The EnviroDIY team created this manual to help you build, program, install, and manage an EnviroDIY Sensor Station. If you find an error or think something is missing, please send us an email using our feedback form. (If your system does not allow access to the Google Form, please email your comments to firstname.lastname@example.org.)
Version 1.0, 09 November 2018
|EnviroDIY||A community for do-it-yourself environmental science and monitoring. EnviroDIY is part of WikiWatershed®, a web toolkit designed to help citizens, conservation practitioners, municipal decision-makers, researchers, educators, and students advance knowledge and stewardship of fresh water.|
|WikiWatershed||WikiWatershed, an initiative of Stroud™ Water Research Center, is a web toolkit designed to help citizens, conservation practitioners, municipal decision-makers, researchers, educators, and students advance knowledge and stewardship of fresh water.|
|Mayfly Data Logger||User-programmable microprocessor board that is fully compatible with the Arduino IDE software; data logger used in EnviroDIY Sensor Station|
|SL###||EnviroDIY Sensor Station ID (“Logger ID”), Stroud Logger 0-# – Running tally of all EnviroDIY Sensor Stations that have been built and deployed by Stroud Water Research Center. Each individual EnviroDIY Sensor Station has a unique SL number.|
|Site ID||Stroud Center-assigned site identification used for preparing and processing samples and general accounting|
|Time-series data||Data points collected over time at equally spaced intervals. For EnviroDIY Sensor Stations measurements are recorded every five minutes.|
|Continuous data||Term used interchangeably with “time-series data” — refers to the continuous stream of data generated at specific time intervals.|
|Discharge||Quantity of water flowing in a stream over time (m3/s or ft3/s)|
|Turbidity||Measure of light passage through water (i.e., measure of the cloudiness of water). Measured in Nephelometric Turbidity Units (NTU) and others (e.g., JTU, FNU)|
|Conductivity||Electrical conductivity — measure of how well water conducts electricity — directly related to the amount dissolved material (ions) in the water. Measured as microsiemens per centimeter (μS/cm)|
|CTD||Conductivity (μS/cm), Temperature (degC), Depth (mm) — refers to the sensor that measures these parameters|
|Chloride (Cl–)||Anion (negative charge) part of common road salt (NaCl) — commonly measured (mg/L) in streams to assess road salt inputs|
|Total Suspended Solids (TSS)||Concentration (mg/L) of undissolved materials in water, includes sediment (sand, silt), organic debris, and any other particles that do not dissolve in water.|
|Supplemental sampling||Term used here to indicate additional sampling intended to enhance EnviroDIY Sensor Station data — primary intention is for supplemental data to be used for developing rating curves.|
|Rating curve||Graph and correlation of one variable against another on x- and y-axes. Once developed, the relationship is used to infer information about one variable based on the other — here, discharge is inferred based on depth, TSS based on turbidity, and chloride based conductivity.|
|Field Visit Data Form||General data sheet used for recording EnviroDIY Sensor Station management actions, grab sample collection information, and other key sampling information|
|Stream Discharge Data Form||Data sheet used for recording measurements used for calculating discharge|
|Cross-Section Discharge Form||Data sheet used for delineating stream cross-section profile for use in predicting wetted cross-sectional area in unwadeable conditions|
|Discharge Rating Curve Calculator||Excel spreadsheet developed by Stroud Water Research Center into which Stream Discharge Data form data are entered, discharge is calculated, and rating curve generated.|
|Stage to Area Predictor||Excel spreadsheet developed by Stroud Water Research Center that contains channel profile information and which calculates predicted wetted cross-sectional area for use in estimating discharge in unwadeable conditions (used in association with Discharge Rating Curve Calculator)|
|Load Calculator||Excel spreadsheet developed by Stroud Water Research Center that uses rating curve equations and time-series data from EnviroDIY Sensor Station to calculate TSS and chloride loads|
|MonitorMyWatershed.org||Web data portal to which all data from EnviroDIY Sensor Stations with cellular coverage are transmitted|
|Campbell OBS-3+ Turbidity sensor||Turbidity sensor made by Campbell Scientific — measures turbidity in Nephelometric Turbidity Units (NTU)|
|Meter Hydros 21 CTD sensor||Conductivity, Temperature, Depth sensor made by Meter Group (formerly Decagon CTD-10)|
|Hologram||Technology company providing cellular data plans for transmitting data on the internet|
EnviroDIY is a community sharing do-it-yourself ideas for environmental science and monitoring. Our vision is that the sharing of ideas and experiences by the EnviroDIY community will result in open-source hardware and software solutions for observing our environment that are low cost, easy to learn, and easy to use. Our goal is that this resource will create a wealth of high-quality, real-time data that transforms the practice of environmental science, resource monitoring, and watershed protection.
All EnviroDIY members can showcase their environmental-sensing gadgets or describe their own homegrown approaches to monitoring, sensor calibration, installation hardware, radio communication, data management, training or any number of other topics. Members can post and answer questions through an EnviroDIY forum and can network within interest groups to collectively develop new devices, tutorials, or other useful products.
EnviroDIY was inspired by DIY Drones, an online community of about 60,000 amateur designers and builders of unmanned aerial vehicles; Weather Underground, a website that aggregates data from the backyard weather stations installed by citizen scientists; and Public Lab, a collaborative community developing open, citizen-science tools for environmental exploration and investigation.
