments about situations that could cause the NIST-calibrated
field thermometer measurement to be faulty (e.g., perhaps the crew could not take the
measurement in close proximity to the temperature sensor due to high flow conditions, or used
daylight savings time instead of standard time when recording the time of the measurement). If
no issues with the NIST-calibrated field thermometer measurement are documented, consider
replacing the sensor. After the sensors are retrieved and brought back to the office or laboratory,
conduct a post-deployment accuracy check on each sensor, using one of the pre-deployment
accuracy check procedures described in Section 2.2.1.
2.6.2. Error screening
Erratic readings with temperature sensors can occur for a number of reasons, including:
• they could become dewatered during low flow conditions
• high flow events could bury them in sediment
• high flow events could move them
• they could become fouled from debris, aquatic vegetation, or algae
• humans could cause interference
Therefore, a standard set of procedures should be performed to ensure the quality of the data and
make necessary corrections. A series of error screening checks based on guidance from Dunham
et al. (2005) and Sowder and Steel (2012) is described in Table 13, and Appendix H contains a
data screening checklist. The first step involves removing observations recorded before and after
the sensor is correctly positioned. This can be done via a visual inspection of data and by
referencing field notes indicating the exact times of deployment and recovery. While reviewing
the field notes, also look for comments about situations that could cause the sensor to record
questionable readings (e.g., during a mid-deployment check, the sensor was found to be
dewatered or buried in the sand) and flag those data accordingly.
Next, plot all temperature measurements versus date/time and visually check the data. Look for
missing data and abnormalities, Consider plotting air and stream temperature on the same graph,
as well as stream temperature and stage (depth) data (if available). Specific errors to watch for
include missing data; a close correspondence between water and air temperature (this indicates
that the stream sensor might have been out of the water); and diel fluxes with flat tops (this
indicates that the sensor might have been buried in sediment). In addition, consider performing
the following optional checks: graphically compare data across sites; graphically compare data
across years (dramatically different data from one year indicates possible errors); graphically
compare with data from the nearest active weather station (if appropriate). If discrepancies occur,
use the plots to specify the time and duration of errors in the raw data files.
32
-------
Table 13. Error screening procedure (Dunham et al. 2005, Sowder and Steel 2012, Personal
communication, Michael Kashiwagi (MDDNR), April 17, 2014, Personal communication,
Dustin Shull (PADEP), February 2, 2014)
Task
Procedure
Remove pre- and
post-deployment
observations
Use field notes indicating the exact times of deployment and recovery to
remove observations recorded before and after the sensor is correctly
positioned in the stream channel.
Visual checks
Plot individual data points versus date/time to look for missing data and
abnormalities
Graphically compare stream to air temperature (if available); a close
correspondence between water and air temperature is a strong indication that
the stream sensor was out of the water
Optional: Graphically compare data across sites
Optional: Graphically compare data across years; when data from one year
are dramatically different, data errors might be present
Optional: Graphically compare with stage data (if available)
Automated checks
• Calculate upper and lower 5th percentiles of the data
• Flag data points for potential errors if they:
o Exceed a thermal maximum of 25 °C*
o Exceed a thermal minimum of -1 °C*
o Exceed a daily change of 10°C*
o Exceed the upper 5th percentile of the overall distribution
o Fall below the lower 5th percentile of the overall distribution
*These values should be adjusted to thermal limits appropriate for each location.
Additional checks are described in Table 13 and Appendix H. These include flagging data points
for potential errors if they fall outside expected thermal limits (e.g., if water temperature
measurements exceed a thermal maximum of 25°C or a daily change of 10°C). Thresholds
should be adjusted to thermal limits that are appropriate for each site.
Various software applications can be used to help with the error screening checks, data
management, and calculation of various summary statistics. The software purchased with the
temperature sensors is likely to have some applications like this. Other options include free
applications that can be downloaded from the U.S. Forest Service, Rocky Mountain Research
Station stream temperature website (2014)
(http://www.fs.fed.us/rm/boise/AWAE/proj ects/stream temperature.shtml), which include a SAS
temperature data processing macro and the Thermo Stat 3 software package (Jones and Schmidt
2012). The Washington State Department of Ecology also has developed the Freshwater
Monitoring Unit (FMU) Access® Data Logger Database, available upon request (Ward 2011).
Errors should be addressed on a case-by-case basis. In general, three possible actions can be
taken: 1) leave data as is; 2) apply correction factor; or 3) remove data. If you are inexperienced
at addressing errors with continuous temperature data, consider seeking guidance from someone
33
-------
with more experience and consult references like Wagner et al. (2006). Table 14 provides a
general summary of different types of problems that can occur (e.g., missing data, failed
accuracy check) and recommended actions for addressing them.
Table 14. General summary of different types of problems that can occur with continuous
temperature data and recommended actions for addressing them
Problem
Recommended action
Missing data
Leave blank
Water temperature sensor was
dewatered or buried in sediment for part
of the deployment period
Use the plot to determine the period during which the problem
occurred. Exclude these data from analyses.
Recorded values are off by a constant,
known amount (e.g., due to a
calibration error)
Adjust each recorded value by a single, constant value within
the correction period.
A large amount of drift is present. It is
unknown when and by how much the
sensor was "off (when drift occurs, the
difference between discrete
measurements and sensor readings
increases over time)
The data should be removed.
Discrepancy between sensor reading
and discrete measurement taken during
an accuracy or fouling check
General rules:
• If the errors are smaller than the sensor accuracy quoted
by the manufacturer and cannot be easily corrected (e.g.,
they are not off by a constant amount), leave the data as
is.
• If the sensor fails a mid-deployment accuracy check,
review field notes to see if any signs of disturbance or
fouling were noted, and also look for notes about the
quality of the QA/QC measurement (e.g., was the
measurement taken by a National Institute of Standards
and Technology [NIST] traceable or calibrated reference
thermometer? Did environmental conditions prevent the
measurement from being taken in close proximity to the
sensor?). Also check whether the same time setting was
used for both the sensor and discrete measurements
(daylight savings time vs. standard time). Based on this
information, use your best judgment to decide which
action is most appropriate (leave as is, apply correction, or
remove).
• If a sensor fails a post-retrieval accuracy check, repeat the
procedure. If it fails a second time, use your best
judgment to decide which action (leave as is, apply
correction, or remove) is most appropriate.
34
-------
Corrections should not be made unless the cause(s) of error(s) can be validated or explained in
the field notes or by comparison with information from nearby stations. Accurate field notes and
accuracy check logs are essential in the data correction process. Any discrepancies should be
documented in your data file and any actions you take should also be carefully documented.
2.6.3. Record keeping
Even if the set of standardized procedures described in Section 2.6.2 is followed, the data
cleaning process is inherently subjective, so both the original and the cleaned data files should be
maintained and archived. Large amounts of stream temperature data will accumulate quickly so a
central database should be developed and maintained from the initial stages of monitoring. Any
changes you make to the data should be carefully documented, and all forms should be
organized, easily accessible, and archived in a way that allows for safe, long-term storage.
3. HYDROLOGY
The protocols in this section describe how to install and maintain pressure transducers and staff
gages to measure continuous stage (depth). If additional measurements are taken, stage data can
be converted to discharge (flow). Section 3.8 contains a brief description of how to collect the
discharge measurements that are used to develop stage-discharge curves. Protocols and methods
discussed in the following sections are based on those used by Massachusetts Division of
Ecological Restorations River Instream Flow Stewards Program (RIFLS) (Chase 2005, Division
of Ecological Restoration 2010), the USGS (Rantz et al. 1982), and the Washington Department
of Ecology (Shedd 2011, Shedd and Springer 2012).
3.1. Equipment
This section describes equipment used to collect stream stage data, using a combination of
sensors for making continuous measurements and graduated staff gages for making discrete field
measurements. Many different types of sensors can provide continuous monitoring and logging
of stream stage. In this document, we focus on pressure transducers because most entities will be
using these at RMN sites.
3.1.1. Basic components
Two basic components are required to measure stream stage over time:
• A submersible pressure transducer, which provides continuous monitoring and logging of
water level (Section 3.1.4)
• A staff gage, which is a graduated measuring instrument from which stage can be
visually read (Section 3.1.3)
Pressure transducers record absolute pressure (air pressure + water pressure), from which air
pressure must be subtracted to yield water pressure. Software programs can then be used to
convert water pressure to water level using the density of water. Because atmospheric pressure
changes with weather and altitude, compensating for barometric variations is necessary; failure
to account for these variations could result in errors of 0.6 m (2 ft) or more.
35
-------
Adjustments can be made in two ways: 1) if the transducer is vented, it has a vented cable that
references and automatically corrects for atmospheric pressure; or 2) if the transducer is non-
vented, barometric pressure readings must be obtained from a separate device that is located
nearby on land, and after both sets of data are downloaded, software is used to correct the water
level data for the barometric variations. Vented and non-vented transducers are discussed in
more detail in Section 3.1.2.
To access the data, a data offload device is required. This device can be a cable supplied with the
transducer that connects directly from the data logger to a computer or a data shuttle that is
purchased separately. The data shuttle used for the Onset temperature sensors (shown in Section
2.1.3 and Figure 2) can also be used with Onset pressure transducers by attaching a different
coupler. Additional software, specific to the pressure transducer, is necessary to view the data.
Additional equipment needed for installation can be found in Section 3.3 and Appendix I.
3.1.2. Considerations when choosing pressure transducers
Pressure transducers are of two main types:
• Vented pressure transducers, which collect and automatically correct data for
barometric pressure. These typically have transducers that are connected via vented
cables to data loggers that are installed on land.
o Pros:
The transducer does not need to be removed from the stream to download data (data
are downloaded directly from the on-land logger). Data downloads are quick and
convenient. You do not risk accidentally putting the transducer back at a different
elevation, which would necessitate the collection of new correction data.
• Data are automatically corrected for barometric pressure, allowing for an
immediate comparison of transducer data to gage data. This facilitates immediate
detection and troubleshooting of any data quality issues in the field.
o Cons:
• Maintenance of the vented cable can include changing desiccant and ensuring the
cable is not damaged (e.g., ensuring animals did not chew cable). Extra
precautions must be taken to ensure ice does not crimp vented cable in streams
with ice cover.
• Cable lengths are fixed, meaning that cables might not be long enough to allow
the logger to be placed in a convenient location or might be so long that long
loops must be stored in unwieldy coils.
• The transducer can be subject to higher rates of vandalism compared to non-
vented transducers due to visibility of the on-land data sensor and cable.
36
-------
• Non-vented pressure transducers, which do not automatically correct data for
barometric pressure. This type of transducer typically has an internal data logger.
o Pros:
• Data loggers are internal and the transducer can be easier to hide in the stream
than a vented transducer.
• There is no vented cable to maintain. These transducers might be easier to use in
streams with extended ice cover due to the lack of cable.
o Cons:
• A second, identical pressure transducer must be installed on land to collect
barometric pressure for the correction.
• Data must be corrected for barometric pressure post-download using the
manufacturer-provided software. Data cannot be viewed in real time and unless
data correction is done in the field, stage data cannot be viewed immediately.
• Data loggers are typically located inside the transducer and therefore must be
removed from the stream to download data. If the transducer is not returned to the
same location and elevation, new correction data will need to be collected.
Transducers at RMN sites should meet the
following specifications:
• Accuracy of < 0.015 ft Accuracy is
a function of the operational range of
the pressure transducer. As the range
increases, the accuracy decreases.
Manufacturers specify a full-scale
accuracy for a given transducer;
multiply this by the expected range of
depths to be experienced by the
transducer to assess whether the
equipment will provide an
appropriate level of accuracy. For
example, a transducer with a full
scale accuracy of ±0.1% should not
be used over a range of depths
greater than 15 feet.
• Range to encompass the maximum
expected range of stream stages for
that location. A site visit prior to
purchasing the transducer will help
determine the expected range of
flows. Because accuracy decreases as
range increases, choose the smallest
possible range. A depth range of less
than 15 feet should be suitable at most
If I am using a non-vented transducer, is it ok
to use barometric pressure data from the
nearest active weather station instead of
deploying an on-land transducer?
Barometric pressure can vary over very small
distances due to factors such as elevation and
topography, so this should be evaluated on a site-
by-site basis. The best way to evaluate how well
the weather station data approximate on-site
conditions is by collecting on-land transducer data
for a year and comparing those data to the weather
station data. If the two datasets are closely
matched, the on-land transducer could be removed
from the site. Additional considerations should be
taken into account as well (e.g., Is the weather
station expected to remain in operation during the
period of deployment? What quality assurance and
control measures are performed on the weather
station data? At what rate are data collected, and
do they need to be matched or interpolated to
match with the in-stream transducer data?).
Whichever data source is used, it is critical that the
barometric pressure readings accurately represent
on-site conditions because failure to account for
pressure variations will result in erroneous water
level measurements.
RMN sites.
37
-------
• Sufficient cable length to meet installation requirements (vented transducers only).
Some transducers come with a standard 25-foot cable while others must be specified. A
site visit prior to purchase will determine the length needed. If a site visit is not possible,
a 50-foot cable is typically sufficient (see Section 3.3.2.2).
