EPA/600/R-13/170
External Review Draft
September 2013
Best Practices for Continuous Monitoring of Temperature
and Flow in Wadeable Streams
NOTICE
THIS DOCUMENT IS A PRELIMINARY DRAFT. It has not been formally
released by the U.S. Environmental Protection Agency and should not at this
stage be construed to represent Agency policy. It is being circulated for comment
on its technical accuracy and policy implications.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
-------
DISCLAIMER
This document is distributed solely for the purpose of pre-dissemination peer review under
applicable information quality guidelines. It has not been formally disseminated by EPA or the
U.S. Geological Survey. It does not represent and should not be construed to represent any
Agency determination or policy. 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 lack of continuous temperature and flow data for minimally disturbed, free-flowing
freshwater wadeable streams is an impediment to analyses of long-term trends in biological,
thermal, and hydrologic data. In recent years, there has been substantial interest in developing
Regional Monitoring Networks with states and EPA Regional offices to detect long-term climate
change-related impacts on aquatic communities in freshwater streams. Current participants,
including states in the northeast, mid-Atlantic and southeast, are initiating collection of thermal,
hydrologic and biological data from targeted sites in each state. To help further this effort, the
U.S. EPA and collaborators have written a guidance document to facilitate more uniform and
effective collection of continuous temperature and water depth data at ungaged sites in wadeable
streams. This document addresses questions related to equipment needs, configuration,
placement, installation techniques, data retrieval, and data processing. The collection of these
data will further efforts to detect and track climate change-related impacts over the long term,
further our understanding of how biological, thermal, and hydrologic conditions vary spatially
and temporally and inter-relate to one another, and help inform state and federal agencies on how
to attribute altered environmental conditions to climate change versus other stressors.
This document is a draft for review purposes only and does not constitute Agency policy.
ii DRAFT—DO NOT CITE OR QUOTE
-------
TABLE OF CONTENTS
TABLES vi
FIGURES vii
1. PUPPOSE AND SCOPE 1
2. TEMPERATURE 3
2.1. Equipment 3
2.1.1. Basic components 3
2.1.2. Considerations when choosing temperature sensors 4
2.1.3. Considerations when choosing portable data offload devices 8
2.1.4. Considerations when choosing radiation shields 9
2.2. Pre-Deployment 12
2.2.1. Calibration/accuracy check 12
2.2.2. Sensor configuration and launch 14
2.3. Deployment of Water Temperature Sensors 16
2.3.1. Guidelines for placement within the stream 16
2.3.2. Installation 17
2.3.2.1. Underwater epoxy 18
2.3.2.2. Cabling the sensor to rebar or stable instream structures 22
2.3.3. Documentation 23
2.3.4. Common problems 27
2.4. Deployment of Air Temperature Sensors 29
2.5. Maintenance/Mid-Deployment Checks and Data Offload 30
2.6. Quality Assurance and Control 33
2.6.1. Mid- and post-deployment accuracy checks 34
2.6.2. Error screening 34
2.6.3. Record keeping 36
3. STREAMFLOW 36
3.1. Equipment 36
3.1.1. Basic components 37
3.1.2. Considerations when choosing pressure transducers 37
3.1.3. Staff gage 42
3.1.4. Protective housing 42
3.2. Pre-Deployment 43
3.2.1. Calibration 43
777/5 document is a draft for review purposes only and does not constitute Agency policy.
iii DRAFT—DO NOT CITE OR QUOTE
-------
3.2.2. Sensor configuration and launch 44
3.3. Deployment of Instream Pressure Transducers 45
3.3.1. Selecting a location 45
3.3.2. Staff gage installation 46
3.3.3. Pressure transducer installations 50
3.3.3.1. Fixed object 50
3.3.3.2. Streambed/rebar 53
3.4. Deployment of On-Land Components 55
3.5. Elevation Surveys and Documentation 57
3.6. Maintenance/Mid-Deployment Checks and Data Offload 61
3.7. Quality Assurance and Control 63
3.7.1. Accuracy checks 63
3.7.2. Error screening 63
3.8. Developing Frames of Reference 64
3.8.1. Stage-discharge rating curves 65
3.8.1.1. When to measure discharge 65
3.8.1.2. Equipment 67
3.8.1.3. Site selection 67
3.8.1.4. Measurements 68
3.8.1.5. Documentation 69
3.8.1.6. Quality assurance and control 70
3.8.1.7. Making flow rating curves 70
3.8.2. Channel cross-section measurements and modeling 71
3.8.2.1. Equipment 71
3.8.2.2. Site selection 71
3.8.2.3. Measurements 72
3.8.2.4. Documentation 72
3.8.2.5. Quality assurance and control 72
3.8.2.6. Modeling 73
4. LITERATURE CITED 73
This document is a draft for review purposes only and does not constitute Agency policy.
iv DRAFT—DO NOT CITE OR QUOTE
-------
APPENDICES
A. How to construct PVC housings for water temperature sensors with cable installations A1
B. How to make and install homemade radiation shields for air temperature sensors B1
C. Temperature sensor calibration forms Cl
D. Temperature sensor deployment & tracking forms Dl
E. Equipment lists for temperature sensor procedures El
F. Examples of alternate temperature sensor installation techniques Fl
G. Temperature sensor mid-deployment check forms Gl
H. QA/QC checklist for temperature sensor data HI
I. Equipment lists for pressure transducer procedures II
J. Field forms for water level and flow measurements Jl
This document is a draft for review purposes only and does not constitute Agency policy.
v DRAFT—DO NOT CITE OR QUOTE
-------
TABLES
Table 1. Specifications for temperature sensors used in the RMNs 4
Table 2. Examples of commercially available temperature sensors 7
Table 3. Pre-deployment calibration procedure 13
Table 4. Equipment needs for the pre-deployment accuracy check 14
Table 5. Quick guide for doing underwater expoxy installations (Isaak et al. 2013) 19
Table 6. Equipment list for doing underwater epoxy installations (Isaak et al. 2013). Monumenting is
discussed in Section 2.3.3 21
Table 7. Guidelines for documenting the installation 25
Table 8. Equipment list for documenting sites 27
Table 9. Tips for minimizing the chance of sensors being lost or damaged 28
Table 10. Checklists for performing maintenance/mid-deployment checks and data downloads 32
Table 12. Error screening procedure (based on Sowder and Steel 2012 and Dunham et al. 2005) 35
Table 13. Examples of commercially available pressure transducers 41
Table 14. Equipment list for staff gage installation 49
Table 15. Equipment list for transducer installation 53
Table 16. Quick guide for elevation surveys of staff gages and transducers using an auto level 59
Table 17. Equipment list for elevation surveys 59
This document is a draft for review purposes only and does not constitute Agency policy.
vi DRAFT—DO NOT CITE OR QUOTE
-------
FIGURES
Figure 1. Examples of commercially available temperature sensors 6
Figure 2. Additional equipment is needed to offload data from the temperature sensors onto a computer.. 9
Figure 3. Stream temperature measurements from four sensors at the same site during eight days in July
2010 10
Figure 4. Unassembled components of an inexpensive PVC canister radiation shield for water temperature
sensors (described in more detail in Isaak et al. 2013) 10
Figure 5. Examples of radiation shields for air temperature sensors 11
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) 15
Figure 7. Examples of large rocks (a and b) and cement bridge pilings (c and d) that provide good sensor
attachment sites 20
Figure 8. Close-up of a PVC solar shield with a temperature sensor glued to a rock in a stream 20
Figure 9. Equipment needed to permanently install temperature sensors in streams using underwater
epoxy 22
Figure 10. Photos of a rebar installation (from Mauger 2008) 23
Figure 11. Example of a hand-drawn map from a field form used by the Washington State Department of
Ecology (taken from Ward 2011) 26
Figure 12. 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) 26
Figure 13. Examples of commercially available pressure transducers 40
Figure 14. Example of aUSGS style staff gage (Type A) marked in 0.02 foot increments 42
Figure 15. Example of a protective housing made of PVC pipe 43
Figure 16. Staff gage readings provide a quality check of transducer data 44
Figure 17. Examples of controls downstream of staff gages 46
Figure 18. Examples of gage installation techniques 48
Figure 19. Example of a 0.5 x 3.75 inch wedge anchor with bolt and washer 48
Figure 20. Non-vented pressure transducer installation 51
Figure 21. Non-vented (left) and vented (right) pressure transducers attached to staff gage board using
conduit hangars 52
Figure 22. Examples of streambed installations of pressure transducers using rebar 54
Figure 23. Vented transducer data logger installation 56
Figure 24. Barometric pressure sensor installation for a non-vented transducer 56
Figure 25. Example of auto level and tripod used for elevation survey (A) and of a permanent structure
used as a benchmark (B) 58
Figure 26. An example of a completed elevation survey form 60
Figure 27. 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) 62
Figure 28. Example of a transducer download field datasheet 62
Figure 29. A culvert replacement downstream of a stream gage on Gulf Brook in Pepperell, MA caused
enough of a channel change to necessitate anew rating curve 66
Figure 30. Examples of good cross-sections for making discharge measurements 68
This document is a draft for review purposes only and does not constitute Agency policy.
vii DRAFT—DO NOT CITE OR QUOTE
-------
Figure 31. Layout of a channel cross-section for obtaining discharge data, using the velocity-area
procedure 69
Figure 32. Example of regular (left) and log-log scale (right) rating curves created using Aquatic
Informatics'AQUARIUS software 71
This document is a draft for review purposes only and does not constitute Agency policy.
viii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS
GCIA Global Change Impacts and Adaptation
GPS Global Positioning System
NIST National Institute of Standards & Technology
PVC Polyvinyl Chloride
PZF Point of zero flow
RIFLS River Instream Flow Stewards Program
RMNs Regional Monitoring Networks
QAPPs Quality Assurance Project Plans
QAQC Quality Assurance Quality Control
SOP Standard Operating Procedures
US EPA United States Environmental Protection Agency
USGS United States Geological Survey
This document is a draft for review purposes only and does not constitute Agency policy.
ix DRAFT—DO NOT CITE OR QUOTE
-------
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, and the Global
Change Assessment Staff in the Air, Climate, and Energy Program at the U.S. Environmental
Protection Agency. It is meant to facilitate 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.
This document is a draft for review purposes only and does not constitute Agency policy.
x DRAFT—DO NOT CITE OR QUOTE
-------
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 is
responsible for publishing this report. This document was prepared by Tetra Tech, Inc. under
Contract No. EP-C-12-060; EPA Work Assignment No. 0-01. Dr. Britta Bierwagen served as the
Technical Project Officer. Dr. Bierwagen provided overall direction and technical assistance, and
she contributed as an author.
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, Office of Research and Development Washington DC
Britta G. Bierwagen
REVIEWERS
U.S. EPA Reviewers
Richard Mitchell, Joe Flotemersch, Lil Herger and Margaret Passmore
ACKNOWLEDGMENTS
The authors would like to thank the many state and federal partners who reviewed early versions
of this draft for clarity and usefulness. Their comments and input have substantially improved
this document.
This document is a draft for review purposes only and does not constitute Agency policy.
xi DRAFT—DO NOT CITE OR QUOTE
-------
EXECUTIVE SUMMARY
The lack of continuous temperature and flow data for minimally-disturbed, free-flowing
freshwater wadeable streams is an impediment to analyses of long-term trends in biological,
thermal, and hydrologic data. To address this, a number of state biomonitoring programs have
expressed an interest in incorporating annual monitoring at targeted, minimally disturbed sites
into their existing programs, and in coordinating these efforts at a regional level in order to pool
resources and increase efficiency. In response, the United States Environmental Protection
Agency (US EPA) has been collaborating with its regional offices and states to develop Regional
Monitoring Networks (RMNs) to detect changes in stream biota and thermal and hydrologic
regimes. This guidance document is meant to facilitate more uniform and effective deployment
of continuous temperature sensors and water level sensors at ungaged sites in wadeable streams.
It addresses questions related to equipment needs, sensor configuration, sensor placement,
installation techniques, data retrieval, and data processing. The data collected through these
efforts will further our understanding of how biological, thermal, and hydrologic conditions vary
spatially and temporally and inter-relate to one another, and will help inform state and federal
agencies and tribes on how to attribute altered environmental conditions to climate change versus
other stressors.
The temperature section of this document describes protocols for measuring continuous water
and air temperature. At least one water temperature sensor should be deployed at RMN sites.
Collection of air temperature data is encouraged as well, since air temperature is an important
component of many modeling efforts and can be used to determine if water temperature sensors
are dewatered during their deployment. Two installation techniques for water temperature
sensors are described: the underwater epoxy method and a method in which cabling is used to
attach the sensor to rebar or stable instream structures. After the installation is complete, it is
critical that the sensor be accurately geo-referenced and documented in a way that allows field
personnel to re-locate it during subsequent visits.
The streamflow section of this document discusses how to install and maintain pressure
transducers and staff gages to measure continuous stage (depth). Protocols and methods are
based on those used by Massachusetts Division of Ecological Restoration River Instream Flow
Stewards program (RIFLS), the United States Geological Survey (USGS), and the Washington
Department of Ecology. Pressure transducers record absolute pressure, which software then
converts to water level. Data are corrected for barometric pressure. Methodologies for this
correction depend on the type of pressure transducer used. Staff gages should be installed at
RMN sites because they provide a means to check the accuracy of the transducer data.