Stroud Water Research Center developed an open-source data logger that can provide the core functions needed to operate environmental sensors. The Mayfly Data Logger provides the scientific community (researchers, citizens, and educators) a low-cost, open-source hardware solution for connecting with environmental sensors, recording measurements, and relaying data to the internet. This manual provides one example of how to construct a water quality monitoring station using the Mayfly Data Logger. This popular design is comprised of sensors for measuring conductivity, temperature, depth, and turbidity, a logger box (containing the Mayfly Data Logger) and solar panel, and a staff gauge mounted in the stream (Figures 1.1 and 1.2). Instructions are provided below on how build, install, and maintain this EnviroDIY Sensor Station; some familiarity with electronics, computer software, hand tools, and power tools will be helpful in completing this project.
The EnviroDIY Sensor Station is one embodiment of many possible designs for environmental science and monitoring using the Mayfly Data Logger. It will not be appropriate for all water-monitoring purposes because each water-monitoring project has unique goals and special requirements for how the data is used to address the research questions being asked. These goals and data requirements should guide the selection of the sensors, parameters, and the sampling frequency used for monitoring water.
EnviroDIY Mayfly Data Logger
The Mayfly Data Logger (Figure 2.1) is a powerful, user-programmable microprocessor board that is fully compatible with open-source Arduino software. The memory card socket, real-time clock (RTC), and solar charging features make it easy to use the Mayfly Data Logger for recording data and low-power operation. The Mayfly can be powered by a 3.7 V lithium battery or through the microUSB port. Programming can be done through the microUSB port or via the FTDI header.
The Mayfly Data Logger features:
- Atmel AVR ATmega 1284p processor
- 128K Flash memory, 16K RAM
- 28 digital I/O pins, eight analog pins, plus
- Two hardware serial UART ports
- Maxim DS3231 I2C connected high precision Real-Time Clock (RTC)
- Independent CR1220 battery socket to power RTC
- Texas Instruments ADS1115 I2C connected 16-bit Analog-Digital Converter
- Adds four additional high-resolution ADC pins
- MicroSD memory card socket
- Solar LiPo battery charging
- Low power consumption (6.5 mA when on but idle, 0.27-0.43 mA when in sleep)
- XBee module socket
- Two user-programmable LEDs, one user-programmable push button
- 3.3 V main board voltage, additional 5 V boost circuitry for external devices
- Two 20-pin headers for accessing all available I/O pins
- Six Grove-style sockets for easy connections to sensors and devices
The Mayfly Data Logger runs a program defined and uploaded by the user. This program (or “sketch”) is generated by the user in an integrated development environment (IDE) with a graphical user interface (GUI). While other options are available, this manual provides instructions for the use of the Arduino IDE. Arduino can be installed on Windows, IOS, and Linux operating systems. Arduino is an open-source project supported by a community of users and developers. This manual provides instructions on using Arduino to program the Mayfly Data Logger, but users are encouraged to learn more about Arduino by starting with these links:
In addition to the Arduino software that runs on your computer and programs the Mayfly Data Logger, Arduino also offers a wide range of electronic hardware. The Mayfly Data Logger is a customized printed circuit board and microprocessor that are similar to Arduino hardware products. While some Arduino hardware can be used to perform the same functions as the Mayfly Data Logger, Stroud Water Research Center created the Mayfly Data Logger to be a simple, inexpensive, and extendable alternative specifically designed for connecting and controlling environmental sensors. Nevertheless, the Mayfly Data Logger is compatible with the easy-to-use and community-supported Arduino IDE.
The programs that run a Mayfly Data Logger are diverse. Each combination of sensors, measurements, and data management requires a unique program. However, there are many programs written by Stroud Water Research Center and other Mayfly Data Logger users that can be copied, pasted into an Arduino sketch, and uploaded to the Mayfly Data Logger with no (or minimal) editing required. These programs can be found at github.com/envirodiy, and are ideal for new users with little experience who want to get a Mayfly Data Logger up and running quickly and for users who want to share their programs and associated files such as helpme documents and instructions for use. EnviroDIY relies on GitHub as a repository for programming files and libraries, instructional documents, and further instructions.
A wide variety of environmental sensors can be connected to and controlled by the Mayfly Data Logger: sensors for atmospheric, soil matrix, groundwater, surface water, and other environments are available from commercial suppliers. Stroud Water Research Center does not develop sensors, only the Mayfly Data Logger to which sensors are connected. There are two main requirements for a sensor to be compatible with the Mayfly Data Logger:
- A known physical connection mechanism (i.e., labeled wires or a documented plug) and
- a known communication protocol.
This manual instructs users on connecting and controlling two sensors:
- A combined conductivity, temperature, and depth sensor (CTD; available from Metergroup), and
- a turbidity sensor (available from Campbell Scientific) (Figures 2.2, 2.3, and 2.4).
Both sensors are factory calibrated and should not need recalibration. However, for quality-control purposes it is recommended that calibrated handheld sensors (e.g., LaMotte PockeTester 1749) be used to cross-check the Hydros 21 CTD and Campbell OBS-3+ (see Sensor Station Management: Quality Control).