Examples and pictures, respectively, of some commercially available pressure transducers are
found in Table 15 (on the next page) and Figure 16. Research your options carefully, and be sure
to account for the fact that non-vented transducers might require purchasing a second transducer
to measure barometric pressure on land. Also consider battery type (replaceable vs. non-
replaceable) and memory capacity.
B)
C)
D)
Figure 16. Numerous pressure transducers are commercially available. Some examples
include: A) Onset Hobo© Water Level Data Logger; B) Global Waters Water level logger
(WL16); C) INW Submersible Pressure & Temperature Smart Sensor (PT2X); and D) In-Situ
Level TROLL.
3.1.3. Staff gage
Staff gages allow for instantaneous readings in the field, verification of transducer readings, and
correction of transducer drift. In addition, if securely installed, the staff gage can potentially
provide an attachment point for the transducer. Gages at RMN sites should be USGS style
(Type A) and marked every 0.02 feet (Figure 17).
38
-------
Table 15. Examples of commercially available pressure transducers
Manufacturer and
sensor model
Onset Hobo© Water
Level Data Logger
(U20)
Global Waters
Water Level Logger
(WL16)
INW Submersible
Pressure & Temp.
Smart Sensor with
Datalogging (PT2X)
In-Situ Level TROLL
500 & 700
Type1
Non-
vented
Vented
Vented
or Non-
Vented
Vented
Ranges available
(ft) that meet
accuracy
standards2
0-13, 0-30
0-3, 0-15
Vented: 0-2.3, 0-
5.8, and 0-12
Non-vented: 0-12
0-11.5
Accuracy3
±0.05% Full
Scale (typical);
±0.1% Full scale
(maximum)
±0.1% Full Scale
at constant
temperature,
±0.2% over 35°F
to 70°F
±0.05% Full
Scale (typical),
±0.1% Full scale
(maximum)
±0.05% Full
Scale (at 15° C),
±0.1% Full Scale
(maximum)
Battery life
(typical use) and
replaceability
5 years with 1
minute or greater
logging interval;
factory replaceable
Up to 1 year
(depending on
recording intervals);
user replaceable
(9V DC batteries)
18 months at 15-
minute interval;
user replaceable
(AA batteries)
10 years or 2
million readings
Logger
memory
Approx.
2 1,700 records5
(64K bytes)
Approx. 81,759
records
Available in
130,000-,
260,000-, and
520,000-record
versions
130,000
records (2.0
MB)
Approx. price4
($)
$495
(instream) + $495
(on land)
$989 (with 25 ft of
cable)
$1,095 (plus
$2.35/ftcable)
$1,1 70 (plus cable
if needed)
Website
www .onsetcomp . com
www .globalw . com
http://inwusa.com
www.in-situ.com
transducers can be either vented or non-vented (see Section 3.1.2 for more information). Non-vented transducers require an additional transducer to collect barometric
pressure data. Some manufacturers sell both vented and non-vented versions of the same transducers.
2Only ranges that meet accuracy standards are included in this table. Smaller depth ranges have higher accuracy.
3 Accuracy will vary based on selected depth range of transducer. Accuracy is calculated as a percentage of the "full scale" (depth range) of the transducer. Small depth
ranges will have the highest accuracy.
4As of January 2013 and subject to change; reduced prices might be available for bulk orders.
5Readings can be taken at 15-minute intervals for approximately 112 days before the memory capacity is reached.
39
-------
Figure 17. Example of a USGS style staff gage
(Type A) marked in 0.02-foot increments.
3.1.4. Protective housing
Instream pressure transducers at RMN sites
should be encased in housings to protect them
from currents, debris, ice, and other stressors.
Inexpensive housings can be constructed from
1.25-in or 1.5-in diameter Schedule 40 (or
stronger) PVC pipe. Multiple 1/2-inch or larger
holes should be drilled into the PVC to allow
water to fluctuate at the same rate as in the
stream (Figure 18). Factory-slotted PVC pipe
with 0.010 slots, which is widely available for
use in monitoring wells and piezometers, would
also work well because it allows the free
exchange of water while also keeping out fine
sediments, debris, and biota. Some transducers,
like the one shown in Figure 19, come in
protective metal cases and do not need to be
encased in an additional housing. In streams that
are subject to extended ice cover, the entire
length of the transducer and the vented cable
can be encased to prevent potential damage. If
vandalism is a concern, the housings can be
painted black or camouflage to make them less
visible.
Figure 18. Example of a protective housing made of PVC pipe. This one is for a
vented pressure transducer, which is secured inside the pipe with zip ties.
40
-------
3.2. Pre-Deployment
Prior to deploying the pressure transducer, check the
transducer to ensure the batteries will last until the
next field visit. If the sensor has previously been
deployed, gently clean it (see transducer manual for
cleaning methods) prior to redeployment.
3.2.1. Accuracy Check
Most transducers will be factory calibrated prior to
being shipped. The accuracy of the transducers
should be checked over time by comparing transducer
data to staff gage data or by measuring the depth of
water over the transducer with a stadia rod or other
measuring device. Data should be compared over a
variety of water depths to ensure the transducer is
accurate over the full range of depths. Most entities
lack access to the type of facility that is needed to
calibrate the transducers, so post-deployment field
checks are typically used to check the accuracy of the
transducers.
Figure 20 illustrates how staff gage readings can be
used to check the accuracy of the transducer data over
time. The pressure transducer data are compared with
the staff gage readings at similar times to determine
the offset between the two. If transducer data do not
correspond to staff gage data and water depth, the
transducer might be fouled, it might need to be
recalibrated, or it might need to be resurveyed (Section 3.5). Consult the transducer manual for
details on how to calibrate transducers when recorded data are not accurate. If you recalibrate the
transducer, record detailed notes on pre- and post-calibration values and changes that were made.
If a transducer cannot be calibrated and appears to be recording inaccurate data, contact the
manufacturer for further instructions, as it might need to be returned for service.
3.2.2. Sensor configuration and launch
Configure the pressure transducer and data logger to record temperature and pressure every 15
minutes, starting on the hour. This is the same recording interval that is used at USGS gages.
Some transducers allow the user to enter an offset value so that logged data are automatically
adjusted to the staff gage. If using this option, record the details in the field notebook. Consult
the transducer manual for specific details on configuration and launch.
The following practices will make data processing and screening easier and more efficient:
• Set the transducers up so that they start recording on the hour (xx:00), half hour (xx:30),
or quarter hour (xx:15 or xx:45).
Figure 19. Some transducers come in a
protective metal case and do not need
to be encased in an additional housing.
41
-------
4.0
3.5
3.0
2.5
2.0
Transducer data
Staff gage readings
Stp
Oet
Nov
0«C
Jan 2013
Date
Figure 20. Staff gage readings provide a quality check of transducer data. In this example,
staff gage readings stopped matching transducer readings in November, indicating that the
transducer or gage could have changed elevation.
• If you are using unvented pressure transducers, set both transducers up so that they
record at the same time.
• Consider using military time (if this is an option) to avoid potential confusion with
a.m./p.m..
• Consider using standard time (e.g., UTC-5 for sites in the Eastern Time zone) instead of
daylight savings time. Regardless of which you choose, make sure that any discrete
measurements taken for accuracy checks are consistent with this setting.
Bringing a computer into the field is helpful to launch the transducer and check the logged data
once it is installed. Alternatively the transducer may be launched prior to deployment provided
you note the date and time of actual deployment in the stream (this is important for data
screening; see Section 3.7.2).
42
-------
3.3. Deployment of Instream Pressure Transducers
Installation of both staff gages and pressure transducers are covered in this section because they
are typically installed during the same visit and pressure transducers are sometimes attached to
the staff gages. During deployment, safety should always be a consideration, and transducers
should only be placed in streams that can be safely accessed and waded by crew members. Install
staff gages where they can safely be read at all stages. This might require that you install two
sections of a staff gage in different locations (e.g., one in the low flow channel and one in the
high flow channel) if the stream experiences a wide range of stages. Prior to installing the
equipment, obtain permissions from all relevant parties. This often includes the local Department
of Public Works (particularly for bridge installations), the local conservation commission (or
equivalent), and abutting property owners if accessing sites will involve crossing private
property. Field crews should make efforts to minimize site impacts.
3.3.1. Selecting a location
Prior to gage and transducer installation, conduct a site reconnaissance survey to identify suitable
stretches of river for installation. At some sites, natural streambed installations will be most
appropriate, while at others, attaching equipment to built structures like bridges might be best.
During the site visit, determine what type of installation is best so that appropriate equipment can
be obtained prior to installation. If the transducer is collecting data that are representative of the
characteristics of the reach from which the biological data are being collected, the transducer
does not have to be located in the exact area where the biological data are being collected (e.g.,
perhaps the best attachment point for the transducer is a bridge or other built structure, and the
biological sample is collected 200 meters upstream from the bridge).
Key considerations for siting instream transducers are listed below. These considerations are
similar to those for water temperature sensors (see Section 2.3.1). More detailed information on
site selection and controls can be found in Rantz et al. (1982).
• The water level data should be representative of the characteristics of the reach from
which the biological data were collected. Note that for general monitoring purposes, the
sensor need not be placed in the exact location where biological data are being collected.
• Ensure that the site is not in the immediate vicinity of tributaries entering the river, and
that no water is entering or exiting between the pressure transducer and the biological
sampling site (e.g., through tributaries, pumping, or diversions). The goal is to minimize
potential impacts from backwater during high flows (tributary downstream) or unevenly
distributed streamflow across the channel (tributary upstream).
• The gaging equipment should be installed in a pool where turbulence is minimal to
increase accuracy of gage and transducer readings. The pool should have a downstream
control feature that allows for stable stage measurements and ensures that the equipment
will be submerged during low flows (Figure 21, Rantz et al. 1982). Natural controls
might include a downstream riffle, bedrock outcrop, or other stream feature that controls
flow. Unnatural controls might include a bridge or culvert that is narrower than the
stream channel and constricts flow. Note that the feature controlling the stage-discharge
relationship can change at different flow levels; such changes will be reflected in the
rating curve.
43
-------
The site should not have extensive aquatic vegetation, beaver activity, or unstable
streambeds and banks. These factors can change or result in unstable stage
measurements.
Figure 21. Examples of controls downstream of staff gages include A. riffle and B. culvert.
3.3.2. Staff gage installation
Staff gages and pressure transducers are typically installed during the same site visit, and
transducers are sometimes attached to the staff gages. Staff gages can be attached to a fixed
object in the stream (e.g., bridge, boulder, or weir) or installed in the streambed (Figure 22). If
suitable conditions are present and proper equipment is available, attaching gages to a bridge or
other fixed structure is preferable, as this minimizes the chance that the gage will shift in high
flows or in the presence of large debris. The streambed installation method works best in smaller,
higher order streams that do not experience extreme high flows that could knock over the gage.
A list of equipment needed for installation is in Table 16.
Figure 22. Examples of gage installation techniques include A. streambed, B. fixed object
bridge wing wall, and C. fixed object - boulder.
44
-------
Table 16. Equipment list for staff gage installation
Location
Supply list
All
Staff gage
Gage board
Screws (stainless steel or brass)
Screwdriver
Assorted drill bits for wood
Level
Stepladder (if installing more than one section of a staff gage)
Datasheets and field notebook
Survey equipment (auto level or laser level and stadia rod, paint
marker or nails for marking benchmarks)
Fixed Object-
additional items
• Concrete wedge anchor and nuts (stainless steel or galvanized)
• Hammer or rotary drill
• Concrete drill bit
• Hammer
Streambed -
additional items
• Galvanized or stainless steel pole with cap
• Galvanized or stainless steel conduit straps
• Pole driver or sledge hammer
• Pry bar
Gages are typically available in 3.33-foot sections; multiple gage sections can be combined to
encompass the entire range of stages expected to occur in that section of the stream (additional
sections can be added if the river tends to rise higher than expected). Stream gages should be
attached to a board that is at least as high and wide as the gage and able to withstand being
submerged in water (good materials include oak, plastic wood, or pressure-treated wood). Use
screws and a level to attach the gage to the board. Care should be taken to level the combined
board/gage before it is attached to the fixed object or pole.
For fixed object installations, the gage and board are directly attached to a bridge or other
permanent structure, using the following procedure:
• Level the gage board on the fixed object and mark locations for concrete wedge anchors
(Figure 23). Use concrete anchors on both sides of the board and at varying elevations to
ensure the board's stability. Two anchors are typically sufficient for a 3-foot gage and
three for a 6-foot gage.
• Use a hammer drill or rotary hammer with a concrete drill bit to drill holes in the marked
locations.
• Pound concrete wedge anchors (bolts) into the holes using a hammer until they are
secure.
• Drill holes in the gage board for concrete wedge anchors.
Place the board over the bolts and screw it into place using the nut.
45
-------
For streambed installations, the gage and board are attached to a pole that has been driven into
the streambed, using the following
procedure:
• Use a pry bar to test the
streambed in the pool for
locations where the pole can be
driven into the bed.