Procedures for selecting sites, installing pressure transducers and staff gages, surveying their
locations and maintaining equipment are covered in the document. In addition, we briefly
describe two approaches for converting stage measurements into streamflow: 1) developing a
flow rating curve; and 2) modeling flow based on a survey of a cross-sectional profile of the
stream. 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, it is most useful to convert stage measurements into streamflow.
This document is a draft for review purposes only and does not constitute Agency policy.
xii DRAFT—DO NOT CITE OR QUOTE
-------
After equipment is installed at a site, the site should be revisited within the first month to confirm
that the temperature sensor and pressure transducer installations are holding properly. After these
initial deployment checks, sites should be visited annually if possible (e.g. in conjunction with
the biological sampling events) to check the condition of the sensors, gather data for mid-
deployment accuracy checks, and offload data. More frequent visits are encouraged, particularly
to check for movement of the staff gage and transducer after high flow events and periods of
extended ice cover.
After data are offloaded, quality assurance and control procedures should be performed to verify
the quality of the data and to check for potential errors. In this document we describe a series of
automated and visual checks that should be performed as a part of quality control procedures for
the data. Because large amounts of data will accumulate quickly, a central database should be
developed and maintained from the initial stages of monitoring, and all calibration and field
forms should be organized, easily accessible, and archived in a way that allows for safe, long-
term storage.
This document is a draft for review purposes only and does not constitute Agency policy.
xiii DRAFT—DO NOT CITE OR QUOTE
-------
1. PURPOSE AND SCOPE
Few monitoring agencies have collected adequate time-series data to support analyses of long-
term trends in biological, thermal, and hydrological data in minimally disturbed, free-flowing
freshwater streams (Mazor et al. 2009; Jackson and Fureder 2006; Kennen et al. 2011). Such data
are necessary to further our understanding of how conditions vary across sites and over time.
Moreover, these data can also inform state and federal agencies on how to assess the relative
importance of climate change as a stressor for environmental health. Climate-related impacts are
occurring now and are expected to increase. They include rising temperatures, changes in the
amount, intensity, frequency, and type of precipitation, alterations in stream flows, and greater
risk of droughts and floods (Karl et al. 2009).
To address these needs, a number of state biomonitoring programs have expressed an interest in
incorporating annual monitoring at targeted, minimally disturbed sites into their existing
programs. Such efforts would ideally be coordinated on a regional level in order to pool
resources and increase efficiency. In response, the United States Environmental Protection
Agency (US EPA) has been collaborating with states from various regions to develop Regional
Monitoring Networks (RMNs) to detect long-term changes in stream biota and thermal and
hydrologic regimes. Analyses conducted in collaboration with northeastern states are being used
to inform design decisions on sample size, classification, site selection, spatial distribution of
sites, indicators, and trend detection time (Bierwagen et al. 2013, in review). Based in part on
these analyses, regional working groups are trying to initiate continuous temperature and flow
sampling and annual biological sampling at targeted sites in each state. Objectives of these
monitoring efforts are to detect climate change-related impacts over the long term, further our
understanding of how biological, thermal, and hydrologic conditions vary spatially and
temporally and inter-relate to one another, and to inform state and federal agencies on how to
attribute altered environmental conditions to climate change versus other stressors.
Year-round temperature monitoring is fundamental to understanding aquatic ecology. Moreover,
there is evidence of significant departures in temperature from historical conditions in response
to a warming climate (Isaak et al. 2012, Kaushal et al. 2010). Although considerable amounts of
stream temperature data are now routinely collected using inexpensive temperature sensors,
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 since it captures a critical time
period for most aquatic species' survival. However, summer-only data provides 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.
In addition to temperature, flow regime (magnitude, frequency, duration, timing, and rate of
change) also has a strong influence on stream ecology (Poff et al. 1997). The United States
Geological Survey (USGS) has been measuring flow in streams since 1889, and currently
maintains over 7,000 continuous gages. This large network provides critical, long-term
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). In
This document is a draft for review purposes only and does not constitute Agency policy.
1 DRAFT—DO NOT CITE OR QUOTE
-------
addition, many of the stream gages deliver data instantaneously through the internet, providing
real-time information for a variety of consumers.
At this time, sampling efforts in the RMNs are primarily focused on medium to high-gradient,
minimally disturbed headwater and mid-order streams (generally with drainage areas less than
100 square kilometers). Monitoring flow in headwater and mid-order streams is important
because streams of this size play a critical role in connecting upland and riparian systems with
river systems (Vannote et al. 1980). Not only do they provide sources of water, sediments and
biota (Sidle et al. 2000), they are also critical sites for processing organic matter and nutrient
cycling (Bilby and Likens, 1980, Wipfli et al. 2007, Clarke et al. 2008). Moreover, small upland
streams are likely to experience substantial impacts from climate change (Durance and Ormerod
2007). Headwaters in particular may be vulnerable because their flow dynamics are often closely
tied to precipitation patterns. Headwaters may also serve as refugia from the extremes in
temperature and flow that are projected to occur (Meyer et al. 2007).
Efforts are being made to co-locate as many RMN sites as possible with active USGS gages
since USGS gage data represent the highest quality flow data available. However, many USGS
stream gages are located in large rivers that have multiple human uses. Thus, only a limited
number are minimally disturbed streams, and in most locations high-quality flow data must be
obtained via alternate methods.
The primary purpose of this document is to provide guidance on how to collect accurate,
continuous temperature and flow data at ungaged sites in wadeable streams. It addresses
questions related to equipment needs, sensor configuration, sensor placement, installation
techniques, data retrieval, and data processing. It describes techniques for attaching sensors to
natural objects in streambeds, and also covers installations in which sensors are attached to man-
made structures like bridges. Although it was written for the RMNs, much of the information and
detail (e.g., data forms) can be used by entities that are developing their own Quality Assurance
Project Plans (QAPPs) and standard operating procedures documents (SOPs) for measuring
stream temperature and flow.
This document differs from existing SOPs in that both temperature and flow information are
compiled into one place, and deployment techniques specifically address challenges posed by
year-round deployment. Efforts were made to include only methods that are formally published
and/or have been shown to be effective in field tests. Because limited resources are available for
implementation of the RMNs, whenever possible, protocols describe simple methods and
reference inexpensive tools needed to collect data. If resources permit, there are many
possibilities for collection of additional data at RMN sites. For example, if multiple temperature
sensors are available for deployment at a site, they could be placed in different riffles to provide
a measure of within-reach temperature variability. Or, for quality assurance checks, duplicate
sensors could be placed at 10% of the sites, and/or a new sensor that will be replacing an older
sensor could be deployed early so that there is overlap between deployment periods. While not
covered in this document, the collection of these additional types of data would be very
informative.
This document is a draft for review purposes only and does not constitute Agency policy.
2 DRAFT—DO NOT CITE OR QUOTE
-------
We attempted to make this document as comprehensive as possible, but acknowledge that there
will be situations in which an alternative methodology, procedure, or process is warranted.
Future updates and improvements will be made to this document 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 a
minimum, at least one water temperature sensor should be deployed at each monitoring site. If
resources permit, additional water temperature sensors may be useful at sites that have high
reach-scale variability in temperature. Collection of air temperature data at individual sites is also
useful, since air temperature is an important component of many modeling efforts (e.g., Hill et al.
2013) and can be used to determine if water temperature sensors are dewatered during their
deployment (Bilhimer and Stohr 2009, Sowder and Steele 2012).
The protocols described in this section are based on a review of existing temperature protocols
from the Washington State Department of Ecology (Ward 2011, Bilhimer and Stohr 2009, Ward
2003), the US Forest Service (Dunham 2005, Isaak et al. 2013, Isaak and Koran 2011, Sowder
and Steel 2012, Holden et al. in press), Alaska (Mauger 2008), and Maryland (MDNR, no date).
2.1. Equipment
In this section we discuss equipment that is needed for collecting air and water temperature data,
as well as considerations that go into selecting the equipment.
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.
This document is a draft for review purposes only and does not constitute Agency policy.
3 DRAFT—DO NOT CITE OR QUOTE
-------
2.1.2. Considerations when choosing temperature sensors
Factors such as durability, accuracy, resolution, measurement range, memory, and battery life
should be considered when selecting temperature sensors. For the RMNs, water temperature
sensors should be durable (able to withstand years of use in challenging conditions), waterproof
(i.e., IP code level 8), and have a minimum accuracy of ± 0.5°C (Table 1). In addition, proper
calibration and accuracy checks (Sections 2.2.1 and 2.6.1) should be performed to ensure that
sensors meet the specifications quoted by the manufacturers. If these steps are taken, the
replacement of sensors every few years should not result in systematic bias, and any errors
associated with equipment differences will be extremely small compared to the temperature
variability within and among sites.
Air temperature sensors should have a minimum accuracy of ± 0.5°C as well. Water temperature
sensors with accuracies of ± 0.2°C are currently available and should be used if possible at RMN
sites (Table 2). Both water and air temperature sensors should have a measurement range that
captures the full range of expected temperatures, a memory that is sufficient to record
measurements at 30-minute intervals during the deployment period, and adequate battery life.
The lifespan of the sensor is another consideration. Some sensors are made to last 5 years or
longer before their batteries run out and/or their cases start to leak. If sensors with non-
replaceable batteries are used, be sure to
document the sensor's use so you know when
to take them out of circulation and budget/plan
for their replacement. Regardless of what
model is used, steps should be taken to
minimize the number of different sensors
deployed at the same site over time, so that
Definitions of accuracy, precision and bias
Accuracy refers to how close the temperature
measurement is to its "true" value. Precision is
the variance or "tightness" of the temperature
measurements. Bias refers to whether there is a
systematic offset between the measured value
and the "true" value.
inter-instrument error can be minimized. In
addition, proper calibration and accuracy
checks (Sections 2.2.1 and 2.6.1) should be performed to ensure that sensors meet the
specifications quoted by the manufacturers. If these steps are taken, the replacement of sensors
every few years should not result in systematic bias, and 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
Resolution3
Measurement range - able to capture
the full range of expected
temperatures
Water Sensor
yes1
yes
±0.5°C
<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 may be necessary
(a typically available range)
This document is a draft for review purposes only and does not constitute Agency policy.
4 DRAFT—DO NOT CITE OR QUOTE
-------
Memory
Sufficient to record measurements at 30-minute intervals
during deployment period
Battery life
Sufficient to remain active during deployment period
1 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; expected to commonly experience
Resolution is the smallest detectable increment that the sensor can measure; it needs to be less than the accuracy.
The same sensors that are used for measuring
water temperature can be used to measure air
temperature as long as 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 may be necessary).
Less expensive, non-waterproof sensors can
also be used to measure air temperature, as
long as the sensors are protected from the
elements (see Table 2). Radiation shields,
which are discussed in Section 2.1.4, can
serve this purpose.
If the pressure transducer that I am
purchasing has the capacity to 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, then it is not
necessary to deploy a separate temperature
sensor. However, even with an accurate
temperature sensor in a transducer, a separate
temperature sensor can be deployed as a back-
up.
Examples of some currently commercially
available temperature sensors can be found in Table 2, and pictures of these sensors are shown in
Figure 1.
A)
B)
C)
Mfunct
^^
F)
777/5 document is a draft for review purposes only and does not constitute Agency policy.
5 DRAFT—DO NOT CITE OR QUOTE
-------
Figure 1. Numerous temperature sensors are commercially available. Examples include: A) Onset Hobo© Water
Temp Pro v2; B) Onset Tidbit© v2; C) Gemini Tiny tag Aquatic 2; D) Thermoworks LogTag; E) MadgeTech Temp
101 A; 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, as long as the proper calibration and accuracy checks are performed to ensure that sensors
meet the specifications quoted by the manufacturers (see Sections 3.2.1 and 3.7.1). Any errors
associated with equipment differences are extremely small compared to the temperature variability
within and among sites.
This document is a draft for review purposes only and does not constitute Agency policy.
6 DRAFT—DO NOT CITE OR QUOTE
-------
Table 2. Examples of commercially available temperature sensors
Manufacturer (web site)
Onset (onsetcomp.com)
Gemini
(geminidataloggers.com)
Maxim Integrated
Products
(maximintegrated . com)
Thermoworks
(the rmo works . com)
MadgeTech
(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)
Thermochron
ibutton
(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) &
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 may be found on the manufacturer specification sheets
2As of January 2013 and subject to change; reduced prices may be available for bulk orders
777/5 document is a draft for review purposes only and does not constitute Agency policy.
1 DRAFT—DO NOT CITE OR QUOTE
-------
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). In order to view
the data, the appropriate software needs to be installed on the computer.
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. For the RMNs, instream data offloads are generally not possible
because sensors should be 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.
Should I purchase a waterproof data offload device?
For added expense, one can purchase a waterproof data offload device (sometimes called a shuttle). It
can be used to download data while the sensor remains in the water. However, at RMN sites, the
sensors will be housed in protective shields, so the sensors will need to be removed from the cases
(and the stream) prior to data download. In these situations, the main benefits of the waterproof
shuttles versus the non-waterproof devices (often called base stations) 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.