Documentation and instructions for connecting a wide variety of sensors are available at https://github.com/EnviroDIY/ModularSensors. The www.envirodiy.org/forums/ and www.envirodiy.org/blogs/ are also useful resources for learning how other users have connected and controlled a wide variety of sensors.
EnviroDIY Sensor Station
This manual provides instructions for making, deploying, and maintaining an EnviroDIY Sensor Station that measures water temperature, conductivity, depth, and turbidity. This combination of sensors is one example of many different configurations that are possible for measuring surface water quality. In this example (and in many configurations of the Mayfly Data Logger and sensors), a deployed EnviroDIY Sensor Station will consist of a waterproof box containing:
- a Mayfly Data Logger
- a battery
- a cell phone modem and antenna
- a solar panel to charge the battery
- a series of sensors connected to the Mayfly Data Logger inside the waterproof box (Figure 1.2)
Preparing the Mayfly Data Logger
- Download and install FTDI virtual COM port (VCP) drivers on your computer from FTDI-VCP.
- NOTE: Windows users must have administrative privileges to be able to install these drivers.
- Use the free Arduino IDE programming software to program the Mayfly Data Logger. Download the latest version from Arduino.cc at https://www.arduino.cc/en/Main/Software. If you already have the IDE installed on your computer, ensure that you are using version 1.6.5 or newer.
- NOTE: Windows users may not be able to install this software without administrative privileges.
- Once the Arduino IDE is installed, you will need to add the Mayfly board to the list of available boards. To do this, start the Arduino software, then click on the “File” menu, then “Preferences” and paste the following URL into the box labeled “Additional Boards Manager URLs”:
- In the IDE, click on Tools > Board > Boards Manager. Use the dropdown box to select “Contributed.” You should then see an option for “EnviroDIY ATmega Boards.” Click the Install button to add the EnviroDIY boards to your IDE.
- Now when you click Tools > Board you will see the EnviroDIY Mayfly 1284P listed either at the top or bottom of your list of available boards.
Connecting a Computer to the Mayfly Data Logger
- Plug the small black microUSB cord into the Mayfly Data Logger, then plug the USB end of the cord into a free USB port on your computer (Figure 2.1).
- Turn the power switch to the on position on the Mayfly Data Logger; red and green LEDs will blink (Figure 2.1). If this is the first time your board has been connected to your computer, wait while the computer recognizes the board and assigns it a COM port number.
- Open the Arduino IDE. Select the COM port assigned to your Mayfly by selecting Tools > Port > port XXX.
- NOTE: It is highly unlikely that your board will be assigned as COM1. This port is typically reserved as a hardware RS232 port, even though very few computers have those ports available. If the only port you see in the port menu is COM1, verify that your Mayfly is correctly plugged in to a USB port, the power switch on the Mayfly is on, that you have correctly installed the FTDI VCP drivers, and that the computer has had enough time to recognize your board. You may also have to close and reopen the Arduino IDE.
- In the Arduino IDE, open the serial monitor (button in upper right corner of Arduino window or via Tools > Serial Monitor). At the lower right of serial monitor window set the baud rate to 57600. All of the Mayfly boards have been preprogrammed with a sample sketch. This sketch instructs the DS3231 real-time clock on the Mayfly to measure temperature and report it through the serial monitor.
Preparing sketches for programming the Mayfly Data Logger is simplified by the use of libraries: packages of functions that serve as shortcuts when writing sketches. A set of many helpful libraries has been compiled at github.com/envirodiy/libraries. Not all of these libraries are required, but many of them will be needed for programming an EnviroDIY Sensor Station to send data to the online data portal (MonitorMyWatershed.org). To download all of the libraries together:
- Go to github.com/envirodiy/libraries.
- Scroll down to the ReadMe.
- Follow the instructions for installing libraries in the Arduino IDE.
Creating and Uploading Sketches
The Mayfly Data Logger must be programmed with a “sketch” telling it what to do before it will work as a data logger. The Arduino IDE allows you to compose sketches and then compiles them and sends them to the data logger board when complete. There are dozens of tutorials online that demonstrate how to use the IDE, often using a “blink” or similar sketch. Arduino’s short guide to their IDE is available at https://www.arduino.cc/en/Guide/Environment.
The first step in preparing the Mayfly Data Logger for use is to set its DS3231 real-time clock. Doing this requires installation of the DS3231 library, either alone or with the other libraries as described above (see Preparing the Mayfly Data Logger: Installing Libraries). Before setting the clock, slide a CR1220 battery (available at most grocery stores and hardware stores) into the RTC battery holder (marked L on the Mayfly diagram, Figure 2.1). The “+” side of the battery goes up. (Usually the lettering is on the “+” side.) This battery will only power the clock and can allow the clock to keep time for several years, even if the rest of the board is unpowered.
The sketch for setting the clock is one of the examples in the DS3231 library found at https://github.com/EnviroDIY/Sodaq_DS3231/tree/master/examples/PCsync. Follow the steps in the ReadMe section to synchronize the clock automatically with the current Network Time Protocol (NTP, https://en.wikipedia.org/wiki/Network_Time_Protocol) time.