• Drive a galvanized steel pipe
with cap into the streambed
using a pole (fencepost) driver
or sledgehammer. Poles of 6
feet and 9 feet are generally
sufficient for 3.33-foot and
6.66-foot gages, respectively.
The pole should be driven into
the streambed at least 3 feet or
until it is very stable and
Figure 23. Example of a 0.5-inch by 3.75-inch wedge
anchor with bolt and washer.
unlikely to be knocked over in high flows. The pole should be as straight as possible so
that when the board and gage are attached, they can be leveled. A step ladder can be
helpful for driving the pole.
• Attach the combined board/gage to the pole using galvanized or stainless steel conduit
straps (use at least two for a 3.33-foot gage). The board should be parallel to the flow
and, if possible, positioned so that it can be read from the stream banks.
If a tree is nearby on the stream bank, a metal bracket or brace can be attached from the gage to
the tree to increase stability.
3.3.3. Pressure transducer installations
This document describes two installation techniques for pressure transducers: the fixed object
method, in which the transducer is attached vertically to the staff gage board or to an object like
a bridge, boulder, or weir; and the streambed method, in which the transducer is laid horizontally
on the streambed and is held in place by rebar and rocks. Site-specific conditions dictate which
technique is most appropriate.
3.3.3.1. Fixed object
With the fixed object installation, the pressure transducer is attached vertically to the staff gage
board or to an object like a bridge or boulder. Before the pressure transducer can be attached to
the fixed object, it must be enclosed in a protective housing (Section 3.1.4). For vented
transducers, which do not need to be removed from the stream to download data, the housing
should be cut to a length at least slightly longer than the transducer. If PVC pipe is used, attach
the transducer to the housing with heavy-duty zip ties to ensure it does not move (Figure 24). Do
not leave slack in the zip ties. If the unit can move, it can wear through the zip tie.
46
-------
c.
Figure 24. Non-vented pressure transducer installation. A. Non-stretch cable is attached to
the transducer, B. A loop is made at the non-transducer end of the cable, C. A cable with
transducer is suspended over the bolt from the top of the PVC pipe.
Non-vented pressure transducers typically should be removed from the stream to download data.
To facilitate data removal and downloading during high water periods, install the transducer into
a PVC pipe as follows:
• Use a PVC pipe that is approximately the same height as the gage board.
• Attach a non-stretch cable or rope (e.g., coated stainless steel cable) to the transducer and
make a loop at the non-transducer end (Figure 24). The cable should be long enough that
the transducer will always be under water but not so long that it will come into contact
with the streambed.
• Place the looped cable over a long bolt with a wing nut that runs through the top of the
PVC (Figure 24). Clearly mark the holes for the bolt so that the transducer is placed back
at the same elevation every time it is removed.
Once the pressure transducer has been installed in the protective housing, the transducer/housing
can be attached vertically to the staff gage board (Figure 25), a bridge abutment, or fixed object,
or horizontally to rebar or other pipe in the stream bed. An equipment list for installations is
found in Table 17.
47
-------
Figure 25. Non-vented (left) and vented (right) pressure transducers attached to staff gage
board using conduit hangars.
Table 17. Equipment list for transducer installation
Location
Supply list
All (continued)
PVC pipe for transducer (drilled with 1/2-in or larger holes)
PVC pipe for data logger or barometric pressure logger
Galvanized or stainless steel conduit straps, hangars, and hose clamps
Zip ties/cable ties
Screws (stainless steel or brass)
Screwdriver
Drill
Assorted drill bits for wood and PVC
Level
Datasheets and field notebook
Survey equipment (see Section 3.3.5)
48
-------
Table 17. Equipment list for transducer installation (continued)
Location
Fixed object -
additional items
Supply list
If using a non-vented transducer, you also need:
• Non-stretch cable or wire
• Wire rope clamps
• Long bolt and wing nut
• Extra-long PVC (should extend out of water in high flows)
• Solar shield (if using barometric logger for air temperature data, see
Appendix B for instructions)
• PVC with caps for barometric transducer
If using a vented transducer, you also need:
• Garden staples
• PVC pipe for data logger
• PVC cap or locking well cap
• Stainless steel conduit straps
• Long lag screws
• Wire cable
• Replacement desiccant (if necessary)
• Hammer or rotary drill
• Concrete drill bit
• Concrete wedge anchor and nuts (stainless steel or galvanized)
• Hammer
To install the transducer/housing on a staff gage board:
• Place the PVC pipe with transducer on the downstream side of the board directly next to
the galvanized steel pole, or whichever side is the least turbulent.
• Attach the PVC pipe to the board vertically using galvanized or stainless steel conduit
straps or hangars.
• Ensure that the PVC pipe and transducer are installed slightly above the streambed to
reduce chances of fouling by fine sediments.
To install the transducer/housing on a fixed object (e.g., bridge abutment):
• Install the transducer close to the staff gage so that stage changes are comparable.
• Use concrete anchors and conduit straps, hangars, or hose clamps to attach the PVC
vertically to a bridge or boulder.
• Ensure that the PVC pipe and transducer are installed slightly above the streambed to
reduce chances of fouling by fine sediments.
49
-------
3.3.3.2. Streambed/rebar
With the streambed installation, the transducer is laid horizontally on the streambed and is held
in place by rebar and rocks (Figure 26). Compared to the fixed object method, transducers are
more prone to moving or being swept away during high flows, and ensuring that transducers are
returned to the exact same location after being removed for data downloads can be more
difficult.
Staff gage
PVC housing
holding
transducer
Rebar
Figure 26. Examples of streambed installations of pressure transducers using rebar.
Streambed installations require rebar stakes and a sledgehammer in addition to the equipment
listed in Table 16. Installation instructions are as follows:
• Find a pool location close to the staff gage and protected from turbulence and debris
(behind a large rock is ideal).
• Drive two rebar stakes into the ground (spaced apart slightly less than the length of the
PVC housing).
• Use cable ties or steel hose clamps to attach the PVC to the rebar in a relatively
horizontal position.
• For added stability, pin the PVC pipe between the rebar and the rock, or place large rocks
over the PVC pipe, or do both.
• Ensure that the PVC pipe and transducer are installed slightly above the streambed to
reduce chances of fouling by fine sediments.
50
-------
3.4. Deployment of On-Land Components
Both vented and non-vented transducers have components that must be installed on land. Vented
transducers are typically connected to data loggers on land via a vented cable. When using non-
vented transducers, a separate, identical transducer must be deployed on land to measure
barometric pressure. This section describes a general installation approach for both components
and concludes with specific considerations for data loggers and for barometric pressure sensors.
Both devices should be located well above where high flows or the snow pack will reach, but not
so high that they are inaccessible for data downloads. Trees are generally the most suitable
attachment points for these devices. If no trees are present, install a stainless steel pole (with a
cap) and attach the data logger to it. Use the same type of pole used for the streambed staff gage
installations. Hide the devices from view to reduce the chance of vandalism.
The devices should be encased in a PVC pipe. The device may sit on the bottom of the pipe or be
suspended on a wire or cable. If the data logger is not supported by the PVC pipe, suspend it on a
wire or cable so no pressure is placed on the vented cable. Leave enough slack in the data logger
cable to allow the logger to be lifted slightly out of the PVC pipe to download data.
Installation instructions are as follows:
• Attach a PVC pipe to a fixed object on the bank. The pipe can be attached to a tree with
U-shaped conduit straps and long screws or zip ties (Figure 27). Adjust the length of the
pipe to adequately cover the device with some extra room.
Figure 27. Vented transducer data logger installation. The transducer is attached to the
tree with stainless steel conduit straps.
51
-------
• Place the device inside the PVC pipe.
• Place a PVC cap or a locking well cap on top of the PVC pipe to protect the equipment
and discourage vandalism.
Specific considerations for the data logger include:
• For vented transducers with excess cable, coil the cable and zip tie it to the tree or PVC
pipe.
• Use garden staples (4-inch U-shaped steel staples), rocks, and leaf litter to hide the data
logger cable. For extra protection, the cable can be encased in a PVC pipe.
Specific considerations for the barometric pressure sensor include:
• Place the device as close as possible to the instream pressure transducer.
• The PVC housing should be drilled with holes to facilitate air flow (similar to those used
to protect the instream pressure transducer [Figure 28]).
.' •«. V
Figure 28. Barometric pressure sensor installation for a non-vented transducer, A.
without the PVC cap and B. with the PVC cap. Holes should be drilled into the bottom
cup of the PVC housing. This prevents condensation and laterally blown rain and snow
from filling the cup to a depth sufficient to inundate the ports through which the
barometric pressure is compensated. Another option is to deploy the barometric
pressure sensor upside down, so that the port tip is on top.
52
-------
• Holes should also be drilled in the bottom of the PVC housing to prevent condensation
and laterally blown rain and snow from filling the cup to a depth sufficient to inundate
the ports through which the barometric pressure is compensated. If this occurs, the
"barometric pressure" becomes barometric pressure plus a small amount of pressure due
to this accumulated water, and the data cannot be used. This issue can also be addressed
by deploying the barometric pressure sensor upside down, so that the port tip is on top.
3.5. Elevation Surveys and Documentation
After the installation is complete, georeference and document the pressure transducer location in
a way that enables field personnel to relocate the sensor during subsequent visits. Take
photographs from different perspectives (e.g., upstream and downstream); photos are important
for relocating the transducers and documenting changes to the monitoring location during the
course of the study. Record all documentation information on a field form that includes
information such as station number, waterbody name, date, time, crew members, driving
directions, serial number/s of transducer/s, time of deployment, transducer installation technique,
image numbers/file names for the photographs, and detailed descriptions of transducer
placement. Be sure to record the time and date of deployment accurately, as this information will
be used in the error screening procedure described in Section 3.7.2.
In addition, survey the elevation of the staff gage and pressure transducer to establish a reference
point for the gage and transducer. This allows for monitoring of changes in the location of the
transducer, which is important because if the transducer moves, stage data will be affected and
corrections will need to be applied. If possible, conduct elevation surveys at least once a year to
identify if and when movement occurs. This is particularly important after high flow events and
periods of extended ice cover. Steps for conducting elevation surveys using an auto level are
summarized in Table 18, and Table 19 contains a list of equipment needs. For additional
information on surveying techniques see Harrelson et al. 1994.
Table 18. Quick guide for elevation surveys of staff gages and transducers using an auto level
Procedure
1. Set up tripod and auto level in location where you can view benchmark and all survey points.
Level auto level by centering the level bubble using the foot screws.
2. Have person hold stadia rod on top of benchmark. Rod must be level.
3. Read elevation at which crosshairs in auto level cross the stadia rod.
4. Record this elevation as the starting backsite (B.S.).
5. Have person hold stadia rod on survey point # 1. Rod must be level.
6. Read elevation at which crosshairs in auto level cross the stadia rod.
7. Record this elevation as the foresite (F.S.).
8. Repeat steps 6 to 8 for additional survey points (including the top of the staff gage and
transducer).
9. Have person hold stadia rod on top of benchmark. Rod must be level.
10. Read elevation at which crosshairs in auto level cross the stadia rod and record as final B.S.
11. If final B.S. does not match starting B.S. (within 0.01 ft), repeat the entire survey.
53
-------
Table 19. Equipment list for elevation
surveys
Supply list
Perform the elevation survey on the day of staff
gage and transducer installation. Identify and
establish a benchmark and one or two other
permanent markers at the site. The benchmark
will serve as the predominant reference point for
the gage/transducer; the additional permanent
marker(s) provide a backup in case the benchmark
is destroyed and allows for a check of benchmark
movement. Knowing the absolute elevation of the
benchmark is unnecessary as the purpose is to
detect changes in elevation of the equipment
relative to the benchmark. The benchmark
therefore can be given a relative elevation
(e.g., 100 feet) as part of the calculations. Reliable
permanent markers can be found on bridges and other permanent structures (e.g., a specific
corner on a part of the bridge structure) (Figure 29); nails in trees are less stable but a good
alternative in relatively secluded sites. Boulders should be used only if they are very unlikely to
shift and if a clear point on the boulder can be identified and marked with survey paint. All
permanent markers should be visible and accessible during all expected stream stages and
vegetation covers.
Auto level or laser level
Tripod
Stadia rod
Survey paint
Survey nails
Datasheet
Figure 29. Examples of auto level and tripod used for elevation survey (A) and permanent
structure used as a benchmark (B).
54
-------
Once markers are established, sketch their location on the data sheet (see example in Figure 30),
clearly documenting both markers and relevant site features that can be used to locate the
markers in subsequent surveys. Take pictures of the benchmark with clear identifying features.
Survey the benchmark, permanent marker(s), and gage and transducer elevations, repeating the
survey of the benchmark after all other points (see Table 18). If the benchmark height is different
at the end of the survey compared to the beginning, repeat the entire survey. For surveying the
staff gage, hold the stadia rod on the top of the metal gage (not the board). For surveying the
transducer, make clear notes of where the stadia rod was placed on the PVC pipe (e.g., top, river
right, river left). Calculate the elevations of all points relative to the benchmark using the
following equations:
Height of instrument (H.I.) = Backsite (B.S.) + Elevation of benchmark (B.M.)