Compatibility is important to consider when purchasing equipment. Ensure that the data offload
device and software are compatible with the model of the temperature sensor. If purchasing
multiple sensors, it is often most cost-effective to buy the same model since reduced prices may
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 lune 2013).
This document is a draft for review purposes only and does not constitute Agency policy.
8 DRAFT—DO NOT CITE OR QUOTE
-------
Tidbit-
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 that are 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.
2.1.4. Considerations when choosing radiation shields
Temperature sensors must be outfitted with radiation shields because sunlight striking the
sensors biases temperature readings
(Figure 3) (Isaak and Horan 2011,
Dunham 2005). These shields can also
serve as protective housings and can
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.
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 may also increase
vandalism rates, especially in heavily trafficked
areas. If vandalism is a concern, use radiation
shields that are neutral colors or paint them to reduce
their visibility.
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 2
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 that epoxy can be wrapped around as the epoxy will not hold the PVC to
the rock without this. More details on how to construct PVC canister shields, including specific
part numbers, can be found in Isaak et al. (2013). An alternate method for constructing PVC
canisters is described in Appendix A (Mauger 2008).
This document is a draft for review purposes only and does not constitute Agency policy.
9 DRAFT—DO NOT CITE OR QUOTE
-------
50
100
150
200
250
300
350
400
30 minute interval
Figures. 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 during times
when sunlight struck the sensor (reproduced from Isaak and Horan 2011).
Figure 4. Unassembled components of an inexpensive PVC canister radiation shield for water temperature
sensors (described in more detail in lsaaketal.2013).
PVC canisters should not be used as solar shields for air temperature sensors because they lack
sufficient air flow. 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). It is a series of plates that reflect incoming radiation
and passively radiate accumulated heat via conduction (see Figure 5 A). In areas with persistent
This document is a draft for review purposes only and does not constitute Agency policy.
10 DRAFT—DO NOT CITE OR QUOTE
-------
winter and spring snowpack, the radiation shield should block incident radiation not just from
above, but also from below, since radiation reflecting off the snow can directly strike sensors and
bias temperature readings (Holden et al. in review).
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 (Holden 2012, unpublished; Appendix B; Figure 5B) costs approximately
$2.50-3.00 per shield. A comparison of the performance of this custom shield to commercially-
available shields is described in 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.voutube.com/watch?v=LkVmJRsw5vs.
A)
B)
Figure 5. Examples of radiation shields for air temperature sensors include A) the Gill-style Onset RS1 solar
radiation shield (www.onsetcomp.com) and B) custom design by Zachary Holden (Appendix B).
777/5 document is a draft for review purposes only and does not constitute Agency policy.
11 DRAFT—DO NOT CITE OR QUOTE
-------
2.2. Pre-Deployment
2.2.1. Calibration/accuracy check
Before field deployment, a calibration procedure should be performed in-house to check the
accuracy of the temperature sensors and to ensure that the sensors are launching and
downloading data properly.
Table 3 presents a step-by-step procedure for doing a multiple-point temperature calibration.
First, sensors should be set up to record at the same sampling intervals that will be used in the
field (in this case, every 30 minutes). Next, the sensors should be exposed to alternating warm
and cold cycles that approximate the temperatures and duration of diurnal fluxes that the sensors
will be exposed to in the field. At various points during this process, temperature measurements
should be taken with a National Institute of Standards & Technology (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, per 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.
All calibration data should be recorded on a calibration datasheet, similar to the one shown in
Appendix C. These records and the digital data from the calibration test should be retained in a
safe place for future reference since they can be used to assess drift rates if sensors are deployed
for multiple years. Equipment needs for this procedure are listed in Table 4.
This document is a draft for review purposes only and does not constitute Agency policy.
12 DRAFT—DO NOT CITE OR QUOTE
-------
Table 3. Pre-deployment calibration procedure
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 get 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 thermometer; record the value
on a calibration datasheet (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 calibration 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 (this should
be < 0.5 C for the water temperature sensors and ± 0.5 C for the air
temperature sensors). Sensors that have anomalous readings
should be set aside and returned to the manufacturer for
replacement.
Record all of the calibration information on a calibration data
13
sheet (like the one shown in Appendix C).
This document is a draft for review purposes only and does not constitute Agency policy.
13 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4. Equipment needs for the pre-deployment accuracy check
Task
Performing the
calibration
procedure
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 (i.e. red liquid) thermometers (optional)
• Containers to hold the sensors
• Water
• Refrigerator
• Clock or watch
• Calibration data sheet
• Computer that has the
temperature sensor
(Appendix C)
appropriate software for reading the
2.2.2. Sensor configuration and launch
When configuring the sensors, it is helpful to set them to local standard time, as this will simplify
data processing (Ward 2011). Sensors can be programmed to start before or after the planned
deployment time, or can be launched on-site. Whatever the timing, it is critical that field
personnel record the exact time the sensor is correctly positioned in the stream channel so that
observations recorded before and after that time can later be removed during data processing
(Section 2.6.2).
For the RMNs, the sensors should be configured to record point temperature measurements in
degrees Celsius at intervals of 30 minutes on the hour and half hour (e.g., 5:00, 5:30, 6:00, 6:30).
If monitoring agencies are not able to visit a site annually and the sensor has insufficient memory
capacity to record at 30-minute intervals for the deployment period, sensors should be
programmed to record point measurements every 60 or 90 minutes. Figure 6 suggests that 60
minute intervals are likely sufficient, though 30 minute intervals may more closely match other
continuous data collection efforts such as water levels.
The configuration and launch information for each sensor should be recorded on a datasheet,
similar to the one shown in Appendix D. It should include serial number, programmed
deployment time, and recording interval. For the RMNs, if the sensor has "sensor high, low and
multiple sampling" features and "wrap-around-when-full, overwrite oldest data" functions, these
should be turned off.
This document is a draft for review purposes only and does not constitute Agency policy.
14 DRAFT—DO NOT CITE OR QUOTE
-------
Considerations when choosing a sampling interval
Choosing a sampling interval is a balancing act. If long intervals are used, you may miss the
maximum and minimum daily temperature since these may only be observed for a short time
within a day. This would result in underestimating the warmest temperatures and overestimating
the coldest temperatures. If too short an interval is used, it could use up too much memory, and
requires 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, there is not much bias. Sites with larger diel fluctuations have a greater probability of
missing the true maximum than those with smaller diel fluctuations.
For the RMNs, the 30-minute interval should be re-evaluated after the first year to determine
whether a longer interval (i.e. 60 or 90 minutes) can adequately capture the thermal regimes at
sites.
0.20
4 hr interval
3 hr interval
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).
This document is a draft for review purposes only and does not constitute Agency policy.
15 DRAFT—DO NOT CITE OR QUOTE
-------
2.3. Deployment of Water Temperature Sensors
At a minimum, one water temperature sensor should be deployed at each site. When deploying
sensors, safety should always be a consideration, and sensors should only be placed in streams
that can be safely accessed and waded by crew members. If possible, avoid reaches with very
high gradients (>7%) since sensor retention rates are inversely related with slope (Isaak et al.
2013). Prior to installing the sensors, be sure to obtain permissions from all relevant parties (e.g.,
property owners if accessing sites will involve crossing private property, the local Department of
Public Works if the water temperature sensor will be attached to a bridge abutment). Efforts
should be made 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 man-
made structures like bridges. If a situation occurs in which a sensor is attached to a bridge or
other man-made structure, the biological sample should be collected at least 200 meters from the
area of human disturbance. As long as the sensor is collecting data that are representative of the
characteristics of the reach from which the biological data are being collected, the sensor does
not have to be sited in the exact location that biota are being collected from.
2.3.1. Guidelines for placement within the stream
There is not a simple straightforward formula for selecting a location since each site is different,
but in general, areas of well-mixed water moving through runs and pools are preferable over
riffles. In addition, sensors should be deployed 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 was
collected (e.g., not below additional warm
or cold water sources). Note that for
general monitoring purposes the sensor
does not have to be placed in the exact
location that biota are being collected
from.
Well-mixed horizontally and vertically
Of sufficient depth to keep the sensor
submerged year-round
Stable and easy to re-locate
Protected from physical impacts
associated with high flow events (i.e. 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, anything that is moving is going
to be well-mixed. To verify this, take
numerous instantaneous temperature
measurements in the vicinity of the
deployment location. If the stream can be
easily waded, then 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, since variability in
these measures could indicate sources of
thermal variation (Dunham 2005). If there
is a high degree of variability in these
measures, consider moving to a different
deployment location.
This document is a draft for review purposes only and does not constitute Agency policy.
16 DRAFT—DO NOT CITE OR QUOTE
-------
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 may 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 hotsprings
o groundwater seeps
When possible, sensors should be deployed 6 inches (<0.5 ft) above the stream bottom (per
Schuett-Hames et al. 1999). There may be situations (e.g., small, shallow streams) where you
have no choice but to place a sensor near the stream bottom to ensure that it remains submerged
during low flows. Note on the field form when situations like this arise, because influences from
groundwater, subsurface flow, and the substrate can cause subsurface temperatures to deviate
from temperatures in the well-mixed portion of the stream (Zimmerman and Finn 2012). Sensors
should never be intentionally buried.
Also, when possible, sensors should be installed on the downstream side of the structure to
which the sensor is being attached (e.g., a large rock or log), since high water velocities and
associated substrate movement and transport of debris 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, since this will allow 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.
Site-specific conditions will dictate which installation technique is most appropriate. Below, we
briefly describe two methods - the underwater epoxy method (Isaak and Horan 2011, Isaak et al.
2013) and a method in which cabling is used to attach the sensor to rebar or stable instream
structures such as large rocks or boulders, woody debris, or roots (Ward 2011, Mauger 2008).
Equipment needs for both installation techniques are listed in Appendix E.
If site conditions permit, the preferred method for year-round deployments is the underwater
epoxy method, since it requires minimal effort and materials and provides durable installations
that withstand floods and associated bed-load movement. Moreover, once a sensor site is
successfully established, it is easily maintained in years thereafter simply by replacing sensors in
This document is a draft for review purposes only and does not constitute Agency policy.
17 DRAFT—DO NOT CITE OR QUOTE
-------
PVC housings that remain in place (based on field trials, 80% - 90% of sensors installed
correctly remained in place after 1- and 2-year intervals (Isaak et al. 2013).
2.3.2.1. Underwater epoxy
The underwater epoxy method can be used in multiple environments as long as there is a suitable
anchor point, such as a large rock or cement structure (e.g., bridge support). Table 5 describes
the step-by-step procedure for doing 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 surely dislodge it, not to mention the
safety risk associated with lifting large boulders). Ideally, on the downstream side of the
attachment point, there should be pockets of relatively calm water with smaller substrate sizes (if
large rocks and cobbles are on the downstream side, it is likely that similarly large substrates will
move there again during the next flood, and these could dislodge or break the sensor). Cement
bridge pilings at road crossing are also good attachment points for this protocol. Photos of good
attachment points are shown in Figure 7.
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).
Selection of 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 and worked well in field conditions. The epoxy works well during installations
in water temperatures ranging from 2 to 20°C, but becomes less cohesive as temperatures warm.
If applying this type of epoxy in water temperatures significantly exceeding 20°C, run tests to
ensure the epoxy sets within 24 hours of installation. Alternatively, installations with FX-764
could be done during times of the day when water temperatures are relatively cool to allow the
epoxy to set. If a sensor needs to be reattached to a structure that has old Fox bonding agent on it,
it is best to use a different attachment point if possible.
Crews that are inexperienced with this technique should first do some practice runs in the
laboratory, and then do initial field installations at a few easily accessible locations that can be
checked after a few days.
To monument a site, the epoxy can also be used to attach a metal forestry tag near the sensor
location, which can make the sensor easier to relocate during subsequent site visits (Figure 7).
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 Horan 2013).
Equipment needs for doing underwater epoxy installations are summarized in Table 6 and Figure
9. Additional details on doing underwater epoxy installations are available in Isaak et al. (2013)
and in the YouTube training video (accessed 15 June 2013):
This document is a draft for review purposes only and does not constitute Agency policy.
18 DRAFT—DO NOT CITE OR QUOTE
-------
http://www.voutube.com/watch?v=vaYaycwftnXs&feature=v outu.be
Table 5. Quick guide for doing underwater expoxy installations (Isaak et al. 2013)
Task
Select
attachment
point
Installation
Monument
Procedure
1 . The attachment point should have the following features:
• Is protruding a foot or more above the water surface at
low 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
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 (at least 2-3 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.
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.
This document is a draft for review purposes only and does not constitute Agency policy.
19 DRAFT—DO NOT CITE OR QUOTE
-------
Figure 7. Examples of large rocks (a and b) and cement bridge pilings (c and 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 taken from Isaak et al. 2013).
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.
This document is a draft for review purposes only and does not constitute Agency policy.
20 DRAFT—DO NOT CITE OR QUOTE
-------
Table 6. Equipment list for doing underwater epoxy installations (Isaak et al. 2013). Monumenting is
discussed in Section 2.3.3
Task
Supplies list
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
Plumber's tape
Wire brush
Zip ties, 4"
Metal mirror
Monument
Metal forestry tags
Spray paint
GPS
Camera
Data sheet (similar to Appendix D)
This document is a draft for review purposes only and does not constitute Agency policy.