Building an EnviroDIY Sensor Station
Assembling the Mayfly Data Logger
The electronic components of the EnviroDIY Sensor Station are shown in Table 4.1. Plug the vertical microSD card adapter board provided in the Mayfly Data Logger Starter Kit into the SPI jack on the Mayfly Data Logger (near the main power switch, Figure 2.2). Insert the cellular SIM card into the slot on the cellular radio module (GPRSbee, 3Gbee, etc, Figure 2.2). Plug the cellular module into the Mayfly Data Logger’s Bee port, making sure to get the pins lined up correctly. Connect the small JST power jumper cable between the cell module and one of the jacks labeled “LiPo Batt” on the Mayfly Data Logger (Figure 2.2). Solder the 2 mm JST connector with leads to the 18” solar panel power cable extension. Be sure to use small heat-shrink tubing on the two solder joints.
|Table 4.1. Electronic components of the EnviroDIY Sensor Station.|
|Component||Vendor Link||Unit Cost (2018)||Quantity||Description|
|EnviroDIY Mayfly Data Logger Starter Kit||Amazon||$90.00||1||Includes waterproof enclosure with clear lid, 0.5 W solar panel, custom microSD connector board that plugs into Mayfly for easy access to the memory card, 4GB microSD card and adapter, one-meter microUSB cable, and two Grove cables.|
|CR1220 12 mm Diameter – 3 V Lithium Coin Cell Battery||Adafruit||$0.95||1||Lithium batteries for the Mayfly board so they’ll retain the clock time after programming|
|Lithium Ion Battery – 3.7 V 2000 mAh w/ pre-attached two-pin JST-PH connector||Adafruit||$12.50||1||Basic battery for sunny location (option for bigger solar panel and longer mAh battery for shady locations)|
|Option: Lithium Ion Battery Pack – 3.7 V 4400 mAh||Adafruit||$19.95||Big battery: This lithium ion pack is made of two balanced 2200 mAh cells for a total of 4400 mA capacity|
|GPRSbee rev.6 cell wireless module (2G cell module)||SODAQ||$35.00||1||A GPRS/GSM expansion board with the “bee” form factor and can be used in any system that has a bee socket like the SODAQ boards, Seeeduino Stalker or the Arduino Fio. The GPRSbee uses SIM cards of the MicroSIM form factor. The core of this board is powered by the SIM800H module. This module, like most other GPRS/GSM modules, has an operating voltage of 3.5 – 4.5 V and can draw up to 2 A of power during broadcasts bursts.|
|Hologram Global SIM Card||Hologram||$5.00||1||SIM card required for 2G or 3G communication|
|6 V 2 W Solar Panel||Adafruit||$29.00||1||Standard for CTD sensor install|
|Option: 6 V 3.5 W Solar panel||Adafruit||$39.00||Option for larger solar panel if in shady location or addition power load is added to station|
|JST two-pin cable (for solar panel; requires soldering)||Adafruit||$0.75||6||Required for connecting solar panel to EnviroDIY Mayfly Logger Board (soldering required)|
|Power cable for solar panel (solder to JST 2-pin cable)||Voltaic Systems||$4.00||1||Extension with Exposed Leads
Connect a Voltaic solar panel to a breadboard or other custom connector. Female 3.5 x 1.1 mm connection on one end, tinned positive and negative cables on the other. Cable is 21” (53 cm).
Pelican Case Assembly
The logger box contains the “guts” of the EnviroDIY Sensor Station (Figure 4.1). Drill three large holes in the side of the Pelican case for the cable glands for the sensors and solar power cable. Drill two small holes in the back of the case for the mending plate mounting bracket. Install the cable glands on the case, being sure to use the small O-rings that come with the cable glands. Tighten the cable glands from the outside using an adjustable wrench, but not too tight that the threads get stripped. Use a 15/16” and 1-1/16” socket to tighten the nuts on the cable glands inside of the case. Then install the bracket on the back of the Pelican case. Cut holes in the Pelican case foam using a small X-Acto knife.
Insert the JST connector on the solar panel cable into the small cable gland and route it up towards the hole where the Mayfly box will be placed in the foam. Once the JST connector is inserted into the cable gland, it usually can’t be pulled back out without breaking, so use great care if you need to remove it for some reason.
Put the small Hammond enclosure into the large hole in the middle of the foam. Most people don’t use the Hammond enclosure lid when being used in the Pelican box. Put the Mayfly Data Logger in the small enclosure and connect the solar panel cable to the jack labeled “SOLAR” on the Mayfly Data Logger. Put the LiPo battery in the battery slot in the foam and connect it to either of the two jacks labeled “LiPo Batt” on the Mayfly Data Logger. See Appendix A for discussion of battery options and solar charging capability.
Contents of the EnviroDIY Logger Box
Figure 4.1 A. Pelican 1120 case with foam. Pelican cases are plastic containers that seal with an airtight and watertight gasket. There is a barometric relief valve made of Gore-Tex to prevent pressure damage to the case, during transportation or when the air pressure in the environment changes. This specific Pelican case costs $24.10. The foam inside allows the Mayfly Data Logger and the battery to fit nicely without moving around.
Figure 4.1 B. Lithium ion polymer battery. 3.7 V 4400 mAh lithium ion polymer battery from adafruit.com (batteries also come in different sizes; for more information on different size batteries please see the post-installation manual or look it up on adafruit.com).