Elevation of survey point = H.I. - Foresite (F.S.),
where B.S. is the height of the stadia rod at the benchmark and F.S. is the height of the stadia rod
at each survey point (e.g., permanent marker, top of gage, transducer housing) and the elevation
of benchmark is the relative elevation (e.g., 100 feet). To calculate the actual elevation of the
transducer, subtract the distance from the transducer housing (at the point where the stadia rod
was placed) to the sensor face from the elevation of the transducer housing.
After elevations of all survey points are calculated, compare beginning and end H.I. to ensure no
movement of the auto level during the survey. Next, compare elevation of all survey points to
previously surveyed elevations to determine if movement occurred.
3.6. Maintenance/Mid-Deployment Checks and Data Offload
After a transducer is installed at a site, try to revisit the site within the first month to confirm that
the installation is holding properly. After these initial deployment checks, at a minimum, visit
sites annually (e.g., in conjunction with the biological sampling events). More frequent visits are
encouraged, however, particularly to check for movement of the staff gage and transducer after
high flow events and periods of extended ice cover. More frequent site visits will help ensure the
longevity of gaging stations and data quality.
During site visits, field personnel should:
• Check the condition of the transducer and perform necessary maintenance.
• Take staff gage readings or measure the depth of water over the transducer with a stadia
rod or other measuring device, ideally at different flow conditions, to check the accuracy
of the transducer data. To minimize the chance of a faulty measurement, take the
measurement as close as possible to the time that the pressure transducer is recording a
measurement, and get as stable a reading as possible (if flows are fluctuating rapidly at
the time of the measurement, note this on your field form and try to record the depth as
accurately as possible). Be sure that no one is standing upstream during these
measurements, as this will affect the readings.
• Offload data.
55
-------
Site:
r M
River Instream Flow Stewards (RIFLS)
Staff Gauge Survey Sheet
Date: \3-IS/!.'
Gage (ft): 6J
Notes:
,
BS = backsite, HI = height of instrument, FS = fofesite, BM = benchmark, TP = turning point
H.I. = B,S. + Elevation, Elevation = H.I. - F.S.,
-s -Hlift
F^
lUPT ( P
t>y
IHfi.
ISE35
ajar
Notes and Site Sketch:
NT
Figure 30. An example of a completed elevation survey form.
56
-------
Typical maintenance issues include:
• Ensuring that the instream pressure transducer is submerged and not buried in sediment.
• Clearing leaf litter and debris that might pile up against the gage, transducer, and
downstream control. At the beginning of a site visit, clear this material as it could impact
gage height. Note the stage before and after debris clearing to check for any changes.
Stage data may be corrected if changes are detected.
• Checking for transducer and staff gage movement after high flows and floods. The
difference in elevation between the staff gage and the transducer can change if sediment
accumulation or scour occur near the gage and transducer. If movement is detected,
secure the equipment (if necessary) and resurvey. Note any differences in gage and
transducer elevation in a field notebook along with the date and time. If a gage is
constantly shifting, consider an alternative location.
• Checking for impacts from ice cover. Water temperature and stage data should be
evaluated for potential impacts from ice cover and data should be flagged accordingly.
• Cleaning sediment or algae from the pressure transducers. These can cause fouling and
inaccurate readings. Consult the transducer manual for specific instructions on cleaning
and maintenance.
• Cleaning the staff gage with a scrub brush, especially during the summer months, so that
the gage can be accurately read. For especially dirty gages, baking soda or native sand (if
available in situ) can help in cleaning the gage. Paint over any rust marks on the gage
with enamel paint to improve durability.
• Checking the condition of desiccantpackets (vented transducers only). These are needed
to keep the vented cable dry. Different transducers use different types of packets, and the
lifespan of these packets varies depending on site-specific conditions (e.g., how much
moisture is present in the air).
If possible, download data during each site visit. Frequent data downloads, as well as frequent
stage measurements from the staff gage, will help in early identification of transducer drift. To
download data, a computer and a cable or data shuttle (to connect the transducer to the computer)
is typically needed in the field (see Figure 31). Prior to download, record the stage at the staff
gage and note any factors that might have affected the data since the previous deployment. After
download, spot-check the data for accuracy. If no cleaning or troubleshooting is necessary, re-
launch the transducer following procedures outlined in Section 3.2.2. Each time data are
downloaded, complete a form like the one shown in Figure 32.
Also check battery life. Batteries in some sensors need to be replaced each year, while others last
5 years or longer under normal use. Plan for these replacements.
Record information from mid-deployment checks in a field notebook or on a field form. That
good field notes be taken during every site visit is essential. Entries should include notes about
the condition of the transducer, staff gage (or other water level) readings, and whether any
unusual measurements appeared during the data spot-check (if available). Taking photos during
57
-------
each site visit is also good practice, as this will help document changes to the monitoring
location during the course of the study.
Figure 31. Data download from vented transducer with external data logger on land using
computer and cable (left) and non-vented transducer with internal data logger using
computer and data shuttle (right).
Site:
Date:
Crew:
Weather:
Gage (ft.):
Photos taken?
Battery Status:
File name:
Notes:
Transducer Download Field Data Sheet
ftJLb &\I?$- AT SOOTH CiMA/IM B.OA£> PLcX^i ^A . AAA-
ujs/i-z.
L.'PJrtlCtTP-,
rtJE^A-ST .(
!.•&<;
i^Brtt) ( 11 -T
tx^NUvVOK
0?- ^AC,E /T
Time: IS'lp
M. C(Jrt-CirtrJ!WVPifi- - cW
^S^dr.W° ™™
JTtDULiWfi
Figure 32. Example of a transducer download field data sheet.
3.7. Quality Assurance and Control
Proper QA/QC is essential to collecting accurate long-term data. The procedures discussed in the
following sections will help detect and correct for errors in transducer and stage data.
3.7.1. Accuracy checks
Pressure transducer readings can drift over time, which can result in deviation of transducer data
from that observed at the staff gage. As mentioned in Section 3.2.1, making periodic staff gage
readings to detect any shift or drift in the transducer data is important. The stage should be
recorded every time the site is visited for a discharge measurement or data download. Ideally,
gage readings should be done at least every month, if possible, to ensure a variety of stages are
58
-------
captured throughout the deployment period. Frequent gage readings facilitate error screening and
early detection and correction of transducer problems that help minimize data loss. If this
frequency is not possible due to the remoteness of the field site, local volunteers or other state
agency collaborators could assist with gage readings. Local watershed groups or other
organizations might be able to help to identify local volunteers. At the minimum, try to visit sites
after large storm events, which could impact the transducer.
3.7.2. Error screening
Pressure transducers might record erroneous readings during deployment for a variety of reasons,
including:
• they can become dewatered during low flow conditions
• high flow events can bury them in sediment
• high flow events could move them
• they might become fouled from debris, aquatic vegetation, or algae
• humans might cause interference
• they could become encased in ice
• if moisture enters the cable of a vented transducer, it could result in erratic readings or
readings of zero water depth
• if the cable of a vented transducer becomes kinked or plugged, it could result in the data
not being corrected for barometric pressure
Because these errors may occur, data should be screened. A series of error screening checks is
described in Table 20, and Appendix K contains a data screening checklist. Various software
applications can be used to help with the error screening checks (e.g., Microsoft Excel, Aquatic
Informatics' AQUARIUS (http://aquaticinformatics.com/products), including software that is
purchased with the transducers(s)).
The first step in the error screening process involves removing observations recorded before and
after the transducer is correctly positioned. This can be done via a visual inspection of data and
by referencing field notes indicating the exact times of deployment and recovery. While
reviewing the field notes, also look for comments about situations that could cause the transducer
to record questionable readings (e.g., during a mid-deployment check, the transducer was found
to be buried in the sand) and flag those data accordingly.
Next, plot all of the measurements and visually check the data. Look for missing data, outliers
and other abnormalities, Specific errors to watch for include: missing data; values of 0 (which
could mean that the pressure transducer was dewatered, or in the case of vented transducers,
could mean that moisture entered the cable); values flat-lining at 0°C/32°F (the stream pressure
transducer is likely encased in ice); negative values (if unvented transducers are being used, this
could indicate that the barometric pressure correction is off); and outliers or rapidly fluctuating
values (the stream pressure transducer could have moved (e.g., due to a high flow event or
vandalism). Also consider performing the following additional checks: graphically compare data
59
-------
Table 20. General summary of different types of problems that can occur with pressure
transducer data and recommended actions for addressing them
Problem
Recommended action
Missing data
Leave blank
Stream pressure transducer was
dewatered or buried in sediment
for part of the deployment period
Use the plot (and temperature data, if available) to determine the
period during which the problem occurred. Exclude these data when
calculating summary statistics.
Recorded values are off by a
constant, known amount (e.g.,
due to a calibration error)
Adjust each recorded value by a single, constant value within the
correction period.
Drift is large and when and how
much the sensor was 'off cannot
be determined (when drift occurs,
the difference between staff gage
or depth readings and pressure
transducer readings increases over
time)
The data should be removed.
General rules:
Discrepancy between pressure
transducer reading and discrete
measurement taken during a staff
gage or depth check
If the errors are smaller than the accuracy quoted by the
manufacturer and cannot be easily corrected (e.g., they are not
off by a constant amount), leave the data as is, and include the
data in summary statistics calculations.
If the transducer fails a staff gage or depth accuracy check,
review field notes to see if any signs of disturbance or fouling
were noted, and also look for notes about the quality of the gage
measurement (Were flows fluctuating rapidly at the time of the
measurement?). Also check whether the same time setting was
used for both the transducer and gage or depth measurements
(daylight savings time vs. standard time). Based on this
information, use your best judgment to decide which action
(leave as is, apply correction, or remove) is most appropriate.
A shift is detected and an
elevation survey reveals that the
stream pressure transducer has
moved
Stage readings can be adjusted by adding or subtracting the difference
in elevation. If the exact date of the elevation change is unknown,
compare gage data to transducer data to observe any shifts. If no gage
data for the time period are available, transducer data should be
examined for any sudden shifts in stage. Changes in the elevation
typically occur during high flows, so closely examine all data during
these time periods.
The sensitivity of the pressure
transducer changes with stage
(e.g., the transducer is less
sensitive or accurate at high
stages)
Sensitivity drift can be detected by graphing the difference between
transducer and staff gage readings against the gage height and plotting
a linear trend line through it. A strong correlation between the data
sets and a positive or negative trend line as stage increases or
decreases could indicate a sensitivity shift. Based on this information,
use your best judgment to decide which action (leave as is, apply
correction, or remove) is most appropriate.
60
-------
across years (when data from one year are dramatically different, errors might be present);
graphically compare with precipitation data from the nearest active weather station; and
graphically compare with data from the nearest USGS stream gage, if appropriate. If
discrepancies occur, use the plots to specify the time and duration of errors in the raw data files.
Additional checks (e.g., for accuracy and drift) are described in Table Kl and Appendix K. If
you are inexperienced at addressing errors with continuous stage data, consider seeking guidance
from someone with more experience and consult references like Wagner et al. (2006). Table 20
provides a general summary of different types of problems that can occur (e.g., missing data,
failed accuracy check) and recommended actions for addressing them. Corrections should not be
made unless the cause(s) of error(s) can be validated or explained in the field notes or by
comparison with information from nearby stations.
Any changes or corrections that are made to the data should be noted, and both the original and
the "cleaned" data files should be maintained and backed up. Large amounts of data will
accumulate quickly so a central database should be developed and maintained from the initial
stages of monitoring. Also, all field forms should be organized, easily accessible, and archived in
a way that allows for safe, long-term storage. Accurate field notes and accuracy check logs are
essential in the data correction process.
3.8. Developing Stage-Discharge Rating Curves
The previous section describes how to install and maintain pressure transducers to measure
continuous stage (depth). Taken alone, stage measurements yield some information about
streamflow patterns, including the timing, frequency, and duration of high flows (McMahon et
al. 2003, Roy et al. 2005). Stage data per se, however, do not provide quantitative information
about the magnitude of streamflows or flow volume, which makes comparing data between
streams difficult. Furthermore, the channel shape can change from year to year, such that a
similar stage measured in a given location during two separate years could represent different
flows. Thus, to assess patterns and changes in stream hydrology, converting stage measurements
into streamflow is most useful. Different approaches are available for converting stage to
discharge. The most common approach is to develop a stage-discharge rating curve, which we
present here.
A rating curve enables the user to convert stage measurements to streamflow. To develop a
rating curve, a series of discharge (streamflow) measurements is made at a variety of stages,
covering as wide a range of flows as possible. In this section, we briefly describe how to take
discharge measurements in wadeable streams. This involves measuring the depth and velocity of
the water passing through several segments along a given cross-section of stream. Each
measured velocity is multiplied by its contributing flow area; the resulting flows are summed
across the cross-section to produce a total flow. A full treatment of this topic is beyond the scope
of this document. For more detailed guidance consult Rantz et al. (1982), Chase (2005), or Shedd
(2011).