21 DRAFT—DO NOT CITE OR QUOTE
-------
Fox Indus 410-243-8866
FX-764 Splash Zone Epox
A Component
Mix equal parts lAtolB
Caution: Contains Epoxy R
v/ww foxmd com
Fox Indus 410-24S-88S(
FX-764 Splash Zone Epoi
B Componenl
Mix equal parts 1A to IE
Caution: Contains Epoxy Rfi
www foxmd corr
d
Figure 9. Equipment needed to permanently install temperature sensors in streams using underwater epoxy
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, and (g) plastic viewing box, (h) wire brush, and (i) metal forestry tree tag
(lsaakandHoran2013).
2.3.2.2. Cabling the sensor to rebar or stable instream structures
If the underwater epoxy method cannot be used at a site, sensors may be cabled to rebar or stable
instream structures like large rocks or boulders, roots, or woody debris. These techniques are
commonly used and have been shown to be effective for seasonal summer deployments (Email,
Bill Ward to Jen Stamp, January 4, 2013). A downside is that if these techniques are used for
year-round deployments in streams that experience annual high flow events, sensors may
sometimes be buried or dislodged by moving substrates.
Ideally the sensors can be attached to the downstream side of the instream structure, as this will
shield the sensor from moving rocks or debris during floods. Cable ties and/or wire are used to
attach the sensors to the structures. If you think the structure might move during high flow
events, consider cabling or chaining the structure to something on the nearest bank (or to another
stable instream structure).
This document is a draft for review purposes only and does not constitute Agency policy.
22 DRAFT—DO NOT CITE OR QUOTE
-------
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 (i.e. no near-surface bedrock or consolidated sediments), the rebar
method is commonly used. A 2-3 foot length of rebar 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 cable ties or wire. Photos of a rebar installation are shown in Figure 10.
More detailed descriptions of these methods can be found in Ward 2011, Mauger 2008, and
Appendix F. An alternate technique that utilizes a low profile concrete base is also described in
Appendix F.
PVC housing
holding sensor
cable
Figure 10. Photos of a rebar installation (from Mauger 2008).
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 2005). Thus, it is critical that each sensor be
accurately georeferenced and that sensor placement is documented in a way that allows field
personnel to re-locate the sensor during subsequent visits.
This document is a draft for review purposes only and does not constitute Agency policy.
23 DRAFT—DO NOT CITE OR QUOTE
-------
Table 7 contains a list of guidelines for documenting sensor locations. First, GPS coordinates
(latitude and longitude) should be recorded for the exact site at which each sensor is deployed, as
well as the datum of the GPS. If unable to obtain GPS coordinates in the field, note the sensor
location on an accurate map and determine the coordinates later. While GPS coordinates are
useful for getting close to the site, they are insufficient by themselves, 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.
Photos should be taken from different perspectives (i.e., upstream and downstream), and should
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 may be appropriate. Detailed
hand-drawn maps like the one shown in Figure 11 (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 12).
All documentation information must be recorded on a field form, like the form shown in
Appendix D. Field forms should include information on station number, waterbody name, date,
time, crew members, driving directions, serial number of sensor/s, time of deployment, sensor
installation technique, image numbers/file names for the photographs, and detailed descriptions
of sensor placement. It is very important to accurately record the time and date of deployment, as
this information will be used in the error screening procedure described in Section 2.7.2.
If the stream can be easily waded, an instantaneous stream temperature measurement 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. If time permits, the following additional
measurements should be taken: the total stream depth at the sensor, the distance from the stream
bottom up to the sensor, distance from water surface to the sensor and wetted width along a
transect that intersects the sensor. In addition, it is helpful to do a cross sectional survey of the
stream temperature, as described in the text box in Section 2.3.1, since these results will help
verify that the stream temperature sensor is measuring representative temperatures and will also
expose any cross-sectional temperature differences. Equipment needs for taking these
measurements and for documenting sensors are summarized in Table 8.
This document is a draft for review purposes only and does not constitute Agency policy.
24 DRAFT—DO NOT CITE OR QUOTE
-------
Table 7. 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 whatever else may be appropriate and/or annotate
photographs when viewing them later on a computer.
Complete field form.
Temperature
and depth
measurements
• Measure the following and record on field form:
• Total stream depth at the sensor.
• Distance from the stream bottom up to the sensor.
• Distance from water surface to the sensor.
• Instantaneous stream temperature at the location of the
sensor, taken 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.
• Wetted width along a transect that intersects the sensor
• If the stream can be easily waded, do a cross sectional
survey of the stream temperature, as described in Section
2.3.1.
This document is a draft for review purposes only and does not constitute Agency policy.
25 DRAFT—DO NOT CITE OR QUOTE
-------
Continuous Temperature Survey Form
Station* QQCltC Station Name: (2 £*>**- M£ J4*il><, &u£6- Samplers: UJfaLb / l^
Interval Frequency 00:30
Water Temperature Logger
i.a# U57 37 3
Water Depth / . S ft Deployment Depth
Height
-------
Table 8. Equipment list for documenting sites.
Task
Georeferencing
& monumenting
Measuring
temperature and
depth
Supply list
• GPS
• Camera
• Map and/or gazetteer
• Metal forestry tags
• Field form (similar to Appendix F
, Figure 3. 3 -4)
• NIST-calibrated field thermometer and/or multi-probe meter
• Meter ruler or calibrated rod/pole (i.e. surveyor's rod)
• Measuring tape
2.3.4. Common problems
As summarized in Dunham 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 9. 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.
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 won't 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, it will not be impacted by chunks of ice
moving downstream.
This document is a draft for review purposes only and does not constitute Agency policy.
27 DRAFT—DO NOT CITE OR QUOTE
-------
Table 9. Tips for minimizing the chance of sensors being lost or damaged
Problem
Tips for minimizing chance of damage or loss
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 and/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 it may also make the data
sensor more difficult to relocate)
Secure the sensor in a 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
This document is a draft for review purposes only and does not constitute Agency policy.
28 DRAFT—DO NOT CITE OR QUOTE
-------
2.4. Deployment of Air Temperature Sensors
If possible, air temperature should be collected at RMN sites. The air temperature sensor should
be located in the riparian zone, as close as possible to the water temperature sensor. Ideally, it
should be installed in a place that is out of direct sunlight and has low potential for vandalism.
As discussed in Section 2.1.4, temperature sensors must be outfitted with radiation shields so that
sunlight does not strike the sensor and bias the temperature readings (Isaak and Horan 2011,
Dunham 2005). Vegetation alone should not be used as the primary radiation shield.
Trees (ideally > 12 inch diameter) are the best attachment points for air temperature sensors
because they provide stability and some degree of shade. If a suitable tree is present, attach the
radiation shield and sensor to the north side of the tree using the simple four-step process
described in Appendix B (Holden 2012, unpublished). 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 (Holden 2010, unpublished), 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 sensor to a 10-foot piece of 1A inch diameter PVC pipe, which can be
purchased at a local building supply store. Using a drill, create a l/8th 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 3-4 foot
piece of metal rebar into the ground to a depth of approximately 1-foot, and slide the PVC pipe
onto the rebar. Please note that this method has only been used experimentally and has not been
extensively tested.
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, daily air temperature observations from
the nearest active weather station should be compiled. 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).
This document is a draft for review purposes only and does not constitute Agency policy.
29 DRAFT—DO NOT CITE OR QUOTE
-------
There is value in obtaining data from the nearest weather station even if an air temperature
sensor is installed at a site. For one, it could be used for quality assurance purposes (e.g., if a
comparison of the two datasets reveals differences in patterns, it may be an indication that the
on-site sensor is malfunctioning). Or, if the comparison shows there to be little difference
between data sets, the weather station data could potentially 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. In the future, instead of weather station data, it may be
possible to use modeled data. Higher spatial resolution, gridded air temperature models are
currently being developed that account for terrain influences on temperature. These data will
likely prove useful for understanding spatio-temporal variation in air temperature and its
influence on streams.
2.5. Maintenance/Mid-Deployment Checks and Data Offload
When a site is first established, it should be revisited 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, sites should be visited annually to check the condition of
the sensors, gather data for mid-deployment accuracy checks (Section 2.7.1), and offload data.
This can be done in conjunction with the annual biological sampling events. If annual site visits
are not possible, visit the site as frequently as your schedule permits, ideally during low flow
conditions.
This document is a draft for review purposes only and does not constitute Agency policy.
30 DRAFT—DO NOT CITE OR QUOTE
-------
Table 10 and the field form in Appendix G contain checklists for performing these
maintenance/mid-deployment checks and data downloads. For maintenance, look for signs of
physical damage, vandalism, or disturbance. Also ensure that the sensor is not buried by
sediment and remove anything that could bias the temperature readings (e.g., debris, aquatic
vegetation, algae). Photographs should also be taken to document any changes to the monitoring
location during the course of the study.
To gather data for the mid-deployment accuracy checks (Section 2.7.1), collect instantaneous
stream temperature measurements near the sensor with a NIST-calibrated field thermometer, as
close as possible to the time when the sensor is recording a measurement.
Procedures for data offload will vary depending on the model of the temperature sensor and the
data offload device. Typically, start by attaching the temperature sensor to a base station or
shuttle and then connect the data offload device to a computer that has the appropriate software
(Figure 2). 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 may affect its ability
to connect. Be careful not to scratch the sensor optic communication area when doing this. Once
the connection is established, follow the manufacturer's downloading procedures. The data
transfer should be done in a way that minimizes the disruptions/discontinuities in the long-term
temperature record.
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 there are unusual readings, 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 need to be
replaced each year, while others, like the Onset TidbiT v2, last five years or longer under
normal use); and
• back up the data (e.g., onto a flash drive) before clearing the sensor memory.
If it is impractical to bring a laptop into the field, 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, redeploy the same sensor at
the site, as this will minimize inter-instrument error. If the sensor needs to be removed from the
site (e.g., due to low battery life), before leaving the site, mark the sensor with a temporary tag
identifying the site, date, and time of retrieval, and replace it with another calibrated sensor. It is
very important to accurately record the time and date of retrieval, 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.
This document is a draft for review purposes only and does not constitute Agency policy.
31 DRAFT—DO NOT CITE OR QUOTE
-------
If conditions permit, the stream temperature and depth measurements described in Table 7
should be collected during each site visit, as should an instantaneous air temperature
measurement. The instantaneous stream and air temperature measurements should be taken with
a NIST-calibrated thermometer at the sensor locations, as close as possible to the time of the
expected sensor recording. These measurements and information on sensor condition should be
recorded on a field form like the one shown in Appendix G. Equipment needs for taking these
measurements and for conducting maintenance/mid-deployment checks and data downloads are
listed in Table 11.
Table 10. Checklists for performing maintenance/mid-deployment checks and data downloads
Task Procedure
Maintenance/
mid-
deployment
checks
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 isn't, move it to a location where it
is covered by water and will remain so during periods of base-flow.
Make sure the sensor is not buried in sediment. If it is, remove the sediment
and reinstall the sensor in a location 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 towards
cooler temperatures by hyporheic flows. In many cases, temperature
recordings from buried sensors should be destroyed due to the large amount
of bias incurred and because adjustments are difficult to apply with
accuracy.
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 abundantperiphyton or algal growth.
Take photos to document any changes to the monitoring location
(particularly those that may impact 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 field form (like the one shown in Appendix F).
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 may 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 the manufacturer's downloading
procedures.
Clear the sensor memory as necessary to ensure sufficient capacity for
continued deployment.
This document is a draft for review purposes only and does not constitute Agency policy.
32 DRAFT—DO NOT CITE OR QUOTE
-------
Sensor
retrieval
• If a sensor must be removed from a site, before leaving the site, mark the
sensor with a temporary tag identifying the site, date, and time of retrieval.
Temperature
and depth
measurements
• Measure the following and record on field form:
• Total stream depth at the sensor
• Distance from the stream bottom up to the sensor
• Distance from water surface to the sensor
• Wetted width along a transect that intersects the sensor
• If the stream can be easily waded, do a cross sectional survey of the
stream temperature, as described in Section 2.3.1.
Table 11. Equipment list for conducting maintenance/mid-deployment checks and data downloads
Task
Relocating
sensor
Documenting
on-site-
conditions
Data offloads
Measuring
temperature
and depth
Back-up
equipment (in
case a sensor
needs to be
replaced)
Supply list
• GPS
• Map and/or gazetteer
• Annotated photos and/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) & data back-up device (e.g. flash drive)
• NIST-calibrated field thermometer and/or multi-probe meter
• Meter ruler or calibrated rod/pole (i.e. surveyor's rod)
• Measuring tape
• Calibrated replacement sensors and other necessary deployment
equipment
2.6. Quality Assurance and Control
Quality assurance and control procedures must be performed after data are offloaded and/or
sensors are retrieved to verify the quality of the data and to check for potential errors. Records of
these procedures must be documented on a form like the one shown in Appendix H and stored
for long-term record keeping (ideally in a central database that someone maintains).
The series of data cleaning steps described below can improve data quality, reduce the time and
effort of data processing, and increase collaboration and comparison across projects, streams,
and regions (Sowder and Steele 2012).
This document is a draft for review purposes only and does not constitute Agency policy.