Figure 4.1 C. Mayfly Data Logger. Created by electrical engineer, Shannon Hicks. The Mayfly Data Logger is a powerful microprocessor that functions using Arduino programming language and can be programmed to accommodate any number of different environmental sensors.
Figure 4.1 D. Cable glands (x3).The two large cable glands (3/4” NPT) are for the CTD and OBS-3+ cables. The small cable gland (½” NPT) is for solar panel cable. Cable glands allow wires to enter the logger box while at the same time remaining waterproof and submersible.
Figure 4.1 E. Solar Panel. Shown here is the 2.0 W solar panel by adafruit.com. The average solar panel used for an EnviroDIY Sensor Station is 3.5 W, these are sufficient to power the stations in most scenarios. These outdoor solar panels are also available in larger sizes if necessary (see post-installation manual for more information on the various solar panel sizes).
Figure 4.1 F. CTD sensor. Meter Hydros 21 CTD Sensor Electrical Conductivity Temperature Depth Made by Meter Group. Measures electrical conductivity, temperature, and depth. This CTD sensor uses a vented differential pressure transducer to measure the pressure from the water column to determine water depth at the location of the sensor, not from the stream bed. The sensor also automatically corrects for changing barometric pressure. Temperature is measured by a specific type of resistor, called a thermistor, in thermal contact with the probe. Conductivity is measured via four metal screws on a red board within the CTD, which form a four-electrode array. The Hydros 21 CTD sensor is made by METER Group Inc.
Figure 4.1 G. OBS-3+ Turbidity Sensor. A submersible, optical backscatter turbidity probe that has sideways-facing optics. Made by Campbell Scientific.
Mounting Bracket and Associated Hardware
Figure 4.2 A. Mending Plate. Available through Lowe’s Home Improvement, these brackets are the perfect size to fit onto the logger box without an excessive amount of extra metal sticking out from the top, or below the logger box. The mending plate is crucial to create a stable attachment location on the logger box. To allow attachment, two ¼” holes need to be drilled into the logger box at exact locations that match the holes in the mending plate.
Figure 4.2 B. Stainless Steel with Neoprene Rubber Sealing Washer. Available through McMaster-Carr (www.mcmaster.com: Item number: 94709A317) Because the logger box was drilled to allow attachment to the mending plate, it is no longer waterproof. To ensure that the logger box is waterproof again, rubber sealing washers are necessary to fill the gaps between the mending plate and the logger box. Without the neoprene on the underside of the washer, water would find its way through the gaps, and into the box potentially damaging the contents inside.
Figure 4.2 C. Stainless Steel Phillips Flat Undercut Head Screw. Available through McMaster-Carr (www.mcmaster.com: Item number:91099A453) This particular screw fits flush with the suggested mending plate. The Phillips flat undercut head screw goes through the top of the mending plate, and then through the logger box itself to allow the plate to be securely fastened to the logger box.
Figure 4.2 D. Stainless Steel Thin Nylon-Insert Lock Nut. Available through McMaster-Carr (www.mcmaster.com: Item number:90101A230) The lock nut goes onto the threaded end of the screw on the inside of the logger box. The nylon within the nut ensures that no water is to enter the box, even if water somehow gets past the rubber washer, this is the last line of defense before water will enter the case.
Connecting Sensors to the Mayfly Data Logger
The Campbell OBS-3+ sensor cable has six colored wires that need to be trimmed. Cut approximately three inches off the wires and then strip and tin the wires so that 5 mm of wire is exposed and tinned. Insert the OB3+ cable into the center cable gland in the Pelican case, being sure to insert it past the heat-shrink neck on the cable. Tighten the cable gland securely around the cable. Insert the six wires into the EnviroDIY six-port screw terminal-to-Grove adapter board. Ensure that the correct colors are going to the right terminal on the adapter board. Connect a 20 cm Grove cable to the adapter board and plug the other end into the AA0-AA1 Grove socket on the Mayfly Data Logger. Make sure the small jumper next to the AuxADC ports is moved from its default 3.3 V position to the 5 V position.
The CTD sensor cable ends with a 3.5 mm stereo jack. That stereo jack should be plugged into the EnviroDIY Grove-to-stereo adapter board. Connect a 20 cm Grove cable to the adapter board and plug the other end into the D6-7 Grove socket on the Mayfly board. The CTD also requires a programming step before it is ready to log data: the default SDI-12 address for the CTD sensor is “0”. This needs to be changed to “1” in order for the Mayfly Data Logger to correctly communicate with the sensor. Use the “SDI-12 Change Address” sketch from the SDI-12 Arduino library to change the sensor address from 0 to 1.
Mounting Parts and Solar Panel Attachment
Mounting parts (Table 4.2) include all items that will keep the EnviroDIY Sensor Station in a fixed position on the stream bank. To ensure that the EnviroDIY Sensor Station properly functions long term in outdoor conditions, the logger box is typically mounted on a tall pipe to keep it out of harm’s way during high flow conditions. The length of the pipe can vary based on the specific features of a site (Figures 6.3 and 6.4). Mounting materials that will be installed on land include the galvanized pipe, ¾” coupling, small hose clamps, solar panel bracket, and U-bolt. In addition to the mounting bracket, two small stainless steel hose clamps (range: 13/16” to 1 ½”), are needed to attach the logger box to the galvanized pipe (Figures 6.5 and 6.6).