3.8.1. When to measure discharge
Five to ten discharge measurements should be made to establish a rating curve at a new site. To
construct a rating curve that accurately predicts flow under most conditions, take measurements
61
-------
over as wide a range of flows as possible. After establishing a rating curve, measure discharge at
least once annually, and if possible, also after large storms or any other potentially channel-
disturbing activities, to verify or (if needed) update the curve (see Figure 33. for an example of a
channel-disturbing activity). If new measurements are more than 15% off the rating curve,
follow-up measurements should be made to identify whether a shift has occurred and, if
necessary, to establish a new rating curve.
Figure 33. A culvert replacement downstream of a stream gage on Gulf Brook in Pepperell,
MA caused enough of a channel change to necessitate a new rating curve.
3.8.2. Equipment
This section covers measuring discharge using current meters, which is generally the most low-
cost approach (compared to using acoustic doppler-based instruments, for example). The basic
equipment includes:
• Current meters measure point velocity. Many different types are available, including
mechanical meters (e.g., Price and Pygmy meters, which are vertical-axis meters) and
electromagnetic meters (e.g., Hach/Marsh-McBirney).
• Wading rods are used to measure water depth at verticals and to set the current meter at
the appropriate depth.
• A measuring tape and stakes are used to define the exact location of the cross section at
which depth and velocity are measured.
3.8.3. Site selection
Site selection is critical for making a good discharge measurement. An ideal cross section will
have the following characteristics (see Figure 34 for examples):
• A relatively straight stream channel with defined edges and a fairly uniform shape
• Limited vegetative growth, large cobbles, and boulders
62
-------
Figure 34. Examples of good cross-sections for making discharge measurements.
• No eddies, slack water, or turbulence
• Depths greater than 0.5 feet and velocities greater than 0.5 feet per second
• Similar flow to that at the gaging station (e.g., no tributaries or drainpipes should be
located between the cross section and the gaging station)
Meeting all of these criteria is often not possible (neither cross-section depicted in Figure 34 is
perfect). Some minor alterations of the streambed, such as removing excessive aquatic plants or
large rocks, can significantly improve the quality of a cross section. Rocks can also be moved to
create a more defined stream edge. All such changes must be made before starting
measurements. The location of the "best" cross section will likely vary depending on flow
conditions. Often a culvert or bridge meets many of the above criteria and provides a good
location to measure flow.
3.8.4. Measurements
Streamflow is calculated by summing individual discharge measurements at numerous segments
of the cross section. A measuring tape (tagline) is stretched across the stream perpendicular to
streamflow and anchored at both banks. Distance along the tagline, channel depth, and stream
velocity are measured and recorded at a minimum of 20 points along the cross section. Be sure
that no one is standing upstream during these measurements, as this will affect the flow readings.
Measurement points should be closer together in portions of the cross section where flow is more
concentrated and depths are greatest, in order to ensure that no more than 5% of the total
discharge passes through any single segment. Measurements may be farther apart where the flow
is lowest and depths are shallowest (Figure 35). Include additional segments where velocity or
bottom irregularities are the greatest.
63
-------
Measure distance along tagline,
channel depth, and stream
velocity at a minimum of 20
points along the cross section.
290
34.0
-200
.7 Ct\
^^— Stream Dt p in
• ODs Depth
Figure 35. Layout of a channel cross-section for obtaining discharge data, using the velocity-
area procedure.
The exact methodology for making velocity measurements is somewhat specific to the
instrument being used. For more details refer to Rantz et al. (1982), Chase (2005), City of Salem
(2007), or Shedd (2011), and to instrument manuals.
In addition to measuring streamflow, field personnel should also estimate the gage height of zero
flow (GZF), which provides information about the low end of the rating curve. The GZF is the
water level at the gage at which the stream would cease flowing. To find it, locate the low flow
channel through the hydraulic control downstream of the sensor. Measure the depth of water at
the lowest point in the control and subtract that depth from the stage to calculate the GZF.
Measuring the GZF at each discharge measurement is best, but locating this point at lower flows
might be most practical and effective.
3.8.5. Documentation
At the time of the discharge measurement, take photographs of the staff gage, upstream of the
gage, and downstream of the gage. Detailed notes about the discharge measurements should
include the following, as applicable:
• Date, weather conditions, field team members
• Start and end times
• Stage at the beginning and end of discharge measurements
• Current meter check (e.g., spin tests for Pygmy/Price)
• Equipment used
• Whether the gage and transducer elevation were surveyed
64
-------
• Description of the cross section location and characteristics, description of any observed
changes that may impact the rating curve or streamflow
• Edge of b ank 1 ocati on
• Discharge measured (if calculated automatically by meter), or width, depth, and velocity
for each measurement (if calculating discharge in the lab/office)
• Point of zero flow (PZF, if measured) and location of PZF measurement
Examples of discharge datasheets for different flow meters can be found in Appendix J.
3.8.6. Quality assurance and control
As with other parameters, careful attention to QA/QC procedures is necessary to ensure accuracy
of discharge measurements. Some good resources for discharge QA/QC include Rantz et al.
(1972), Chase (2005), and Shedd (2011). Some major points to keep in mind are as follows:
• Make duplicate measurements, ideally with a different person making each measurement.
The measurements may be along the same or different cross-sections. The difference
between the two measurements should be less than 15%. Be sure that no one is standing
upstream during these measurements, as this will affect the flow readings.
• Periodically check the accuracy of your measurements by making measurements that you
can compare to a standard, such as a real-time USGS gage, or to those obtained by an
experienced hydrographer from the USGS or another agency.
• Major, channel-disturbing events (e.g., floods, new culverts) can alter the rating curve. If
a major event occurs and subsequent points are not aligned with the original rating curve,
a new rating curve might need to be developed and used to convert stage to discharge for
points following that event.
3.8.7. Making flow rating curves
The rating curve is produced by plotting instantaneous flow measurements and stage heights.
They can also be plotted in a basic spreadsheet program such as Microsoft Excel, or using
software designed to produce rating curves, such as Aquatic Informatics' AQUARIUS
(http://aquaticinformatics.com/products). The curve can include one or more break-points to
account for changes in channel morphology at different stages. When drawn on a log-log scale,
the rating curve should be a straight line. If the rating curve does not cover the full range of the
stage recorded, the curve can be extended to equal twice the highest or half the lowest
measurement recorded. For more detail, see Kennedy (1984). An example of a rating curve is
shown in Figure 36.
65
-------
• GULF BROOK, OAK HILL ST NEAR PEPPERELL. MASS. - DD: 1 Rating: 2
- GULF BROOK, OAK HILL ST NEAR PEPPERELL, MASS. - DD: 1 Rating: 2
1.5
tn
Grade
I
5 10
Discharge (ftA3/s)
0.5
0.4
Grade
\
.1
10
Discharge (ftA3/s)
Figure 36. Example of regular (left) and log-log scale (right) rating curves created using
Aquatic Informatics' AQUARIUS software.
4.
LITERATURE CITED
Baxter, C.V., and F.R. Hauer. 2000. Geomorphology, hyporheic exchange and selection of
spawning habitat by bull trout (Salvelinus confluentus). Canadian Journal of Fisheries and
Aquatic Sciences 57:1470-1481.
Bilhimer, D. and A. Stohr. 2009. Standard Operating Procedures for continuous temperature
monitoring of fresh water rivers and streams conducted in a Total Maximum Daily Load
(TMDL) project for stream temperature, Version 2.3. Washington State Department of
Ecology, Environmental Assessment Program. Available online:
http://www.ecy.wa.gov/programs/eap/qa/docs/ECY_EAP_SOP_Cont_Temp_Monit_TM
DL_v2_3EAP044.pdf
Chase, R. 2005. Standard Operating Procedure: Streamflow Measurement. Massachusetts
Department of Environmental Protection, Division of Watershed Management. CN 68.0.
Division of Ecological Restoration. 2010. River Instream Flow Stewards (RIFLS) Quality
Assurance Project Plan. Massachusetts Department of Fish and Game. (Available on-line
soon).
City of Salem, OR. 2007. Quality Assurance Project Plan: City of Salem Stormwater Services,
Continuous Water Quality Monitoring Program. Available online:
http://www.cityofsalem.net/Departments/PublicWorks/Operations/StormwaterServices/D
ocuments/qapp_and_sop_4-07.pdf
Danielson, T. 2006. Protocols for Measuring Continuous Water Temperature Using an Onset
Data Logger. Report DEPLW0700. Maine Department of Environmental Protection.
66
-------
Division of Ecological Restoration. 2010. River Instream Flow Stewards (RIFLS) Quality
Assurance Project Plan. Massachusetts Department of Fish and Game.
Dunham, J., G. Chandler, B. Rieman, and D. Martin. 2005. Measuring stream temperature with
digital data sensors: A user's guide. Gen. Tech. Rep. RMRSGTR-150WWW. Fort
Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research
Station. 15 p. Available online: ttp://fresc.usgs.gov/products/papers/143 l_Dunham.pdf.
Durance I, and SJ. Ormerod. 2007. Climate change effects on upland stream macroinvertebrates
over a 25-year period. Global Change Biology 13:942-957.
Edwards RT. 2001. The hyporheic zone. Pages 399-429 in Naiman RJ, Bilby RE, editors. River
ecology and management. New York: Springer.
Fulford, J.M., K.G. Thibodeaux, and W.R. Kaehrle. 1994. Comparison of Current Meters used
for Stream Gaging. Proceedings of the Symposium on Fundamentals and Advancements
in Hydraulic Measurements and Experimentation, Buffalo, NY. Available online:
http://water.usgs.gov/osw/pubs/CompCM.pdf
Gill, G.C. 1979. Development of a small rugged radiation shield for air temperature
measurements on drifting buoys. Report to NOAA Data Buoy Office for Development
Contract #01-7-038-827 (IF). Bay, St. Louis, MO.
Gordon, N.D., T.A. McMahon, B.L. Finlayson, C.J. Gippel, and RJ. Nathan. 2004. Stream
hydrology: an introduction for ecologists. Second edition. John Wiley and Sons,
Hoboken, NJ.
Harrelson, C.C., Rawlins, C. L. and J.P. Potyondy. 1994. Stream channel reference sites: an
illustrated guide to field technique. Gen. Tech. Rep. RM-245. Fort Collins, CO: U.S.
Department of Agriculture, Forest Service, Rocky Mountain Forest and Range
Experiment Station. 61 p.
Hill, R.A., Hawkins, C.P., and D.M. Carlisle. 2013. Predicting thermal reference conditions for
USA streams and rivers. Freshwater Science 32(l):39-55.
Holden, Z.A., Klene, A., Keefe, R., and G. Moisen. 2013. Design and evaluation of an
inexpensive solar radiation shield for monitoring surface air temperatures. Agricultural
and Forest Meteorologyl80:281-286.
Holden, Z.A., J. Abatzoglou, C. Luce, and L.S. Baggett. 2011. Empirical downscaling of
minimum air temperature at very fine resolutions in complex terrain. Agricultural and
Forest Meteorology 151 (8): 1066-1073.
Holden, Z. 2010. Temperature Sensor Installation Instructions. Available online:
http://www.fs.fed.us/rm/boise/AWAE/proiects/stream temp/blogs/06ThoughtsOn%20mo
nitoringAirTemperaturesInComplexForestedTerrain.pdf
67
-------
Isaak, D. J., D.L. Horan, and S.P. Wollrab. 2013. A Simple Protocol Using Underwater Epoxy to
Install Annual Temperature Monitoring Sites in Rivers and Streams. Gen. Tech. Rep.
RMRS-GTR-314. Fort Collins, CO: U.S. Department of Agriculture, Forest Service,
Rocky Mountain Research Station. 21 p. Available online:
http://www.treesearch.fs.fed.us/pubs/44251
Isaak, D. J., and D.L. Horan. 2011. An assessment of underwater epoxies for permanently
installing temperature in mountain streams. North American Journal of Fisheries
Management 31:134-137. Available online: http://www.treesearch.fs.fed.us/pubs/37476.
Isaak, D.J., S. Wollrab, D. Horan, and G. Chandler. 2012. Climate change effects on stream and
river temperatures across the northwest U.S. from 1980-2009 and implications for
salmonid fishes. Climatic Change 113:499-524. Available online:
http://www.treesearch.fs.fed.us/pubs/39783.
Jackson, J.K., and L. Fureder. 2006. Long-term studies of freshwater macroinvertebrates - a
review of the frequency, duration, and ecological significance. Freshwater Biology
51:591-603.
Jones, N.E., and B. Schmidt. 2012. ThermoStat Version 3.0: Tools for Analyzing Thermal
Regimes. Ontario Ministry of Natural Resources, Aquatic Research and Development
Section. Available online: http://people.trentu.ca/nicholasjones/thermostat.htm.
Kennedy, E.J. 1984. Discharge Ratings at Gaging Stations. Techniques of Water Resource
Investigations of the USGS, Book 3: Applications of Hydraulics, Chapter A10. Available
online: http://pubs.er.usgs.gov/publication/twri03A10.