33 DRAFT—DO NOT CITE OR QUOTE
-------
2.6.1. Mid- and post-deployment accuracy checks
Some sensors will be deployed for multiple years (e.g., Onset TidbiT v2 sensors last 5 years or
longer under normal use). When possible, mid-deployment accuracy checks should be performed
on these sensors by comparing downloaded sensor values to the values of the instantaneous
stream temperature measurements that are collected with a NIST-calibrated field thermometer
during mid-deployment checks (see Section 2.5). If sensors are retrieved (i.e., brought back to
the office/laboratory), a post-deployment accuracy check should be performed, using the
calibration procedure described in Section 2.2.1.
When sensor values are compared to the values from the NIST-certified thermometers, they
should not exceed the accuracy quoted by the manufacturer (± 0.5 C). If a sensor fails this check,
repeat the procedure. If it fails a second time, flag the data with an appropriate data qualifier.
2.6.2. Error screening
Sensors may record erroneous readings during deployment for a variety of reasons. For example,
sensors may come out of the water as a result of low flow conditions, high flow events may bury
sensors in silt, sensors may malfunction, or humans may cause interference. Therefore, a
standard set of procedures should be performed to 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) are described in Table 12. The first step involves removing observations recorded
before and after the sensor is correctly positioned in the stream channel. 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, perform the series of automated and visual checks described in Table 12. The automated
checks will flag missing data and data points that fall outside expected thermal limits. The visual
checks are important because they pick up different types of errors (e.g., a dewatering event in
the spring when air temperatures do not exceed the thresholds of the automated checks), and can
be used to specify the time and duration of errors in the raw data files.
This document is a draft for review purposes only and does not constitute Agency policy.
34 DRAFT—DO NOT CITE OR QUOTE
-------
Table 12. Error screening procedure (based on Sowder and Steel 2012 and Dunham et al. 2005)
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.
Automated
checks
Missing data
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.
Visual checks
Plot individual data points to look for 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
Graphically compare data across sites
Graphically compare data across years; when data from one
year are dramatically different, there may be data errors
Graphically compare with flow data (if available)
If observations are flagged, attempts should be made to re-verify these data with personnel
involved in sensor programming, deployment, and retrieval. Flagged values should not be
removed from the data file unless obvious problems are found, since it is possible that an
extreme or unexplainable event is accurate and important to study. If confirmed anomalies are
deleted from the data file, document these changes (along with reasons for making these
changes).
What if sensor drift has occurred?
If drift occurs, it often occurs in very small amounts (e.g., 0.1 C per year). Those data can be retained.
In situations where there is a large amount of drift and there is no way to tell when and how much the
sensor was 'off by, the data should be removed. This is one of the reasons why the accuracy of the
sensors should be checked at the end of the deployment period.
This document is a draft for review purposes only and does not constitute Agency policy.
3 5 DRAFT—DO NOT CITE OR QUOTE
-------
Various software applications can be used to help with the error screening checks, data
management, and calculation of various summary statistics. The software that is purchased with
the temperature sensor(s) is likely going to have some applications like this. Other options
include the following free applications that can be downloaded from the U.S. Forest Service,
Rocky Mountain Research Station stream temperature website
(http://www.fs. fed.us/rm/boi se/AWAE/proj ects/stream temperature. shtml):
o Thermo Stat 3 software package (Jones and Schmidt 2012)
o A SAS temperature data processing macro
A macro for R software is not yet available.
In addition, the Washington State Department of Ecology has developed the FMU Access® Data
Logger Database, which is available upon request (Ward 2011).
2.6.3. Record keeping
Even if the set of standardized procedures described in Section 2.7.2 are followed, the data
cleaning process is inherently subjective, so both the original and the cleaned data files should be
maintained and backed up. Large amounts of stream temperature data will accumulate quickly so
a central temperature database should be developed and maintained from the initial stages of
monitoring. Also, all calibration and field forms should be organized, easily accessible, and
archived in a way that allows for safe, long-term storage.
3. STREAMFLOW
The protocols in this section discuss how to install and maintain equipment to measure
streamflow in wadeable streams, by measuring continuous stage (depth) and converting it to
discharge (streamflow). 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) (Division of Ecological Restoration 2010, Chase 2005), the United States
Geological Survey (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. There are many different types of sensors that can provide continuous monitoring
This document is a draft for review purposes only and does not constitute Agency policy.
36 DRAFT—DO NOT CITE OR QUOTE
-------
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
There are two basic components that are required to measure stream stage over time:
• A submersible pressure transducer, which provides continuous monitoring and logging of
water level. At RMN sites, transducers should be encased in housings to protect them
from currents, debris, ice, and other stressors (Section 3.1.4).
• A staff gage, which is a graduated measuring tool from which stream stage (depth) can be
read (see Section 3.1.3).
Pressure transducers record absolute pressure, which software then converts to water level.
Because atmospheric pressure changes with weather and altitude, it is necessary to compensate
for barometric variations; failure to account for these variations could result in errors of 0.6 m (2
ft) or more. 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; and 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 can be a cable that comes with the
transducer and connects directly from the data logger to a computer or it may be a data shuttle
that is purchased separately. The data shuttle used for the Onset temperature sensors (as 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
The two main types of pressure transducers are:
• 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 is downloaded directly from the on-land logger).
This document is a draft for review purposes only and does not constitute Agency policy.
37 DRAFT—DO NOT CITE OR QUOTE
-------
Data downloads are quick and there is no risk of placing the transducer
back at a different elevation.
• Data is 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., animals did not chew cable).
Extra precautions must be taken to ensure ice does not crimp vented
cable in streams with ice cover.
• The transducer can be subject to vandalism due to increased visibility of
on-land data sensor and cable.
• 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 may be easier to hide in the
stream than a vented transducer.
• There is no vented cable to maintain. These transducers may 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
software provided by the manufacturer. 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 in order to download data. Care must
be taken to ensure that the transducer is placed back in the same location
and elevation every time data is downloaded.
Transducers at RMN sites should meet the following specifications:
• Transducer accuracy - at least ±0 1%
• Range of stream stages - slightly larger than the maximum expected range in the stream.
A site visit prior to purchasing the transducer will help determine the expected range of
flows. Since accuracy decreases as range increases, chose the smallest possible range. A
depth range of less than 15 feet should be suitable at most RMN sites.
• Cable length (vented transducers only) - enough to meet installation requirements (see
Section 3.3.2.2). Some transducers come with a standard 25 foot cable while others must
be specified. A site visit prior to purchase will inform the length needed. If a site visit is
not possible, a 50-foot cable is typically sufficient.
This document is a draft for review purposes only and does not constitute Agency policy.
3 8 DRAFT—DO NOT CITE OR QUOTE
-------
Examples and pictures of some commercially available pressure transducers can be found in
Table 13 and Figure 13. Research your options carefully, and be sure to account for the fact that
non-vented transducers may require the purchase of a second transducer to measure barometric
pressure on land. Battery type (replaceable vs. non-replaceable) and memory capacity should
also be considered.
If you are using a non-vented transducer, it may be possible to use barometric pressure data from
the nearest active weather station instead of deploying an on-land transducer. This should be
evaluated on a site-by-site basis, by performing the following steps:
1) Locate the closest active weather station. This can be done using online resources like
Utah State University's Climate Database Server
(http://climate.usurf.usu.edu/mapGUI/mapGUI.php)
2) Determine whether barometric pressure data are available at the station. If so, research
the following -
• How often are the data recorded? Does the time interval overlap with the in-
stream transducer? If not, can they be adjusted to match?
• What types of quality assurance and control measures are performed on the
weather station data?
• Who runs the weather station? Where does the funding come from (and is the
funding situation stable)?
• Are there known issues (e.g., expected maintenance, upgrades) that will cause the
station to go out of operation during the period of sensor deployment?
3) Evaluate how well the weather station data are likely to approximate on-site conditions
by:
• Calculating the distance between the weather station and the site (the closer they
are, the better).
• Comparing the elevations of the sites and examining topographical differences (is
one site located in a valley and the other on a mountain?) (the less difference, the
better).
• Examine differences in weather patterns (the sites should be subject to similar
weather patterns).
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, and the weather station is expected to remain in
operation during the period of deployment, the on-land transducer could be removed from the
site and could potentially be deployed elsewhere. Whichever data source is used, it is critical that
the barometric pressure readings accurately represent on-site conditions, since failure to account
for pressure variations will result in erroneous water level measurements.
This document is a draft for review purposes only and does not constitute Agency policy.
39 DRAFT—DO NOT CITE OR QUOTE
-------
There is also value in obtaining precipitation data from the nearest weather station (if available),
since precipitation and flow are often closely linked. As a quality assurance check, patterns in the
precipitation data could be compared to patterns in the water level data. If the patterns differ, this
could be an indication that the water level sensor is malfunctioning. In addition, a similar quality
assurance check could be performed using flow data from the nearest USGS gage.
A)
B)
C)
Figure 13. 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); D) In-Situ Level TROLL.
This document is a draft for review purposes only and does not constitute Agency policy.
40 DRAFT—DO NOT CITE OR QUOTE
-------
Table 13. Examples of commercially available pressure transducers
Manufacturer &
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
Type1
Non-
vented
Vented
Vented
or
Non-
Vented
Vented
or
Non-
Vented
Operation
range
-20° to 50°C
-30° to 85°C
-15° to 55° C
-20° to 80°C
Accuracy2
±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) &
replaceability
5 years with 1
minute or greater
logging interval;
factory
replaceable
Up to 1 year
(depending on
recording
intervals); user
replaceable (2 9V
DC batteries)
18 months at
15 -minute
interval; user
replaceable
10 years or 2
million readings
Logger
Memory
Approx.
21,700
records4 (64K
bytes)
Approx.
81,759 records
Available in
130,000,
260,000 and
520,000 record
versions
130 000
records (2.0
MB)
Approx.
price3
($)
$495
(instream)
+ $495
(on land)
$989 (with
25 ft. of
cable)
$1,095
(plus
$2.35/ft
cable)
$1,170
(plus cable
if needed)
Web site
www.onsetcomp.com
www.globalw.com
http://inwusa.com
www.in-situ.com
1 Transducers can be either vented or non-vented (see Section 3.1.2 for more information on what this means). 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.
2Accuracy 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.
3As of January 2013 and subject to change; reduced prices may be available for bulk orders.
4Readings can be taken at 15-minute intervals for approximately 112 days before the memory capacity is reached.
777/5 document is a draft for review purposes only and does not constitute Agency policy.
41 DRAFT—DO NOT CITE OR QUOTE
-------
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 provide a stable
attachment point for the transducer. Gages at RMN sites should be USGS style (Style A) and
marked every 0.02 feet (Figure 14).
Figure 14. 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. If vandalism is a concern, the housings can be painted
black or camouflage to make them less visible. Inexpensive housings can be constructed from
1.25" or 1.5" diameter Schedule 40 (or stronger) PVC pipe. Multiple /^ inch or larger holes
should be drilled into the PVC to allow water to fluctuate at the same rate as in the stream
(Figure 15). In streams that are subject to extended ice cover, the entire length of the transducer
as well as the vented cable may be encased in PVC to prevent potential damage.
This document is a draft for review purposes only and does not constitute Agency policy.
42 DRAFT—DO NOT CITE OR QUOTE
-------
Figure 15. 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. If vandalism is a concern, the PVC can be painted to make it less visible.
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 re-deployment.
3.2.1. Calibration
Most transducers will be factory calibrated prior to being shipped. The calibration 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 16 illustrates how staff gage readings can be used to check the accuracy of the transducer
data over time. The pressure transducer data is compared with the staff gage readings at similar
times to determine the offset between the two. If transducer data does not correspond to staff
gage data and water depth, there may be fouling of the transducer, it may need to be recalibrated,
and/or it may 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 re-calibrate 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 may need to be returned for service.
This document is a draft for review purposes only and does not constitute Agency policy.
43 DRAFT—DO NOT CITE OR QUOTE
-------
•u
to
ts
4.0
3.5
3.0
2.5
2.0
1.5
Transducer data
Staff gage readings
S«P
Oet
Nov
Dec
Jan 2013
Date
Figure 16. 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 may have changed
elevation.
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 is 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.
It is helpful to bring a computer into the field to launch the transducer and check the logged data
once it is installed. Alternatively the transducer may be launched prior to deployment as long as
a note is made of the date and time of actual deployment in the stream (this is important for data
screening - Section 3.7.2).
This document is a draft for review purposes only and does not constitute Agency policy.
44 DRAFT—DO NOT CITE OR QUOTE
-------
3.3. Deployment of Instream Pressure Transducers
Installation of both staff gages and pressure transducers are covered in this section since 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. Prior to
installing the equipment, permissions from all relevant parties should be obtained. 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. Efforts should be made 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, it may be best to attach equipment to man-made structures like
bridges. During the site visit, determine what type of installation is best so that appropriate
equipment can be obtained prior to installation. If a situation occurs in which a transducer is
attached to a bridge or other man-made structure, the biological sample should be collected at
least 200 meters from the area of human disturbance. As long as 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 sited in the exact location that biota are being
collected from.