The location of the sensor station depicts what size solar panel will be used. If the station is located in a open sunny field, a 2 W solar panel will be sufficient. If the station is located in an area with moderate vegetation a 3.5 W solar panel should be used (this is the solar panel size most commonly used and will be sufficient in most cases). If the sensor station is located in extremely dense vegetation for most of the year (i.e abundant evergreen vegetation) a larger 6 W solar panel is available. The solar panels from Voltaic Systems are waterproof, scratchproof, and rugged enough to withstand long-term outdoor deployment. The three different sized solar panels can all be attached to the galvanized pipe using the same solar panel bracket also available through Voltaic Systems. The solar panel is attached to the bracket using two of the four plastic knurled nuts on the back of the solar panel, which can be tightened and secured using a flat head screwdriver (Figure 6.8 shows the most common type of attachment). Multiple slots in the bracket allow for various types of attachment, to accommodate for varying sun angles. Once the solar panel is attached to the bracket, a stainless steel U-bolt is then used to attach the bracket to the galvanized pipe above the logger box (Figure 6.9). The nuts associated with the U-bolt can be tightened using a 7/16” size wrench. Be sure not to over tighten the U-bolt because the flat plate on the U-bolt is made from aluminum and will bend if tightened too much.
|Table 4.2. Mounting parts for EnviroDIY Sensor Station.|
|Item name||Vendor||Item number||Price per unit||Description|
|¾” coupling||Lowe’s Home Improvement||22481- (black)
|$2.33||Used to secure the two different sized galvanized pipes together. Available in black or silver to match color of pipes.|
|Galvanized pipe (also available in black)||Lowe’s Home Improvement||¾” x 36” galvanized pipe – 24012
¾” x 60” galvanized pipe- 24072
|A variety of sizes are available for specific station needs. Provided here, are the two sizes most commonly used. These pipes will be installed on the stream bank.|
|Small hose clamps (x2)||Lowe’s Home Improvement||910974||$1.78||Used to mount the EnviroDIY Data Logger box to the galvanized pipe|
|Universal Solar panel bracket||Voltaic Systems||bracket||$9.00||Compatible with a 2, 3.5, and 9 W solar panel. Used to mount the solar panel to the galvanized pipe|
|Stainless steel U-Bolt with mounting plate||McMaster- Carr||8896T94||$4.16||Used to mount the solar panel and bracket to the galvanized pipe.|
|Black rebar (with predrilled holes)||Lowe’s Home Improvement||44086||$5.98||The outside diameter of this rebar is the perfect size to fit inside of the PVC pipe. This will be hammered directly into the stream bed to mount the sensor bundle.|
Figure 6.8 A. Voltaic Systems universal solar panel bracket. Attach a Voltaic solar panel to any vertical or horizontal pole or vertical pipe. Rugged 3 mm aluminum is built to withstand high winds and rough conditions. The bracket is compatible with our 2, 3.5, 6 and 9 W solar panels.
Figure 6.8 B. Waterproof solar panel wire. This solar panel wire will not plug directly into the Mayfly Data Logger. There is also a 4’ extension wire available through Voltaic Systems for $6 (https://www.voltaicsystems.com/3511-ext-4ft) if the solar panel needs to be mounted higher up on the galvanized pipe. This is the case in most scenarios.
Figure 6.8 C. 3.5 W Voltaic Systems solar panel. The 3.5 W 6 V solar panel is lightweight, waterproof, and designed for long-term outdoor use in any environment. The panel uses high-efficiency monocrystalline solar cells, and is UV- and scratch-resistant. (Link: https://www.voltaicsystems.com/3-5-watt-panel).
Figure 6.8 D. Location to attach solar panel to bracket. These four knobs can be unscrewed using your fingers or a small flat head screwdriver. Simply unscrew, place the threads through the openings on the solar panel bracket and tighten.
Sensor Bundle Parts
The sensor bundle is comprised of the two different sensors (CTD and turbidity), a drilled PVC pipe, black rebar with holes and a spiked end, a large (2 ½”) stainless steel hose clamp, a retaining pin, and black outdoor use zip ties (Table 6.2, Figure 6.10). These materials will be anchored into the bottom of the stream bed once the black rebar is hammered into the substrate. The rebar must be anchored into the stream bed deep enough so that it will not become dislodged under high flow conditions.
Using a power drill and two different sized drill bits (5/32” and 3/16”), make a hole 1 ½” from the top of the PVC pipe that lines up with the hole pre-drilled in the black rebar. The retaining pin is used to secure the PVC pipe to the black rebar (Figures 6.10 and 6.11).