Kennen, J.G., D.J. Sullivan, J.T. May, A.H. Bell, K.M. Beaulieu, andD.E. Rice. 2011. Temporal
changes in aquatic-invertebrate and fish assemblages in streams of the north-central and
northeastern U.S. Ecological Indicators 18:312-329.
MDDNR (Maryland Department of Natural Resources), (nd). Quality Assurance Document for
Temperature Monitoring. Draft Version 1.3. Available online:
http://www.dnr.state.md.us/streams/pdfs/Q A_TemperatureMonitoring.pdf.
Mauger, S. 2008. Water temperature data sensor protocol for Cook Inlet salmon streams. Cook
Inletkeeper, Homer, Alaska. 10 p. Available online:
http://inletkeeper.org/resources/contents/water-temperature-data-sensor-protocol.
Mazor, R.D., A.H. Purcell, and V.H. Resh. 2009. Long-term variability in benthic
macroinvertebrate bioassessments: A 20-year study from two northern Californian
streams. Environmental Management 43:1269-1286.
McMahon, G., J.D. Bales, J.F. Coles, E.M.P. Giddings, and H. Zappia. 2003. Use of stage data to
characterize hydrologic conditions in an urbanizing environment. Journal of the
American Water Resources Association 39:1529-1536.
68
-------
Meyer J.L., D.L. Strayer, J.B. Wallace, S.L. Eggert, G.S. Helfman, andN.E. Leonard. 2007. The
contribution of headwater streams to biodiversity in river networks. Journal of the
American Water Resources Association 43:86-103.
Onset Computer Corporation. 2014. A Quick Temp Accuracy Check (Ice Bath) [webpage].
Accessed 8 August 2014. Available online: http://www.onsetcomp.com/support/tech-
notes/quick-temp-accuracy-check-ice-bath.
Poff, N. L., J.D. Allan, M.B. Bain, J.R. Karr, K.L. Prestegaard, B.D. Richter, R.E. Sparks, and
J.C. Stromberg. 1997. The natural flow regime: a paradigm for river conservation and
restoration. BioScience 47(ll):769-784.
Rantz, S.E. et al. 1982. Measurement and Computation of Streamflow, Volume I: Measurement
of Stage and Discharge and Volume II: Computation of Discharge. USGS Water Supply
Paper 2175. Available online: http://water.usgs.gov/pubs/wsp/wsp2175/.
Roy, A.H., M.C. Freeman, BJ. Freeman, SJ. Wenger, W.E. Ensign, and J.L. Meyer. 2005.
Investigating hydrologic alteration as a mechanism offish assemblage shifts in
urbanizing streams. Journal of the North American Benthological Society 24:656-678.
Schuett-Hames, D., A.E. Pleus, E. Rashin, and J. Matthews. 1999. TFW Monitoring Program
Method Manual for the Stream Temperature Survey. Washington State Department of
Natural Resources and NW Indian Fisheries Commission publication #TFW-AM9-99-
005.
Selker, J. S., L. Thevenaz, H. Huwald, A. Mallet, W. Luxemburg, N. van de Giesen, M. Stejskal,
J. Zeman, M. Westhoff, and M. B. Parlange (2006), Distributed fiber-optic temperature
sensing for hydrologic systems, Water Resources Research 42, W12202,
JO/.-10.1029/2006WR005326.
Shedd, J. R. 2011. Standard Operating Procedure for Measuring and Calculating Stream
Discharge. Washington State Department of Ecology, Environmental Assessment
Program. Available online: http://www.ecv.wa.gov/programs/eap/quality.html.
Shedd, J.R., and C. Springer. 2012. Standard Operating Procedures for Correction of Continuous
Stage Records Subject to Instrument Drift, Analysis of Instrument Drift, and Calculation
of Potential Error in Continuous Stage Records. Washington State Department of
Ecology, Environmental Assessment Program - EAP No. EAP082. Available at:
http://www.ecy.wa.gov/programs/eap/qa/docs/ECY_EAP_SOP_CorrectionOfContinuous
StageRecords_vl_OEAP082.pdf.
Sowder, C., and E.A. Steel. 2012. A note on the collection and cleaning of water temperature
data. Water 4:597-606.
Stanford J.A., and J.V. Ward. 1988. The hyporheic habitat of river ecosystems. Nature 335:64-
66.
69
-------
Torgersen, C., et al. 2001. Airborne thermal remote sensing for water temperature assessment in
rivers and streams, Remote Sensing Environment 76(3), 386-398, doi:10.1016/S0034-
4257(01)00186-9.
U.S. EPA (U.S. Environmental Protection Agency). 2014. Regional Monitoring Networks
(Internal Review Draft). National Center for Environmental Assessment, Office of
Research and Development, Washington, DC; EPA/600/R-14/xxx. Available from the
National Technical Information Service, Springfield, VA, and online at
http://www.epa.gov/ncea.
U.S. Forest Service, Rocky Mountain Research Station. 2014. Stream Temperature Modeling
and Monitoring [webpage]. Accessed 8 August 2014. Available online:
http://www.fs.fed.us/rm/boise/AWAE/projects/stream_temperature.shtml
Vannote, R.L., G.W. Minshall, K.W. Cummins, J.R. Sedell, and C.E. Cushing. 1980. The River
Continuum Concept. Canadian Journal of Fisheries and Aquatic Sciences 37:130-137.
Wagner, R.J., R.W. Boulger, Jr., C.J. Oblinger, and B.A. Smith. 2006. Guidelines and standard
procedures for continuous water-quality monitors—Station operation, record
computation, and data reporting: U.S. Geological Survey Techniques and Methods 1-D3,
51p.+8attachments; accessed August 30, 2011, at http://pubs.water.usgs.gov/tmld3.
Ward, W. 2003. Continuous Temperature Sampling Protocols for the Environmental Monitoring
and Trends Section. Publication No. 03-03-052. Washington State Department of
Ecology Environmental Assessment Program. Available online:
https://fortress. wa.gov/ecy/publications/publications/0303052.pdf
Ward, W. 2011. Standard Operating Procedures for Continuous Temperature Monitoring of
Fresh Water Rivers and Streams, version 1. Washington State Department of Ecology
Environmental Assessment Program. Available online:
http://www.ecy. wa.gov/programs/eap/qa/docs/ECY_EAP_SOP_Cont_Temp_Mon_Ambi
ent_vl_OEAP080.pdf.
Wilde, F.D. 2006. Temperature (ver. 2.0): U.S. Geological Survey Techniques of Water-
Resources Investigations, book 9, chap. A6., sec. 6.1, March 2006, accessed September 6,
2011, from http://pubs.water.usgs.gov/twri9A6/.
70
-------
Appendix A
How to construct PVC housings for water temperature
sensors with cable installations
A-l
-------
Housing Construction
PVC housings are very simple to make, are inexpensive ($10-$ 15), provide
total shade for the logger, protect the logger from moving debris, and provide
for secure attachment with a cable.
The data logger is suspended in a PVC pipe that allows stream water to flow
through but prevents the penetration of solar radiation. Black PVC provides
more camouflage than white PVC for sites where vandalism is a concern. In
clear water streams, white PVC is recommended because black PVC has a
tendency to absorb heat from the daylight.
Materials needed to assemble housing:
2" Schedule 40 ABS pipe (I1 length)
2" DWV clean-out plug
(2) 2" DWV female adaptor
(2) 3/8-" x 4" ZC eye bolt
(1)8" cable ties
Multipurpose cement
Assorted nuts and bolts
Drill
Using the multipurpose cement, attach the female adapters to each end of the
PVC pipe. Drill a 3/8" hole for one eyebolt to go through a clean-out plug.
Drill at least 20 holes in the PVC to allow water flow. Secure the eye bolt
through the cleanout plug with a nut and bolt. Insert a cable tie through two
drilled holes to suspend the data logger in the housing. Additional cable ties
can be used to secure rocks in the bottom of the housing that might be needed
to weigh it down. Screw the cleanout plugs into the female adapters.
Reference: Mauger, S. 2008. Water temperature data sensor protocol for Cook
Inlet salmon streams. Cook Inletkeeper, Homer, Alaska; 10 p. Available online
at https://inl etkeeper.org/resources/contents/water-temperature-data-sensor-
protocol.
A-2
-------
Appendix B
How to make and install homemade radiation shields for
air temperature sensors
B-l
-------
Air Temperature Sensor Installation Instructions
Zack Holden; email: zaholden@fs.fed.us (406) 274.6766
Instructions for building the radiation shield can be found at: www.youtube.COm/watch?v=LkVmJRsw5vs
STEP 1: Mount radiation shield cover
The radiation shield should be installed at a height of 6 feet on the north-facing side of a large (> 12
inch diameter) tree. A coniferous tree that does not lose its leaves in winter is preferred. Bend and
fold the triangular tabs of the radiation shield cover downward to create angled "wings". Nail the
cover to the tree using aluminum nails (FIGURE 1). Hammer nail until nail head touches plastic, or
for trees with thin bark, bend the nail sideways against the plastic shield to hold the shield firmly
against the tree. Make sure the nail goes through both the triangular tab and rectangular back
cover. This will give the shield extra support. Note that the angle of the cover wings can be adjusted
depending on the size of the tree. The cover should be approximately perpendicular to the bole of the
tree (level to the ground) as shown in FIGURE 2. However, when mounted on smaller trees, the
shield will naturally bend downward, point toward the ground, which is not a problem.
STEP 2: Record Logtag sensor ID number
Each Logtag temperature sensor has a unique 10 digit ID
number located just above the barcode sticker (FIGURE 3)
Record the ID number and the GPS location on a
datasheet (Lat /Long or UTM X/Y, include the DATUM).
Sensor ID: 9900301054
STEP 3: Secure Logtag to the 2-plate shield
You will see two holes drilled into the top piece of the 2-plate shield. Using a 4 or 6 inch zip tie, insert
the zip tie through the top piece, through the logtag hole, then back up through the top piece. Secure
the zip tie. Make sure to close the zip tie AS TIGHTLY AS POSSIBLE. The top corner of the logtag must
be pulled snugly against the top of the plastic cover. The opposing corner should hang down and
touch the bottom piece of the 2-plate shield. It is critical to minimize direct surface contact between
the sensor and the plastic shield pieces and to allow as much air flow as possible past the sensor.
B-2
-------
STEP 4: Hang shielded sensor beneath the cover
Using 8 inch zip ties, suspend the 2-plate shield (logtag sensor is now housed inside) from the large
mounted radiation shield cover as follows: push the 8 inch zip ties through the pre-drilled holes in
the top cover and loop them through the 6 inch zip ties that are holding the 2-plate shield together
(FIGURE 4 and FIGURE 5).
STEP 4: Close and adjust radiation shield
Using an 8 inch zip tie, secure the front of the 2-plate shield to the front of the shield cover (FIGURE
6) as follows: insert the zip tie through the top of the cover, downward through each of the two
plates below, and then bring it back up and close the zip tie. It is important that there be adequate
space between the 2 plates that surround the temperature sensor to allow air flow. Use your hands
to create a gap between the two lower plates. Adjust as necessary, using the friction of the zip tie
to maintain space between the 2 lower plates housing the air sensor. There should be
approximately a 1.25 to 1.5 inch gap between the two plates when you're finished. The shield
should be approximately parallel with the ground and sit approximately 6 feet (or higher) off the
ground. When completed, it should look like FIGURE 6.
B-3
-------
Appendix C
Forms for temperature sensor accuracy checks
c-i
-------
Table Cl. Form for doing temperature sensor accuracy checks. The temperature sensors and field thermometers are checked against a NIST-certified
thermometer.
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Initials
Date
Time
Temperature (°C)
NIST
Thermometer
S/N-
Field
Thermometer
S/N-
Field
Thermometer
S/N-
Sensor
S/N-
Sensor
S/N-
Sensor
S/N-
Sensor
S/N-
Sensor
S/N-
C-2
-------
Table C2. Form for checking mean temperature sensor readings against mean temperature from a NIST-certified thermometer.
Technician:
Date/s of accuracy check:
Overall mean temperature (NIST-certified thermometer) (°C):
Sensor S/N
Mean
Temperature
(°C)
Difference from
NIST-certified
thermometer (°C)
Minimum
Temperature
(°Q
Maximum
Temperature
(°Q
Battery
Level
(Pass/Fail)
Accuracy
Check
(Pass/Fail)
Notes
C-3
-------
Table C3. Form for checking temperature sensor readings against one another.
Technician:
Date/s of accuracy check:
Overall mean temperature (all sensors) (°C):
Overall minimum temperature (all sensors) (°C):
Overall maximum temperature (all sensors) (°C):
Sensor S/N
Mean
Temperature
(°Q
Difference
from
overall
mean (°C)
Minimum
Temperature
(°C)
Maximum
Temperature
(°C)
Notes
C-4
-------
Appendix D
Temperature sensor deployment and tracking forms
(fill out one for each sensor)
D-l
-------
S/N: Station number:
Water or air: Waterbody name:
Programmed launch date & time: GPS coordinates for sensor
Deployment date and time: Longitude (dec. degrees):
Expected retrieval date: Latitude (dec. degrees):
Battery type: Datum:
Configuration Filenames of photos:
Recording interval:
Units:
Start time:
Installation technique (e.g., underwater epoxy, rebar):
Description of location (e.g. type of structure it is attached to; type of habitat (riffle, run))
Hand-drawn map (and/or attach print out of annotated photo)
D-2
-------
Task
Pre -deployment accuracy check
Launch/start
Deployment (in position)
Mid-deployment accuracy check
(in Notes, write down NIST-
calibrated field thermometer
reading; then compare this to the
sensor reading)
Data download
Sensor Retrieval
Post-deployment accuracy check
Date
Time
Technician
Notes
Comments
D-3
-------
Appendix E
Equipment lists for temperature sensor procedures
E-l
-------
These are equipment lists for various phases of deployment. These lists do NOT include health
and safety equipment or personal gear like waders.