Key considerations for siting in-stream transducers are listed below. These are similar to those
used for water temperature sensors (see Section 2.3.1). More detailed information on site
selection and controls may 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 was collected. Note that for general monitoring purposes the
sensor does not have to be placed in the exact location that biota are being collected from.
• 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 17, Rantz et al. 1982). Natural controls may
include a downstream riffle, bedrock outcrop, or other stream feature that controls flow.
Unnatural controls may include a bridge or culvert that is narrower than the stream
This document is a draft for review purposes only and does not constitute Agency policy.
45 DRAFT—DO NOT CITE OR QUOTE
-------
channel and constricts flow. Note that the feature controlling the stage-discharge
relationship may change at different flow levels; such changes will be reflected in the
rating curve.
• The site should not have extensive aquatic vegetation, beaver activity, and/or unstable
streambeds and banks. These factors can change or result in unstable stage
measurements.
Figure 17. 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 18). 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 potentially knock over
the gage. A list of equipment needed for installation is in Table 14.
Gages are typically available in 3.33 foot sections; multiple gage sections may be combined to
encompass the entire range of stages expected to occur in that section of the stream (additional
sections may 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.
This document is a draft for review purposes only and does not constitute Agency policy.
46 DRAFT—DO NOT CITE OR QUOTE
-------
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 19). Use concrete anchors on both sides of board and at varying
elevations to ensure the stability of the board. Two anchors are typically
sufficient for a three-foot gage and three for a six-foot gage.
• Use a hammer drill or rotary hammer with concrete drill bit to drill holes in the
marked locations.
• Pound concrete wedge anchors (bolts) into holes using a hammer until they are
secure.
• Drill holes in gage board for concrete wedge anchors.
• Place board over bolts and screw into place using nut.
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 will be able to
go into the bed.
• Drive a galvanized steel pipe with cap into the streambed using a pole (fencepost) driver
or sledgehammer. Six foot and nine foot poles 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 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 may 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 flow and, if
possible, positioned so that it may be read from the stream banks.
If there is a nearby tree on the stream bank, a metal bracket or brace may be attached from the
gage to the tree to increase stability.
This document is a draft for review purposes only and does not constitute Agency policy.
47 DRAFT—DO NOT CITE OR QUOTE
-------
Figure 18. Examples of gage installation techniques include, A. streambed, B. fixed object (bridge wing wall) and
C. fixed object (boulder).
Figure 19. Example of a 0.5 x 3.75 inch wedge anchor with bolt and washer.
This document is a draft for review purposes only and does not constitute Agency policy.
48 DRAFT—DO NOT CITE OR QUOTE
-------
Table 14. Equipment list for staff gage installation
Location
Supplies 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
This document is a draft for review purposes only and does not constitute Agency policy.
49 DRAFT—DO NOT CITE OR QUOTE
-------
3.3.3. Pressure transducer installations
Two types of installation techniques for pressure transducers are described in this document: 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. If site conditions permit,
transducers at RMN sites should be installed using the fixed object method since it is more likely
to withstand floods and associated bed load movement, and it is easier to ensure that the
transducer remains in the same place over its period of deployment.
3.3.3.1. Fixed object
With the fixed object method, 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 PVC housing (Section 3.1.4). For vented
transducers, which do not need to be removed from the stream to download data, the PVC pipe
should be cut to a length at least slightly longer than the transducer. Use zip ties to attach the
transducer to the PVC pipe to ensure it does not move (Figure 15). Do not leave slack in the zip
ties. If the unit can move, it can wear through the zip tie.
Non-vented pressure transducers typically need to be removed from the stream to download data.
To facilitate the removal and download of data during periods of high water, 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 20). 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 20). Clearly mark the holes for the bolt so that the transducer is placed back
at the same elevation every time it is removed.
This document is a draft for review purposes only and does not constitute Agency policy.
50 DRAFT—DO NOT CITE OR QUOTE
-------
Figure 20. Non-vented pressure transducer installation. A. Non-stretch cable is attached to transducer, B. Loop is
made at non-transducer end of cable, C. Cable with transducer is suspended over bolt from top of PVC pipe.
Once the pressure transducer has been installed in the protective housing, the transducer/housing
can then be attached vertically to the staff gage board (Figure 21), a bridge abutment or fixed
object, or horizontally to rebar or other pipe in the stream bed. An equipment list for performing
installations can be found in Table 15.
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.
This document is a draft for review purposes only and does not constitute Agency policy.
51 DRAFT—DO NOT CITE OR QUOTE
-------
Figure 21. Non-vented (left) and vented (right) pressure transducers attached to staff gage board using conduit
hangars.
To install the transducer/housing on a fixed object (e.g. bridge abutment):
• Install the transducer in close proximity to the staff gage so that stage changes are
comparable.
• Use concrete anchors and conduit straps, hangar, 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.
This document is a draft for review purposes only and does not constitute Agency policy.
52 DRAFT—DO NOT CITE OR QUOTE
-------
Table 15. Equipment list for transducer installation
Location
Supplies list
All
• PVC pipe for transducer (drilled with Va" 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).
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)
Fixed object-
additional
items
Hammer or rotary drill
Concrete drill bit
Concrete wedge anchor and nuts (stainless steel or galvanized)
Hammer
3.3.3.2. Streambed/rebar
If it is not possible to attach the transducer to a staff gage board or other fixed object in the
stream (Section 3.3.3.1), a streambed installation may be possible (Figure 22). With this method,
the transducer is laid horizontally on the streambed and is held in place by rebar and rocks.
This document is a draft for review purposes only and does not constitute Agency policy.
53 DRAFT—DO NOT CITE OR QUOTE
-------
Compared to the fixed object method, transducers are more prone to moving or being swept
away during high flows, and it may be more difficult to ensure that transducers are returned to
the exact same location after being removed for data downloads. However, this technique has
been shown to be effective in short-term deployments (Roy et al. 2005).
Streambed installations require rebar stakes and a sledgehammer in addition to the equipment
listed in Table 15. Installation instructions are as follows:
• Find a pool location in close proximity 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, and/or place large
rocks over the PVC pipe.
• Ensure that the PVC pipe and transducer are installed slightly above the streambed to
reduce chances of fouling by fine sediments.
Staff gage
PVC housing Rebar
holding
transducer
Figure 22. Examples of streambed installations of pressure transducers using rebar.
This document is a draft for review purposes only and does not constitute Agency policy.
54 DRAFT—DO NOT CITE OR QUOTE
-------
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 trees are absent, install a stainless steel pole (with a cap on
it) and attach the data logger to this. Use the same type of pole that is used for the streambed
staff gage installations. Do your best to 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
may be suspended on a wire or cable. If the data logger is not supported by the PVC pipe, it
should be suspended on a wire or cable so there is no pressure 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 for
downloading 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 23). Adjust the length of the
pipe to adequately cover the device with some extra room.
• 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:
• 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 may be encased in a PVC pipe.
Specific considerations for the barometric pressure sensor:
• 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 24)).
This document is a draft for review purposes only and does not constitute Agency policy.
55 DRAFT—DO NOT CITE OR QUOTE
-------
Figure 23. Vented transducer data logger installation using A. stainless steel conduit straps and B. cable ties.
Figure 24. Barometric pressure sensor installation for a non-vented transducer A. without the PVC cap on top and
B. with the PVC cap.
This document is a draft for review purposes only and does not constitute Agency policy.
56 DRAFT—DO NOT CITE OR QUOTE
-------
3.5. Elevation Surveys and Documentation
After the installation is complete, the pressure transducer location should be georeferenced and
documented in a way that allows field personnel to re-locate the sensor during subsequent visits.
Photos should be taken from different perspectives (e.g., upstream and downstream); these are
important for relocating the instruments and documenting any changes to the monitoring location
during the course of the study. All documentation information must be recorded on a field form
that includes information such as station number, waterbody name, date, time, crew members,
driving directions, serial number of sensor/s, time of deployment, transducer installation
technique, image numbers/file names for the photographs, and detailed descriptions of transducer
placement. It is very important to accurately record the time and date of deployment, as this
information will be used in the error screening procedure described in Section 3.7.2.
In addition, the elevation of the staff gage and pressure transducer should be surveyed to
establish a benchmark or 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. Elevation surveys should be
conducted at least once a year to identify if and when movement occurs. It is particularly
important to check for movement after high flow events and periods of extended ice cover. Steps
for conducting elevation surveys using an auto level are summarized in Table 16, and Table 17
contains a list of equipment needs.
The elevation survey should be conducted on the day of staff gage and transducer installation. A
benchmark and one to two other permanent markers should be identified or established 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. It is not necessary to know the absolute elevation of the
benchmark as the purpose is to detect changes in elevation of the equipment relative to the
benchmark. The benchmark may therefore 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 25); 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.
Once markers are established, sketch their location on the data sheet (see example in Figure 26),
clearly documenting both markers and relevant site features that may 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 16). 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:
This document is a draft for review purposes only and does not constitute Agency policy.
57 DRAFT—DO NOT CITE OR QUOTE
-------
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, 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
previous surveyed elevations to see if movement occurred.
Figure 25. Example of auto level and tripod used for elevation survey (A) and of a permanent structure used as a
benchmark (B).
777/5 document is a draft for review purposes only and does not constitute Agency policy.
58 DRAFT—DO NOT CITE OR QUOTE
-------
Table 16. 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.
2. Level auto level by centering the level bubble using the foot screws.
3. Have person hold stadia rod on top of benchmark. Rod must be level.
4. Read elevation at which crosshairs in auto level cross the stadia rod.
5. Record this elevation as the starting backsite (B.S.).
6. Have person hold stadia rod on survey point #1. Rod must be level.
7. Read elevation at which crosshairs in auto level cross the stadia rod.
8. Record this elevation as the foresite (F.S.).
9. Repeat steps 6 to 8 for additional survey points (including the top of the staff
gage and transducer).
10. Have person hold stadia rod on top of benchmark. Rod must be level.
11. Read elevation at which crosshairs in auto level cross the stadia rod and record
as final B.S.
12. If final B.S. does not match starting B.S. (within 0.01 ft), repeat the entire
survey.
Table 17. Equipment list for elevation surveys
Supplies list
• Auto level or
• Tripod
• Stadia rod
• Survey paint
• Survey nails
• Datasheet
laser level
This document is a draft for review purposes only and does not constitute Agency policy.
59 DRAFT—DO NOT CITE OR QUOTE
-------
Site:
River Instream Flow Stewards (RIFLS)
Staff Gauge Survey Sheet
Date:
r M
Notes:
f
Gage ttt):_(>. °C\ PT.
BS = backsite, HI = height of instrument, FS = fofesite, BM = benchmark, TP = turning point
H.I. = B,S. + Elevation, Elevation = H.I. • F.S.,
flU-ft
R
ItPT
IfiflL
MI31
51751
Notes and Site Sketch:
NT
Figure 26. An example of a completed elevation survey form.
This document is a draft for review purposes only and does not constitute Agency policy.
60 DRAFT—DO NOT CITE OR QUOTE
-------
3.6. Maintenance/Mid-Deployment Checks and Data Offload
When a site is first established, it should be revisited within the first month to confirm that the
installation is holding properly. After these initial deployment checks, at a minimum, sites should
be visited at least annually (e.g., in conjunction with the biological sampling events), but more
frequent visits are encouraged, 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 as well as data quality.
During site visits, field personnel should:
• Check the condition of the transducer and perform necessary maintenance.
• Take staff gage readings, ideally at different flow conditions, for quality assurance and
control purposes.
• Offload data.
Typical maintenance issues include:
• Ensuring that the instream pressure transducer is submerged and not buried in sediment.
• Clearing leaf litter and debris that may pile up against the gage, transducer, and
downstream control. At the beginning of a site visit, clear this material as it may 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 may change if there is
sediment accumulation or scour 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 alternate 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 off 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) may 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, data should be downloaded during each site visit. Frequent data downloads, as well
as frequent stage measurements from the staff gage, will help in early identification of transducer
This document is a draft for review purposes only and does not constitute Agency policy.
61 DRAFT—DO NOT CITE OR QUOTE
-------
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 27). Prior to download, record the stage at
the staff gage and note any factors which may 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, a form like the one shown in Figure 28 should be completed.
Battery life should also be checked. Batteries in some sensors need to be replaced each year,
while others last 5 years or longer under normal use. Do your best to plan for these replacements,
and try to minimize the number of different transducers that need to be deployed at a site through
time so that inter-instrument error can be minimized.
Information from mid-deployment checks should be recorded in a field notebook or on a field
form. Entries should include notes about the condition of the transducer, staff gage readings and
whether any unusual measurements appeared during the data spot-check (if available). It is also
important to take photos during each site visit, as this will help document changes to the
monitoring location during the course of the study
Figure 27. 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
ffJLh 1&\!WL AT SOOTH CflUUTW t-OAT) PLaftH5»4.MA-
-T-: ' *—.!.*-
S/12-
Time:
t <*,<;
11 T \/
Transducer (ft):
Gage Survey?
Batteries Changed?
Of-
Figure 28. Example of a transducer download field data sheet.
777/5 document is a draft for review purposes only and does not constitute Agency policy.
62 DRAFT—DO NOT CITE OR QUOTE
-------
3.7. Quality Assurance and Control
Proper quality assurance and control 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 may 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, it is important to make
periodic staff gage readings to detect any shift or drift in the transducer data. The gage depth
should be recorded every time the site is visited for a discharge measurement or data download.