A large stainless steel hose clamp is used to attach the two sensors to the PVC pipe. PVC pipe is purchased in 5’ lengths and cut to a specific length using pipe cutters in the field (Figure 6.12 shows both pieces of PVC pipe cut to proper length). Once the black rebar is hammered into the stream bed, the water level at the top of the rebar is used to measure the length to cut on the PVC pipe measuring by where the water level matches the lettering on the PVC pipe (Figure 6.13). Use black, outdoor use zip ties to secure the sensor wires to the PVC pipe, so that debris will not put tension on the sensors.
|Table 6.2. Sensor bundle parts list.|
|Item name||Vendor||Item number||Price per unit||Description|
|Stainless steel hose clamp, 2 ½” diameter (pack of 5)||McMaster-Carr||5011T19||$7.54||Pack of five. Only one is used to secure the two sensors onto the PVC pipe|
|¾” x 5’ PVC pipe||Lowe’s Home Improvement||32972||$2.30||The PVC pipe is used to fit over the black rebar with holes to anchor the two sensors into the stream bed.|
|Stainless steel Retaining pin (5-32” diameter)||McMaster-Carr||90026A110||$2.69||This locking pin is the perfect length and diameter to|
|Hydros 21 CTD sensor by METER Group Inc.||METER Group Inc.||Hydros 21||$475.00||Measures electrical Conductivity, Temperature, and Depth. The Hydros 21 CTD sensor is made by METER Group Inc.|
|OBS-3+ Turbidity Sensor (with cable)||Campbell Scientific||OBS-3+||$1,311||A submersible, optical backscatter turbidity probe that has sideways-facing optics.|
Figure 6.13 A. Steel rebar pin (round steel stake with pre-drilled holes). Black steel stake with a pointed end, available in either 2’ or 3’ lengths at Lowe’s Home Improvement.
Figure 6.13 B. Grey PVC conduit with drilled hole. Plastic electrical conduit, ¾” grade available in 5’ lengths at Lowe’s Home Improvement. The PVC is cut in the field, to a specific length, to fit over the black rebar in the stream bed. A hole must be drilled either 1 ½” (if using 2’ black rebar) or 3¼” (if using a 3’ black rebar) from the top of the PVC to match exactly to the first hole in the black rebar.
Figure 6.13 C. Outdoor black zip-tie to secure the sensor wires. Note: using regular white zip ties in outdoor conditions will not be sufficient as they tend to break easily when subjected to cold/heat/sun.
Figure 6.13 D. Retaining pins. Stainless steel bent wire locking pin available through McMaster-Carr. This is the mechanism that secures the PVC pipe to the black steel rebar at a fixed depth.
Figure 6.13 E. Hydros 21 CTD sensor by METER Group Inc. Electrical Conductivity Temperature Depth sensor.
Figure 6.13 F. Campbell OBS-3+ sensor by Campbell Scientific. Optical backscatter Turbidity sensor uses its sideways-facing optics to emit a near-infrared light into the water. It then measures the light that bounces back from the water’s suspended particles.
Figure 6.13 G. Large hose clamp. Hose clamp with 410 Stainless Steel Screw, ½” Wide Band. available through McMaster-Carr.
A staff gauge (Figure 6.15) is installed with the sensor station to serve as a semi-permanent on-site depth reference that is used to develop hydrologic rating curves (see Supplemental Sampling, Rating Curves, Loads). The staff gauge is a “master reference” for depth over the lifetime of the sensor station deployment. The staff-sensor depth offset is the difference between water level measured on the staff gauge and water depth measured by the CTD sensor. Once installed, any subsequent adjustments to positioning of the staff gauge will require adjustments to the rating curve (if developed).
Many different staff gauge models are available with different units and measurement increments. Staff gauges installed in association with the EnviroDIY sensor stations are in metric units (cm) in half-centimeter increments (Figure 6.15).
Programming and Activating an EnviroDIY Sensor Station
Uploading the Mayfly Sensor Station Sketch
This manual provides instruction for building, deploying, and maintaining an EnviroDIY Sensor Station with CTD and turbidity sensors. The sketches described in this section enable the Mayfly Data Logger to control a Meter Hydros 21 CTD sensor which measures conductivity (μS/cm), temperature (degrees C), and depth (mm), and a Campbell OBS-3+ Turbidity Sensor that measures turbidity (NTU, Nephelometric Turbidity Units). The Mayfly Data Logger also has an internal air temperature sensor (within the logger box) and records battery level; see Appendix B: Example Data). The Mayfly Data Logger is programmed to record data every five minutes to the microSD card. Additionally, the EnviroDIY Sensor Station can report data to MonitorMyWatershed.org if cellular service is available at the deployment location. The standard method for transmitting data to the portals is via 2G cell signal (3G and LTE options are under development).
The sketches needed to run this particular combination of sensors is available on github.com/envirodiy. For EnviroDIY Sensor Stations which will not have 2G cellular service, copy and paste the contents of the ino file DRWI_NoCellular.ino into a new (blank) sketch. More details on this program are available at github.com/EnviroDIY/ModularSensors/tree/master/examples/DRWI_NoCellular. If, instead, the user intends to connect the EnviroDIY Sensor Station to the internet via 2G cellular signal and have data automatically reported to MonitorMyWatershed.org, copy and paste the contents of the ino file DRWI_CitSci.ino into a new (blank) sketch. Edit the sketch with the logger ID, calibration coefficients, Universally Universe Identifiers, and the Sampling Feature as described at github.com/EnviroDIY/ModularSensors/tree/master/examples/DRWI_CitSci. Instructions for obtaining the Registration Token, Universally Universe Identifier, and Sampling Feature that describe a site and associated sensors are provided below in the below (see Registering a Station on MonitorMyWatershed.org).