Equipment needs for the pre-deployment accuracy check
Task
Performing the
accuracy check
Recording
measurements
Supplies List
• Temperature sensor/s
• National Institute of Standards and Technology (NIST) traceable
or calibrated reference thermometer with an accuracy of ±0.2°C
• Field thermometers (optional)
• Containers to hold the sensors
• Water
• Refrigerator
• Clock or watch
• Accuracy check data sheet (see sample forms in Appendix C)
• Computer that has the appropriate software for reading the
temperature sensor
Equipment list for doing underwater epoxy installations (Isaak et al. 2010).
Task Supplies list
Underwater
epoxy
installation
Temperature sensor
Radiation shield (PVC canister, 1-1/2" with screw top, mid-
section and base)
Underwater epoxy (FX-764 Splash Zone Epoxy)
Jars for mixing the epoxy
Underwater viewing box
Lead weights, 1A oz
Neoprene, 3mm
Rubber gloves
Wire brush
Zip ties, 4"
Metal mirror
E-2
-------
Equipment list for doing cable installations (Mauger 2008, Ward 2011).
Task
Cable
installation
Supplies list
• Temperature sensor
• Radiation shield (example in Appendix A)
• Cable or heavy duty cable ties (e.g., 120-lb tensile strength)
• Rebar (3/8 to 1/2", 2-4 ft lengths)
• Rebar pounder (for custom design, see Ward 201 1)
• For more information, reference Ward 201 1 and Mauger
2008
Equipment list for documenting sites.
Task
Supply list
Georeferencing
& monumenting
GPS
Camera
Map and/or gazetteer
Metal forestry tags
Field form (see sample form in Appendix D)
Spray paint (optional)
Equipment list for conducting maintenance/mid-deployment checks and data downloads.
Task
Documenting
on-site-
conditions
Data offloads
Measuring
temperature for
accuracy check
Measuring depth
(optional)
Back-up
equipment (in
case a sensor
needs to be
replaced)
Supply list
• Camera
• Field form (see sample form in Appendix G)
• Base station or portable shuttle
• Laptop (if practical) & data back-up device (e.g., flash drive)
• NIST-calibrated field thermometer and/or multi-probe meter
• Field form (see sample form in Appendix G)
• Meter ruler or calibrated rod/pole (e.g., surveyor's rod)
• Measuring tape
• Field form (see sample form in Appendix G)
• Replacement sensors (that have passed accuracy and battery
checks) and other necessary deployment equipment
E-3
-------
Literature Cited
Mauger, S. 2008. Water temperature data sensor protocol for Cook Inlet salmon streams. Cook
Inletkeeper, Homer, Alaska. 10 p. Available online at
http://inletkeeper.org/resources/contents/water-temperature-data-sensor-protocol.
Ward, W. 2011. Standard Operating Procedures for Continuous Temperature Monitoring of
Fresh Water Rivers and Streams, version 1. Washington State Department of Ecology
Environmental Assessment Program. Available online at
http://www.ecy.wa.gov/programs/eap/qa/docs/ECY EAP SOP Cont Temp Mon Ambient v 1
OEAP080.pdf.
E-4
-------
Appendix F
Examples of alternate installation techniques
F-l
-------
Rebar Method (Mauger, 2008)
This method is preferred for streams with moderate movement of the streambed during high
flows. The protective case or PVC housing is attached by a cable to a rebar stake sunk 3 feet
into the stream bed.
Secure an eyebolt to the rebar (approximately 1 foot from an end) with hose clamps. Secure
one end of the cable to the eyebolt with a wire rope clip. Secure the other end to the
protective case or PVC housing using a wire rope clip. Use a stake pounder to sink the rebar
3 feet into the stream bottom near a large rock or other landmark.
Materials:
1/2" rebar (4' length)
3/8" x 4" ZC eyebolt (1)
9/16-1 1/16 hose clamp (3)
1/8" wire rope clip (2)
1/8" RL uncoated cable (T length)
Stake pounder
Stream Bank-Secured Cable Method (Mauger, 2008)
This method is preferred for streams with significant movement of the streambed during high
flows. In this method, the logger in its protective case or PVC housing is secured to the
stream bank vegetation using plastic-coated wire rope.
The logger is secured to the wire rope (1/8" to 3/8" diameter and 12
feet long) using a wire rope clip. Upon deployment, the cable is
wrapped around a large tree, rocks, bridge supports, or other secure
object within or on the stream bank. The logger is then placed within
the stream channel. Large stream rocks can be placed on top of the
cable near the logger to hold the logger in place within the stream.
The cable should be hidden under bank vegetation to avoid
vandalism or accidental disturbance. Try to avoid locations where
the cable will cross active fishing or wildlife trails.
Sand Bag Method (Mauger, 2008)
This method is preferred for streams with minimal movement of the streambed during
high flow events. Sturdy sand bags can be purchased at most hardware stores. Fill the
bag on-site with any mineral material (large rocks, cobbles, or sand). Avoid organic
material which is often buoyant. The logger, in its protective case or PVC housing, can
be attached to the bag by weaving a cable tie through the mesh. The bag can be tied off
with a rope to the stream bank for extra security. The rope should be hidden under bank
vegetation to avoid vandalism or accidental disturbance. Try to avoid locations where
the rope will cross active fishing or wildlife trails.
Delaware River Basin Commission (DRBC) Low Profile Concrete Base Technique
DRBC uses this technique for year-round deployment of conductivity, temperature, and dissolved
oxygen sensors. Some of their sensors have been deployed for more than 2 years with minimal
maintenance. Their general protocol is as follows:
F-2
-------
To make the concrete base, pour 40-lbs of mixed concrete into a saucer snow sled (the sled is
used as a mold and can be reused). Add some metal to the concrete to provide mounting
points for the protective shield.
Making a concrete base for sensor deployment. Photo taken by John
Yagecic
Using metal straps, mount the protective shield to the concrete base.
Attach the sensor to the protective shield with zip ties. The sensor is affixed to the side of
the shield such that the sensor face is held in the vertical plane. This is consistent with the
manufacturer's recommendations and reduces build-up of silt or bubbles on the sensor
face.
Photo provided by the Delaware River Basin Commission
F-3
-------
When placing the unit in the water, orient the base in such a way that flowing water will encounter
the sensor first. This should reduce any mild impacts to water chemistry from water flowing over
the shield or base. The end of the shield that points upstream should protrude a little further than the
downstream end.
Deployed sensor. Photo taken by Don Hamilton
Data are downloaded and conductivity and temperature sensors are cleaned about once every 3
weeks.
Manufacturer's recommendations for pre- and post-deployment data adjustment are followed.
For more information, contact John Yagecic (John.Yagecicfgidrbc.state.nj.us) and Eric Wentz
(Eric.Wentzfgjdrbc.state.nj.us) from the DRBC.
F-4
-------
Appendix G
Mid-deployment/maintenance check form
(fill out one for each mid-deployment site visit; information for both the air and
water temperature sensor at the site can be recorded on this form)
G-l
-------
Date:
Time:
Technicians:
Data offload (yes/no):
Battery life:
Station Number:
Waterbody Name:
Water temperature S/N:
Air temperature S/N:
Photo file names:
MAINTENANCE CHECK
Are there any signs of physical damage, vandalism or disturbance? If yes, describe.
Is the water temperature sensor dewatered? If yes, describe.
Is the water temperature sensor buried in sediment? If yes, describe.
Is there evidence of fouling (i.e. debris, aquatic vegetation, algae)? If yes, describe.
ACCURACY CHECK
Temperature measurements should be taken with a National Institute of Standards and
Technology (NIST) traceable or calibrated reference thermometer.
NIST traceable or calibrated reference thermometer S/N:
Equipment
Field
Thermometer
Sensor
Water
Time
Temperature
(°Q
Air
Time
Temperature
(°Q
G-2
-------
Appendix H
Post-retrieval QA/QC procedures for temperature sensor
data
H-l
-------
This appendix is organized into five sections:
1. Record keeping and data storage
2. Data evaluation
3. Application of data corrections
4. QA/QC checklist form for temperature sensor data
5. References
HI Record keeping and data storage
• Make sure you set up a good record keeping and data storage system Large amounts
of data will accumulate quickly so a central temperature database should be developed
and maintained from the initial stages of monitoring. Also, all field and accuracy check
forms should be organized, easily accessible, and archived in a way that allows for safe,
long-term storage.
• Original raw data files should be retained for all sites, and should be kept separate
from files in which data have been manipulated. The data should be accessible, since
someone may want to go back and calculate different metrics in the future.
• Carefully document these steps as well as any changes that you make to the data. The
checks can be conducted using a number of different software packages (e.g., Microsoft
Excel, Hoboware, Aquarius).
H2 Data evaluation
Recommended steps:
1. Save the file that you are manipulating with a different file name so that you don't
confuse it with the original raw data file.
2. Format the data so that it is easy to work with.
3. Trim data (as necessary) to remove measurements taken before and after the sensor is
correctly positioned.
4. Plot all of the measurements and visually check the data. Look for missing data and
abnormalities. Consider doing the following, as data permits:
• Plot air and stream temperature data on the same graph.
• Plot stream temperature data with stage data.
Specific things to watch for:
• Missing data.
• A close correspondence between water and air temperature - this indicates
that the stream sensor may have been out of the water.
H-2
-------
• Diel fluxes with flat tops - this indicates that the sensor may have been buried in
sediment.
Optional
• Graphically compare data across sites
• Graphically compare data across years; when data from one year are
dramatically different, there may be errors.
• Graphically compare with data from the nearest active weather station, if
appropriate. The closest active weather stations can be located and the daily
observed air temperature data for those stations can be downloaded from websites
like the Utah State University Climate Server:
http://climate.usurf.usu.edu/mapGUI/mapGUI.php.
Additional checks (also optional)
• If using MS Excel, use pivot tables to check for missing data. If a 30-minute
interval is used, there should be 48 measurements per day. If there are fewer (or
more) measurements than expected, check the original data and your field notes
and try and determine what might have caused this to occur.
• Flag data points for potential errors if they
o exceed a thermal maximum of 25°C*
o exceed a thermal minimum of -1°C*
o exceed a daily change of 10°C*
o exceed the upper 5th percentile of the overall distribution
o fall below the lower 5th percentile of the overall distribution
* These values should be adjusted to thermal limits appropriate for each location.
H3 Application of data corrections
Erratic readings with temperature sensors can occur for a number of reasons:
• They may become dewatered during low flow conditions.
• High flow events may bury them in sediment.
• High flow events may move them.
• They may become fouled from debris, aquatic vegetation, or algae.
• Humans may cause interference.
Errors should be addressed on a case-by-case basis. In general, there are three possible actions:
1) Leave the data as is.
2) Apply the correction factor.
3) Remove the data.
If you are inexperienced at addressing errors with continuous temperature data, consider seeking
guidance from someone with more experience and consult references like Wagner et al., 2006.
Table 14 in the report provides a general summary of the different types of problems that can
occur (e.g., missing data, failed accuracy check) and recommended actions for addressing them.
H-3
-------
H4 QA/QC checklist form for temperature sensor data
Technician conducting QA/QC check:
Sensor S/N:
Water or air:
Station number:
Waterbody name:
Data trimming
Programmed launch/start
Deployment start (in
position)
Out of position (e.g., data
download, maintenance)
Retrieval
Date
Time
Technician
Accuracy checks
Pre -deployment
Mid-deployment
Post-deployment
Pass/fail
Notes
Describe how you addressed questionable data
Did you save the original data (pre-'cleaned1)? If so, what is the file name and where is it located?
What is the file name of the 'cleaned data' and where is it located?
Did you back up both the original and 'cleaned' data files? If so, where are the back-up files
located?
H-4
-------
Checklist
Consistent units - did you check the units to
ensure that they were consistent throughout the
period of deployment (should be °C)?
Missing data - in the Notes field, describe how
you addressed these data and why you think they
occurred.
Trimming data - did you remove observations
that were recorded when the sensor was NOT
correctly positioned?
Graph - did you plot all individual data points and
look for abnormalities?
Graph - did you plot air and stream temperature
on the same graph and look for abnormalities?
Graph - did you plot stream temperature and
water level (if available) on the same graph and look
for abnormalities?
Graph - did you graphically compare data across
sites?
Graph - did you graphically compare data from the
same site across different years?