It is ideal for gage readings to be done at least every month, if possible, to assure a variety of
stages are 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 may assist with gage readings. Local watershed groups or other
organizations may be able to help to identify local volunteers. At the minimum, try to visit sites
after large storm events that may impact the transducer.
3.7.2. Error screening
Pressure transducers may record erroneous readings during deployment for a variety of reasons
(e.g., they may become dewatered during low flow conditions, high flow events may bury them
in sediment, humans may cause interference). The types of errors that can occur and how they
manifest themselves will vary. For example, if moisture gets into the cable, it may result in
erratic readings or readings of zero water depth. If the cable gets kinked or plugged, it can result
in the data not being corrected for barometric pressure. Because these errors may occur, data
need to be screened.
The first step involves removing observations recorded before and after the transducer is
correctly positioned in the stream channel. 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 dewatered or buried in the sand) and flag those data accordingly.
Next, perform the series of checks described below:
• Outliers. Graphing data over time and against precipitation data from a nearby weather
station provides a quick way to identify and flag obvious outliers (e.g., negative numbers
during periods of normal streamflow, very high numbers when there was no precipitation
This document is a draft for review purposes only and does not constitute Agency policy.
63 DRAFT—DO NOT CITE OR QUOTE
-------
or other known cause, data gaps). Evaluate and, if possible, correct the cause of data
outliers.
• Accuracy. Compare staff gage readings to appropriate transducer measurements. If the
difference between staff gage and transducer readings at a given time is more than 5%,
additional analysis will be necessary to determine if a data shift or data drift occurred
during the deployment period.
• Data shift/drift. Data drift is when the difference between the staff gage and transducer
readings changes over time. A data drift may be detected by graphing the difference
between staff gage and transducer readings over the deployment period. If the difference
between the two readings increases over time it is likely that data drift occurred. If the
difference between the two readings suddenly changes and then remains constant over the
deployment, it is possible that a data shift occurred due to changing elevation of the
transducer relative to the gage.
• Transducer sensitivity drift. Sensitivity drift is when the sensitivity of the transducer
changes with stage (e.g., the transducer is less sensitive or accurate at high stages).
Sensitivity drift may 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 may indicate a sensitivity shift.
If a shift is detected and a follow-up elevation survey is performed, water level 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 (see Figure 16).
If there is no gage data for the time period, 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. More detailed information on data drift and correcting datasets
can be found in Shedd and Springer (2012).
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.
3.8. Developing Frames of Reference
Following the procedures described above will help measure stream stage. Taken alone, stage
measurements yield some information about streamflow patterns, including the timing,
frequency, and duration of high flows (McMahon et al. 2003). However, stage data itself does
not give quantitative information about the magnitude of streamflows or flow volume, which
makes it hard to compare data between streams. Furthermore, the channel shape may change
from year to year, such that a similar stage measured in a given location during two separate
This document is a draft for review purposes only and does not constitute Agency policy.
64 DRAFT—DO NOT CITE OR QUOTE
-------
years may represent different flows. Thus, in order to assess patterns and changes in stream
hydrology, it is most useful to convert stage measurements into streamflow.
The most common approach is to use a stage-discharge rating curve. Section 3.8.1 briefly
summarizes how to develop a rating curve, with references from the USGS and several state
agencies which provide much greater detail on discharge measurement methods and rating curve
development. Section 3.8.2 discusses an alternative method which involves less field time, using
a combination of channel cross-section measurements and modeling.
3.8.1. Stage-discharge rating curves
A rating curve allows the user to convert stage measurements to streamflow. In order to develop
a rating curve, a series of discharge (streamflow) measurements are made at a variety of stages,
covering as wide a range of flows as possible. The following discussion summarizes discharge
measurements in wadeable streams. Discharge measurement involves measuring the depth and
velocity of the water passing through a number of 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. For more detailed guidance consult
Rantz et al. (1982), Shedd (2011), or Chase (2005).
3.8.1.1. 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
over as wide a range of flows as possible. After establishing a rating curve, discharge should be
measured at least once annually, and if possible, also after large storms or any other potentially
channel-disturbing activities, in order to verify or (if needed) update the curve (see Figure 29 for
an example of a channel-disturbing activity). If new measurements are more than 15% off of 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.
This document is a draft for review purposes only and does not constitute Agency policy.
65 DRAFT—DO NOT CITE OR QUOTE
-------
Figure 29. 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.
This document is a draft for review purposes only and does not constitute Agency policy.
66 DRAFT—DO NOT CITE OR QUOTE
-------
3.8.1.2. 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 involved includes:
• Current meters measure point velocity. There are many different types 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.
3.8.1.3. Site selection
Site selection is critical to making a good discharge measurement. An ideal cross section will
have the following characteristics (see Figure 30 for examples):
• A relatively straight stream channel with defined edges and a fairly uniform shape.
• Limited vegetative growth, large cobbles, and boulders.
• 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 30 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 may meet many of the above criteria and provide a good
location to measure flow.
This document is a draft for review purposes only and does not constitute Agency policy.
67 DRAFT—DO NOT CITE OR QUOTE
-------
Figure 30. Examples of good cross-sections for making discharge measurements.
3.8.1.4.
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. These
points should be distributed such that an approximately equal percent of total flow is in each
segment (Figure 31). Thus, measurement points should be closer together in portions of the cross
section where flow is more concentrated and depths are greatest and farther apart where the flow
is lowest and depths are shallowest. No more than ten percent of the total flow should be within
any one segment. Include additional segments where velocity or bottom irregularities are the
greatest.
The exact methodology for making velocity measurements is somewhat specific to the
instrument being used. For more details refer to Rantz (1982), Shedd (2011), Chase (2005), or
City of Salem (2007), as well as to instrument manuals.
This document is a draft for review purposes only and does not constitute Agency policy.
68 DRAFT—DO NOT CITE OR QUOTE
-------
Measure distance along tagline,
channel depth, and stream
velocity at a minimum of 20
points along the cross section.
Figure 31. Layout of a channel cross-section for obtaining discharge data, using the velocity-area procedure.
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. It is
best to measure the GZF at each discharge measurement, but it may be most practical and
effective to locate this point at lower flows.
3.8.1.5.
Documentation
At the time of the discharge measurement, take photos 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.
• 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 (if measured) and location of PZF measurement.
This document is a draft for review purposes only and does not constitute Agency policy.
69 DRAFT—DO NOT CITE OR QUOTE
-------
Examples of discharge datasheets for different flow meters can be found in Appendix J.
3.8.1.6. Quality assurance and control
Careful reading and use of the documents in Section 4 (Literature Cited) is important for good
QA/QC of discharge measurements. 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, but be sure that no
one is in the stream upstream during measurements, as this will affect the flow readings.
The difference between the two measurements should be less than 15%.
• 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 an experienced
hydrographer from the USGS or another agency.
• Major, channel-disturbing events (e.g., floods, new culverts) may alter the rating curve. If
a major event occurs and subsequent points are not aligned with the original rating curve,
a new rating curve may need to be developed and used to convert stage to discharge for
points following that event.
3.8.1.7. 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 may 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 32.
This document is a draft for review purposes only and does not constitute Agency policy.
70 DRAFT—DO NOT CITE OR QUOTE
-------
- 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
Grade
0.5
0.4
Grade
.1
10
Discharge (ftA3/s)
Discharge (ftA3/s)
Figure 32. Example of regular (left) and log-log scale (right) rating curves created using Aquatic Informatics'
AQUARIUS software.
3.8.2. Channel cross-section measurements and modeling
Converting stage to discharge via a flow rating curve is the best approach for quantifying a wide
array of hydrologic indicators that can be compared across time or space. However, for short-
term deployments of pressure transducers or sites where making the necessary discharge
measurements to develop a rating curve is infeasible, it is possible to calculate stage-based
parameters that are referenced to the cross-section where stage measurements are recorded. This
approach requires surveying of the cross-sectional profile where the transducer is located, and
then modeling the flow to get a mean hydraulic depth for calculating relative measurements of
magnitude and volume. This approach was used by Roy et al. (2005) where stage was measured
at 30 sites for one year.
3.8.2.1. Equipment
To generate a profile of the channel cross-section, the following equipment is needed:
• An automatic level or electronic total station and tripod, or a clinometer is used to get the
height at a consistent level along the cross section.
• A stadia rod is used to measure heights along the cross section.
• A measuring tape and stakes are used to define the exact location of the cross section and
measure the widths at which points are measured.
3.8.2.2.
Site selection
This document is a draft for review purposes only and does not constitute Agency policy.
71 DRAFT—DO NOT CITE OR QUOTE
-------
A profile of the channel cross-section should be made at the precise location of the pressure
transducer. The measurements should encompass the entire bankfull width, and have sufficient
measurements to get an accurate view of the profile. For stream slope, measurements should be
made for the stream reach encompassing the transducer and, if possible, the biotic sampling.
3.8.2.3. Measurements
For the cross section, start by pinning the measuring tape to the left and right banks
perpendicular to the streamflow and encompassing the entire bankfull width. If a tripod is being
used, place it at a location where the entire width is visible, and where the scope is higher than
the highest elevation. Take height measurements along the entire cross section. Be sure to
include points at the top and bottom of every break in elevation. More frequent points are
necessary where elevation varies. When using an automatic level or tripod, record both the width
and the height at each point (an electronic total station will record the height and distance
automatically).
For stream slope, take elevation points at riffle tops along a 100-meter reach or longer. If using
an automatic level, record longitudinal location (from a tape measure) and height at each point.
The electronic total station will capture height and distance automatically at measured points. If
the tripod must be moved to view the entire reach, be sure to survey a bench mark and one
instream location twice (from each tripod position) and adjust the heights recorded from one
tripod location to match the other.
Additional information about measuring cross-sectional profiles and stream slope can be found
in Gordon et al. (2004).
3.8.2.4. Documentation
The following data should be recorded:
• Date, time, field crew members.
• Staff gage reading.
• Width, height, and type (water, bank, floodplain) at each recorded location along cross
section.
• Longitudinal distance and height for each riffle top for stream slope.
• Bankfull and water width edge locations.
• Picture and hand-drawn sketch of the profile.
3.8.2.5. Quality assurance and control
This document is a draft for review purposes only and does not constitute Agency policy.
72 DRAFT—DO NOT CITE OR QUOTE
-------
Ideally, the transducer will be located in a section with high stability and minimal expected
changes in the cross-section. However, bed scour and sediment deposition caused by high flows
can alter the cross-sectional area and, subsequently, the volume of water associated with certain
stages. Thus, the cross-section should be re-surveyed after major events or at least once annually.
3.8.2.6. Modeling
HEC-RAS (US Army Corps of Engineers, Davis, CA) is a freely-available software that can be
used to model depths of certain recurrence interval floods based on bankfull cross-sectional area,
stream slope, Manning's n (a measure of stream roughness), and discharge.
Stream slope should be determined by plotting the riffle-top elevation vs. longitudinal distance;
the slope of the linear trendline through the points is the stream slope. Manning's n can be
visually estimated based on stream type or comparison to photographic keys (Gordon et al.
2004). Discharge for a certain recurrence interval (RI) flood can be determined for each stream
based on subcatchment area using published flood-frequency formulas. Roy et al. (2005) related
metrics to the 0.5-year RI flood because this level encompassed the mean depth for a majority of
the streamflows over the study period. Bankfull cross-sectional area, stream slope, Manning's n,
and discharge can then be entered into HEC-RAS steady flow analysis to generate a mean
hydraulic depth for a certain RI flood at each site.
4. LITERATURE CITED
Bierwagen et al. 2013, in review. Analytical Foundation for a Monitoring Network to Detect
Climate Change-Related Effects in Streams in the Northeastern United States
Bilby, R.E. and G.E. Likens. 1980. Importance of organic debris dams in the structure and
function of stream ecosystems. Ecology 61: 1107-1113.
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
Butkus, S. 2007. Quality Assurance Monitoring Plan: Streamflow Gaging Network. Washington
State Department of Ecology, Environmental Assessment Program. Available online:
http://www.ecy .wa.gov/programs/eap/flow/shu_main.html
Chase, R. 2005. Standard Operating Procedure: Streamflow Measurement. Massachusetts
Department of Environmental Protection, Division of Watershed Management. CN 68.0.
This document is a draft for review purposes only and does not constitute Agency policy.
73 DRAFT—DO NOT CITE OR QUOTE
-------
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.city ofsal em.net/Departments/PublicWorks/Operations/StormwaterServices/D
ocuments/qapp_and_sop_4-07.pdf
Clarke, A., Mac Nally, R., Bond, N. and P.S. Lake. 2008. Macroinvertebrate diversity in
headwater streams: a review. Freshwater Biology 53: 1707-1721.
Division of Ecological Restoration. 2010. River Instream Flow Stewards (RIFLS) Quality
Assurance Project Plan. Massachusetts Department of Fish and Game. (Available on-line
soon).
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.
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.
Hill, R.A., Hawkins, C.P. and D.M. Carlisle. 2013. Predicting thermal reference conditions for
USA streams and rivers. Freshwater Science 32(1): 39-55.
Holden, Z.A., Klene, A., Keefe, R. and G. Moisen (in press). Design and evaluation of an
inexpensive solar radiation shield for monitoring surface air temperatures. Agricultural
and Forest Meteorology.
This document is a draft for review purposes only and does not constitute Agency policy.
74 DRAFT—DO NOT CITE OR QUOTE
-------
Holden, Z.A., Abatzoglou, J., Luce, C. andL.S. Baggett. 2011. Empirical downscaling of
minimum air temperature at very fine resolutions in complex terrain. Agricultural and
Forest Meteorology.
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
Isaak, D. J.; Horan, D. L.; and Wollrab, S. P. 2013. A Simple Protocol Using Underwater Epoxy
to Install Annual Temperature Monitoring Sites in Rivers and Streams. Gen. Tech. Rep.
RMRS-GTR-###. Fort Collins, CO: U.S. Department of Agriculture, Forest Service,
Rocky Mountain Research Station. ## p. Available online:
http: //www.fs.fed.us/rm/boise/AWAE/proj ects/stream_temp/downloads/2013_StreamSen
sorEpoxyProtocol RMRS GTR.pdf
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.
Karl, T.R., J.M. Melillo, and T.C. Peterson, (eds.). 2009. Global Climate Change Impacts in the
United States. Cambridge University Press, 189 pp. Available online:
http://www.globalchange.gov/publications/reports/scientific-assessments/us-impacts.
Kaushal, S. S., G. E. Likens, N. A. Jaworski, M. L. Pace, A. M. Sides, D. Seekell, K. T. Belt, D.
H. Secor, and R. L. Wingate. 2010. Rising stream and river temperatures in the United
States. Frontiers in Ecology and the Environment 9:461-466.
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.
This document is a draft for review purposes only and does not constitute Agency policy.
75 DRAFT—DO NOT CITE OR QUOTE
-------
Kennen, J.G., DJ. 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.
Larsen, D., P. Kaufmann, T. Kincaid and N. Uruquhart. 2004. Detecting persistent change in the
habitat of salmon-bearing streams in the Pacific Northwest. Canadian Journal of Fisheries
and Aquatic Sciences 61: 283-291.
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., Purcell, A.H., and Resh, V.H. 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.
Meyer J.L., StrayerD.L., Wallace J.B., Eggert S.L., Helfman G.S. andN.E. Leonard. 2007. The
contribution of headwater streams to biodiversity in river networks. Journal of the
American Water Resources Association 43: 86-103.
Mohseni, O., Stefan, H.G. and T.R. Erickson. 1998. A nonlinear regression model for weekly
stream temperatures. Water Resources Research 34(10): 2685-2692.
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/
Rieman, B. E., and D. J. Isaak. 2010. Climate change, aquatic ecosystems and fishes in the
Rocky Mountain west: implications and alternatives for management. U.S. Forest
Service, Rocky Mountain Research Station, GTRRMRS- 250, Fort Collins, Colorado.
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.
This document is a draft for review purposes only and does not constitute Agency policy.
76 DRAFT—DO NOT CITE OR QUOTE
-------
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.ecy.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
Sidle R.C., Tsuboama Y., Noguchi S., Hosoda I, FujiedaM. and T. Shimizu. 2000. Stormflow
generation in steep forested headwaters: a linked hydrogeomorphic paradigm.
Hydrological Processes 14: 369-385.
Sowder, C., and E. A. Steel. 2012. A note on the collection and cleaning of water temperature
data. Water 4: 597-606.
U.S. Forest Service, Rocky Mountain Research Station. 2013. Stream Temperature Modeling
and Monitoring [web page]. Available online:
http://www.fs.fed.us/rm/boise/AWAE/projects/stream_temperature.shtml
Vannote, R.L., Minshall, G.W., Cummins, K.W., Sedell, J.R. and C.E. Cushing. 1980. The River
Continuum Concept. Canadian Journal of Fisheries and Aquatic Sciences 37: 130-137.
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
Wipfli, M.S., Richardson, J.S. and R.J. Naiman. 2007. Ecological linkages between headwaters
and downstream ecosystems: transport of organic matter, invertebrates, and wood down
headwater channels. Journal of the American Water Resources Association 43: 72-85.
Zimmerman, C.E. and J.E. Finn. 2012. A Simple Method for In Situ Monitoring of Water
Temperature in Substrates Used by Spawning Salmonids. Journal of Fish and Wildlife
Management 3(2): 288-295. Available online:
http://www.fwspubs.org/doi/pdf/10.3996/032012-JFWM-025.
This document is a draft for review purposes only and does not constitute Agency policy.
77 DRAFT—DO NOT CITE OR QUOTE
-------
Appendix A
How to construct PVC housings for water temperature
sensors with cable installations
-------
Appendix A (Mauger 2008)
Housing Construction
PVC housings are very simple to make, 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 solar radiation to penetrate. Black PVC provides more camouflage than white PVC for
sites where vandalism is a concern. In clear water streams, heat absorption by the dark surface
may be an issue so white PVC is recommended.
Here's a supply list to make the housings:
"
Sch 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
Multi purpose cement
Assorted nuts and bolts
Glue the female adapters to each end of the PVC pipe. Drill a hole for one
eyebolt to go through a clean out plug. Drill at least 20 holes in PVC to allow
water flow. Secure the eye bolt through the clean out plug with appropriate-
sized nuts and bolt. Use a cable tie through 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 to weigh it down. Screw the clean out plugs into the
female adapters.
-------
Appendix B
How to make and install homemade radiation shields for
air temperature sensors
-------
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.
-------
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 ziptie to
maintain space between the 2 lower plates housing the air sensor. There should be approximately a
1.25-1.5 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.
-------
Appendix C
Temperature sensor calibration forms
-------
Table Cl. Form for doing a multiple-point temperature calibration in which 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
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-
-------
Table C2. Form for checking temperature sensor readings against one another.
Technician:
Date/s of calibration check:
Overall mean temperature (all sensors) (°C):
Sensor S/N
Mean
Temperature
(°Q
Difference
from overall
mean
Minimum
Temperature
(°C)
Maximum
Temperature
(°C)
Notes
-------
Appendix D
Temperature sensor deployment & tracking forms
(fill out one for each sensor)
-------
S/N: Station number:
Water or air: Waterbodyname:
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 File names of photos:
Recording interval:
Units:
Start time:
Installation technique:
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)
-------
Task
Pre-deployment accuracy check
Launch/start
Deployment (in the water)
Mid-deployment accuracy
check (in Notes, write down
NIST-calibrated field
thermometer reading and
sensor reading)
Data download
Sensor Retrieval
Pre-deployment accuracy check
Date
Time
Technician
Notes
-------
Appendix E
Equipment lists for temperature sensor procedures
-------
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
calibration
procedure
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 (i.e. red liquid) thermometers (optional)
• Containers to hold the sensors
• Water
• Refrigerator
• Clock or watch
• Calibration data sheet (Attachment C)
• Computer that has the appropriate software for
temperature sensor
reading the
Equipment list for doing underwater epoxy installations (Isaak et al. 2010).
Task
Supplies list
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
Monument
Metal forestry tags
Spray paint
GPS
Camera
Field form
-------
Equipment list for doing cable installations (Ward 2011, Mauger 2008).
Task
Installation
Supplies list
• Temperature sensor
• Radiation shield (see Attachment A (Mauger 2008))
• Cable
• Rebar (3/8 to 1/2", 2-4 ft lengths)
• Rebar pounder (for custom design, see Ward 201 1 -
Attachment B)
• For more information see checklist in Ward 201 1
(Attachment A) and/or Attachment G (Mauger 2008)
Equipment list for documenting sites.
Task
Georeferencing
& monumenting
Measuring
temperature and
depth
Supply list
• GPS
• Camera
• Map and/or gazetteer
• Metal forestry tags
• Field form
• NIST-calibrated field thermometer and/or multi-probe meter
• Meter ruler or calibrated rod/pole (i.e. surveyor's rod)
• Measuring tape
Equipment list for conducting maintenance/mid-deployment checks and data downloads.
Task
Relocating
sensor
Documenting
on-site-
conditions
Data offloads
Measuring
temperature and
depth
Back-up
equipment (in
case a sensor
needs to be
replaced)
Supply list
• GPS
• Map and/or gazetteer
• Annotated photos and/or hand-drawn map with landmark
references (see Section 3.3.3)
• Camera
• Field form (similar to Attachment E)
• Base station or portable shuttle
• Laptop (if practical) & data back-up device (e.g. flash
• NIST-calibrated field thermometer and/or multi-probe
• Meter ruler or calibrated rod/pole (i.e. surveyor's rod)
• Measuring tape
drive)
meter
• Calibrated replacement sensors and other necessary
deployment equipment
-------
Appendix F
Examples of alternate installation techniques
-------
Appendix F
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 eye bolt (1)
9/16-1 1/16 hose clamp (3)
1/8" wire rope clip (2)
1/8" RL uncoated cable (2' length)
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.
PVC Pipe with
drilled holes
• DIRECTION OF FLOW
000
o o o
0 O 0
i
-
O 0 0 0 0
o o o o o
o o o o o
=
oo oo o o r
o o o o o o •
o o o o o o ~
n
Low profile
concrete base
HOBO-U24 logger
mounted sideways on side of
PVC pipe to minimize
bubbles and sedimentation
on sensor face
Front view
The general protocol is as follows:
• 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.
Makings concrete base for sensor deployment. Phototaken 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.
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. The photo below
shows an old-style base. For the last few years, DRBC has oriented the bars in the other direction and has bent
them flush with the top surface of the concrete base. The PVC pipe is then attached to the bars using large hose
clamps. The new configuration is lower profile and more secure.
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.Yagecic@drbc.state.nj.us) and Eric Wentz
(Eric.Wentz@drbc.state.nj.us) from the DRBC.
-------
Appendix G
Mid-deployment/maintenance check form
(fill out one for each mid-deployment site visit)
-------
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 NIST-calibrated field thermometer.
Field thermometer S/N:
Date
Time
Temperature (°C)
Water
Field
Thermometer
Sensor
Air
Field
Thermometer
Sensor
-------
DEPTH MEASUREMENTS
Measurement
Total stream depth at the sensor:
Distance from the stream bottom up to the sensor:
Distance from water surface to the sensor:
Wetted width along a transect that intersects the
sensor:
Value
Units
(Optional) Cross sectional survey of stream temperature (taken along a transect that intersects the
sensor with a NIST-calibrated field thermometer)
#
1
2
3
4
5
6
7
8
9
10
Distance
from
bank
Temperature
(°Q
Depth (m)
Notes
-------
Appendix H
QA/QC checklist for temperature sensor data
-------
QA/QC CHECK- TEMPERATURE SENSOR DATA
Technician conducting QA/QC check:
Sensor S/N:
Water or air:
Station number:
Waterbody name:
Sensor
Programmed launch/start
Deployment (in water)
Retrieval
Date
Time
Technician
Accuracy checks
Pre-deployment
Mid-deployment
Post-deployment
Pass/fail
Notes
Describe how you addressed flagged 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?
-------
Checklist
Did you check the units to ensure that they were
consistent throughout the period of deployment
(should be °C)?
Were there missing data? If so, in the Notes
field, describe how you addressed these data and
why you think they occurred
Did you remove observations recorded before
and after the sensor was correctly positioned in
the stream channel? (e.g. if it was programmed
to launch before it was deployed in the stream)
Did you flag values if they met the conditions below?
(If you used different thermal limits, write what you
used in the Notes field)
>25°C
<-l°C
> daily change of 10 °C
> upper 5th percentile of the overall
distribution
< lower 5th percentile of the overall
distribution
Did you plot individual data points to look for
abnormalities?
Did you compare the data to other datasets (see
below)?
vs. air temperature
vs. other sites
vs. other years
vs. flow data (if available)
Yes/No
Notes
-------
Appendix I
Equipment checklists for pressure transducer protocols
-------
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 and stadia rod, paint
marker or nails for marking benchmarks)
• Concrete wedge anchor and nuts (stainless
• Hammer or rotary drill
• Concrete drill bit
• Hammer
steel or galvanized)
• Galvanized or stainless steel pole with cap
• Galvanized or stainless steel conduit straps
• Pole driver or sledge hammer
• Pry bar
-------
Equipment list for transducer installation.
Location
All
Fixed object -
additional items
Supplies list
• PVC pipe for transducer (drilled with !/2" 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 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
• Extra-long PVC (should extend out of water in high flows)
• Solar shield (if using barometric logger for air temperature data, see
Attachment 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
• 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
-------
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
-------
Appendix J
Field forms for water level and flow measurements
-------
Elevation/Staff Gauge Survey Sheet
Site: Date:
Crew: Gage (ft):
Notes:
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 HI (+) FS (-) Elevation
Notes and Site Sketch:
-------
Transducer Download Field Data Sheet
Site:
Date: Time:
Crew:
Weather:
Gage (ft.): Transducer (ft):
Photos taken? Gage Survey?
Battery Status: Batteries Changed?
File name:
Notes:
-------
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
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)
[Velocity
(fps)]
[Q (cfs)]
Total Q (cfs):
% of Q
Notes
-------
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
(fps)]
Total Q (cfs):
[Q (cfs)]
% of Q
Notes
------- |