Graph - did you graphically compare with data from
the nearest active weather station?
Flags - did you flag values if they met the conditions
below? (If you used different thermal limits, cross out
the values below and write what you used)
Exceed a thermal maximum of 25 °C
Exceed a thermal minimum of -1°C
Exceed a daily change of 10°C
Exceed the upper 5th percentile of the overall
distribution
Fall below the lower 5th percentile of the
overall distribution
Yes/No
Notes
H-5
-------
H5 References
Dunham, J., G. Chandler, B. Rieman, and D. Martin. 2005. Measuring stream temperature with digital
data sensors: a user's guide. Gen. Tech. Rep. RMRSGTR-150WWW. Fort Collins, CO: U.S. Department
of Agriculture, Forest Service, Rocky Mountain Research Station. 15 p. Available online at
http://fresc.usgs.gov/products/papers/1431 Dunham.pdf.
Maryland Department of Natural Resources, (nd). Quality assurance document for temperature
monitoring. Draft Version 1.3. Available online at
http://www.dnr.state.md.us/streams/pdfs/QA_TemperatureMonitoring.pdf
Radtke, D.B., Kurklin, J.K., and Wilde, F.D., eds. 2004. Temperature (version 1.2): U.S. geological
survey techniques of water-resources investigations. Book 9, chap. A6, section 6.1, 15 p.; accessed July
28, 2004, Available at http://pubs.water.usgs.gov/twri9A6/ (accessed July 28. 2004).
Shull, D. and J. Lookenbill. 2013. Bureau of point and non-point source management continuous instream
monitoring protocol. Prepared by the PA Department of Environmental Protection, Harrisburg, PA.
Sowder, C., and E. A. Steel. 2012. A note on the collection and cleaning of water temperature data. Water
4: 597-606.
Wagner, R.J., Boulger, R.W., Jr., Oblinger, C.J., and Smith, B.A. 2006. Guidelines and standard
procedures for continuous water-quality monitors—Station operation, record computation, and data
reporting: U.S. Geological Survey Techniques and Methods l-D3,51p. plus 8 attachments. Available at
http://pubs.water.usgs.gov/tm 1 d3 (accessed August 30, 2011).
Wilde, F.D. 2006. Temperature (ver. 2.0): U.S. geological survey techniques of water-resources
investigations. Book 9, chap. A6., sec. 6.1, March 2006. Available at http://pubs.water.usgs.gov/twri9A6/
(accessed September 6, 2011).
-------
Appendix I
Equipment checklists for pressure transducer protocols
1-1
-------
Equipment list for staff gage installation.
Location
All
Fixed Object -
additional items
Streambed -
additional items
Supplies list
• Staff gage
• Gage board
• Screws (stainless steel or brass)
• Screwdriver
• Assorted drill bits for wood and PVC
• Level
• Stepladder (if installing more than one section of a staff gage)
• Datasheets and field notebook
• Survey equipment (auto level or laser level
marker or nails for marking benchmarks)
• Concrete wedge anchor and nuts (stainless
• Hammer or rotary drill
• Concrete drill bit
• Hammer
and stadia rod, paint
steel or galvanized)
• Galvanized or stainless steel pole with cap
• Galvanized or stainless steel conduit straps
• Pole driver or sledge hammer
• Pry bar
1-2
-------
Equipment list for transducer installation.
Location
All
Fixed object -
additional items
Supplies list
• PVC pipe for transducer (drilled with !/>" or larger holes)
• PVC pipe for data logger or barometric pressure logger
• Galvanized or stainless steel conduit straps, hangars, and
• Zip ties/cable ties
• Screws (stainless steel or brass)
• Screwdriver
• Drill
• Assorted drill bits for wood and PVC
• Level
• Datasheets and field notebook
• Survey equipment (see Section 4.3.5).
If using a non-vented transducer, you also need:
• Non-stretch cable or wire
• Wire rope clamps
• Long bolt and wing nut
hose clamps
• Extra-long PVC (should extend out of water in high flows)
• Solar shield
• PVC with caps for barometric transducer
If using a vented transducer, you also need:
• Garden staples
• PVC pipe for data logger
• PVC cap or locking well cap
• Stainless steel conduit straps
• Long lag screws
• Wire cable
• Hammer or rotary drill
• Concrete drill bit
• Concrete wedge anchor and nuts (stainless steel or galvanized)
• Hammer
Equipment list for elevation surveys.
Supplies list
Auto level or laser level
Tripod
Stadia rod
Survey paint
Survey nails
Datasheet
1-3
-------
Equipment list for taking discharge measurements.
Supplies list
• Current meter/s
• Wading rods
• Measuring tape and stakes
• Discharge measurement form
Equipment list for generating a profile of the channel cross-section.
Supplies list
• Automatic level or electronic total station and tripod or a
clinometer
• Stadia rod
• Measuring tape and stakes
1-4
-------
Appendix J
Field forms for water level and flow measurements
j-i
-------
Site:_
Crew:_
Notes:
Elevation/Staff Gauge Survey Sheet
Date:
Gage (ft):.
BS = backsite, HI = height of instrument. FS = foresite, BM = benchmark. TP = turning point
H.I. = B.S. + Elevation, Elevation = H.I. -F.S.
BM#
Location
BS Hl(+) FS(-) Elevation
Notes and Site Sketch:
J-2
-------
Transducer Download Field Data Sheet
Site:
Date:
Crew:
Weather:
Gage {ft.}:
Photos taken?
Battery Status:
File name:
Notes:
Time:
Transducer (ft):
Gage Survey?
Batteries Changed?
J-3
-------
Discharge Measurement Field Data Sheet - Marsh McBirney
Site: _
Date:
Crew:
Weather: ~
Start time:
Gage before (ft.):
Equipment:
Gage Survey?
Notes:
Finish time:
Gage after (ft):
Photos taken?
Point
1
2
3
4
5
G
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Tape
Distance
(ft)
[Cell
width
<«)]
Depth
(ft)
[Velocity
ffi»)]
[Q (cfs)]
Tola! Q (cfs):
% of Q
Notes
J-4
-------
Discharge Measurement Field Data Sheet - Pygmy/Price Meter
Site:
Date:
Crew:
Weather: ^^
Start time:
Spin before (seconds):
Gage before {ft.}:
Equipment:
Gage Survey?
Finish time:
Spin after (seconds):
Gage after (ft):
Photos taken?
Point
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Tape
Distance
(ft)
[Cell
width
(ft)]
Depth
(ft)
Clicks/
40 sec
[Velocity
-------
Appendix K
Post-retrieval QA/QC procedures for pressure transducer
data
K-l
-------
This appendix is organized into five sections:
1. Record keeping and data storage
2. Data evaluation
3. Application of data corrections
4. QA/QC checklist form for pressure transducer data
5. References
Kl Record keeping and data storage
• Make sure you set up a good record keeping and data storage system Large amounts
of data will accumulate quickly so a central temperature database should be developed
and maintained from the initial stages of monitoring. Also, all field and accuracy check
forms should be organized, easily accessible, and archived in a way that allows for safe,
long-term storage.
• Original raw data files should be retained for all sites and should be kept separate
from files in which data have been manipulated. The data should be accessible, since
someone may want to go back and calculate different metrics in the future.
• Carefully document these steps as well as any changes that you make to the data. The
checks can be conducted using a number of different software packages (e.g., Microsoft
Excel, Hoboware, Aquarius).
K2 Data evaluation
1. Save the file that you are manipulating with a different file name so that you don't
confuse it with the original raw data file.
2. Format the data so that it is easy to work with.
3. Trim data (as necessary) to remove measurements taken before and after the pressure
transducer is correctly positioned.
4 Plot all stage measurements and visually check the data Look for missing data and
abnormalities
Specific things to watch for -
• Missing data.
• Values of 0 - this could mean that the pressure transducer was dewatered.
Another possibility (with vented transducers) is that moisture got into the cable
and caused readings of zero water depth.
• Values flat-lining at 0°C/32°F - the stream pressure transducer is likely encased
in ice.
K-2
-------
• Negative values - if unvented transducers are being used, this may indicate that
the barometric pressure correction is off. This could occur for a number of
reasons:
o The land-based transducer is not close enough to the stream pressure
transducer to accurately capture barometric pressure.
o If the land-based transducer is housed in PVC pipe that has a solid bottom,
condensation and laterally blown rain and snow can penetrate through the
drilled holes and collect in the bottom, filling the pipe to a depth sufficient
to inundate the ports through which the barometric pressure is
compensated. Thereafter, "barometric pressure" is actual barometric
pressure plus a small amount of pressure due to this accumulated water.
• Outliers or rapidly fluctuating values - the stream pressure transducer may
have moved (e.g., due to a high flow event or vandalism).
In addition, consider doing the following, as data permits -
• Plot stage and temperature data on the same graph. Watch for the following
signals in the temperature data:
o Diel fluxes with flat tops - this indicates that the pressure transducer may
have been buried in sediment.
o A close correspondence between water and air temperature (if air
temperature data are available) - this indicates that the pressure transducer
may have been out of the water.
Optional
• Graphically compare data across years; when data from one year are
dramatically different, there may be errors.
• Graphically compare with precipitation data from the nearest active weather
station, if appropriate. The closest active weather stations can be located and the
daily observed precipitation data for those stations can be downloaded from
websites like the Utah State University Climate Server:
http://climate.usurf.usu.edu/mapGUI/mapGUI.php.
• Graphically compare with data from the nearest USGS stream gage, if
appropriate. The closest active USGS gage can be located and the daily flow data
for those gages can be downloaded from the USGS National Water
Information System (NWIS) website:
http://waterdata.usgs.gov/usa/nwis/rt.
K3 Application of data corrections
Erratic readings with pressure transducers can occur for a number of reasons:
• They may become dewatered during low flow conditions.
• High flow events may bury them in sediment.
• High flow events may move them.
• They may become fouled from debris, aquatic vegetation, or algae.
• Humans may cause interference.
• They may become encased in ice.
K-3
-------
• If moisture gets into the cable of a vented transducer, it may result in erratic readings or
readings of zero water depth.
• If the cable of a vented transducer gets kinked or plugged, it can result in the data not
being corrected for barometric pressure.
Errors should be addressed on a case-by-case basis. In general, there are three possible actions:
1) Leave the data as is.
2) Apply the correction factor.
3) Remove the data.
If you are inexperienced at addressing errors with continuous temperature data, consider seeking
guidance from someone with more experience and consult references like Wagner et al., 2006.
Table 20 in the report provides a general summary of different types of problems that can occur
(e.g., missing data, failed accuracy check) and recommended actions for addressing them.
K-4
-------
K4 QA/QC checklist form for pressure transducer data
Technician conducting QA/QC check: Transducer S/N:
Station number: Waterbody name:
Data trimming
Programmed launch/start
Deployment start (in
position)
Out of position (e.g., data
download, maintenance)
Retrieval
Date
Time
Technician
Describe how you addressed questionable data.
Accuracy
checks
Date
Time
Staff gage
reading
Pressure transducer
reading
Did you save the original data (pre-'cleaned1)? If so, what is the file name and where is it located?
What is the file name of the 'cleaned data' and where is it located?
Did you back up both the original and 'cleaned' data files? If so, where are the back-up files
located?
K-5
-------
Checklist
Consistent units - did you check the units to
ensure that they were consistent throughout the
period of deployment?
Missing data - in the Notes field, describe how
you addressed these data and why you think they
occurred.
Values of 0 or negative values - in the Notes
field, describe how you addressed these data and
why you think they occurred.
Trimming data - did you remove observations
that were recorded when the pressure transducer
was NOT correctly positioned?
Graph - did you plot all individual data points and
look for abnormalities?
Graph - did you plot temperature data on the same
graph and look for abnormalities?
Graph - did you graphically compare data from the
same site across different years?
Graph - did you graphically compare with discharge
data from the nearest USGS stream gage?
Graph - did you graphically compare with
precipitation data from the nearest active weather
station?
Yes/No
Notes
K5 References
Shedd, J.R., and C. Springer. 2012. Standard operating procedures for correction of continuous stage
records subject to instrument drift, analysis of instrument drift, and calculation of potential error in
continuous stage records. Washington State Department of Ecology, Environmental Assessment Program
- EAP No. EAP082. Available at
http://www.ecv.wa.gov/programs/eap/qa/docs/ECY EAP SOP CorrectionOfContinuousStageRecords v
1 OEAP082.pdf.
Shull, D. and J. Lookenbill. 2013. Bureau of point and non-point source management continuous instream
monitoring protocol. Prepared by the PA Department of Environmental Protection, Harrisburg, PA.
Wagner, R.J., Boulger, R.W., Jr., Oblinger, C.J., and Smith, B.A., 2006. Guidelines and standard
procedures for continuous water-quality monitors—Station operation, record computation, and data
reporting: U.S. Geological Survey Techniques and Methods 1-D3, 51p. and 8 attachments. Available at
http://pubs.water.usgs.gov/tmld3 (accessed August 30, 2011).
K-6
-------
United States
Environmental Protection
Agency
National Center for Environmental Assessment
Office of Research and Development
Washington, DC 20460
------- |