After uploading the program to the Mayfly Data Logger, you should use the serial port monitor to ensure that the Mayfly Data Logger is correctly detecting the sensors and making an internet connection (if enabled for cellular connection). The initial “setup” portion of the program after the board is powered or reset takes up to two minutes. Once the setup is finished, you can press the black D21 button (labeled D, Figure 2.2) at any time to enter a sensor testing mode where the current sensor values will be printed to the serial port monitor. When deploying the station in the field, you can use an “OTG” microUSB cable to attach an android smartphone with an app like DroidTerm to the board to use the phone as a serial port monitor. If you do not have the cables/phone in the field you can also look for the following pattern of blinking lights on the Mayfly board itself:
- When the board is first switched on (or you hit the reset button) the green LED by the “Bee” socket (labeled Q in Figure 2.2) will blink several times.
- Immediately after the green LED blinks, the green and red LED’s by the “Bee” socket will blink back and forth quickly.
- The LED on the GPRSBee should then light up. It will blink slowly as it attempts to register to the network and then very quickly once an internet connection is established. After connecting to the internet, it may resume a very slow blink pattern or immediately turn off. This first attempt to connect to the internet may take up to two minutes. If the blink pattern on the GPRSBee never changes from slow to fast, you may not have any cellular internet connection available at your site.
Registering a Station on MonitorMyWatershed.org
Users may want to share their data publicly via the Data Sharing Portal at MonitorMyWatershed.org. In addition to the benefits of sharing data publicly, Monitor My Watershed also allows users to visualize their data using the Time Series Analyst. Time Series Analyst is an interactive graphing program that allows comparison of sites and sensor data. In order for the data from an EnviroDIY Sensor Station to be displayed online on Monitor My Watershed, the user must sign up and register a site. After a site has been established, data from a EnviroDIY Sensor Station can be provided to the website automatically with a cellular connection or manually uploaded. In either case, the first step is to register a site by following these steps:
- Go to MonitorMyWatershed.org
- Click “Log In” in the upper right-hand corner or “Sign Up” and follow the instructions.
- Enter first and last names. Please note that your first and last name will be displayed publicly with your data; if you do not want your name to be visible, please do not sign up for an account.
- Enter email address.
- Choose a username.
- Choose a password.
- Enter your affiliated organization, creating a new organization if the group you belong to does not already exist in the drop-down list (optional).
- Click “REGISTER USER.”
- After registering, click “Log In” and use the information just created in the sign-up section as your new login information.
- Click on “My Sites” next to the Monitor My Watershed logo at the top of the page.
- Scroll down to the blue box and click “REGISTER A NEW SITE.”
- Enter Site Code: a brief and unique text string to identify your site (e.g., “Del_Phil”).
- Enter Site Name: a brief but descriptive name for your site (e.g., “Delaware River near Phillipsburg”).
- Enter Site Type: select the type of site you are deploying from the drop-down menu (e.g, choose from site types such as: Atmosphere, House, Soil hole, Stream, Weather station, Other, and more).
- Enter the following optional information: stream name, major watershed, sub basin, closest town, notes, elevation, and elevation datum.
- Coordinates: Enter the latitude and longitude of your site location in decimal degrees (e.g., 40.02581, -75.82938). If you do not know the latitude or longitude of your site, you can zoom in on the map and select the location of your site on the map. Doing this will automatically populate the latitude, longitude and elevation fields.
- Check the box that says “Notify me if site stops receiving sensor data” if appropriate.
- Confirm the information you provided by clicking “REGISTER SITE”; a new page displays the details of your site.
- Scroll down to “Sensor Observations at This Site” and click “MANAGE SENSORS”.
- Click the blue “+” button to Add New Sensor.
- Select the appropriate drop-down for the sensor manufacturer, model, measured variable, units, and sampled medium.
- Optionally, add the height above or below the water surface in meters and notes about the sensor.
- If you do not see your sensor in the list of equipment models send an email to email@example.com with a request to add a new sensor. Please provide in this email the manufacturer and make/model information for the sensor, as well as a link to a website description if available.
- Click “ADD NEW SENSOR” after entering all of the information.
- Repeat these steps for each additional sensor being used at the site.
- Click the blue “+” button to Add New Sensor.
Once a site and its associated sensors has been generated at Monitor My Watershed, the user has two options for reporting data to that site.
- Cellular-enabled EnviroDIY Sensor Stations will automatically report data to Monitor My Watershed if configured with the appropriate tokens and UUIDs.
- On the Site’s page, click “VIEW TOKEN UUID LIST” at the top right corner.
- Copy and paste the registration token, sampling feature, and sensor UUIDs into the appropriate sections of the Mayfly Data Logger sketch as discussed at github.com/EnviroDIY/ModularSensors/tree/master/examples/DRWI_CitSci.
- EnviroDIY Sensor Stations that are not connected via cellular network will not automatically report data to Monitor My Watershed, however, the user can manually upload data for the site using the following procedure.
- After data has been logged to the microSD card, remove the card from the Mayfly Data Logger, use an SD card adapter to insert the microSD card into a computer.
- Save the data file on the microSD card as a comma-delimited file to your computer and open it as a spreadsheet.
- On the site’s page at Monitor My Watershed, click “VIEW TOKEN UUID LIST” at the top right corner.
- Using the Sampling Feature and sensor UUIDs from this list, edit the .csv file as follows:
Sampling Feature: [sampling feature UUID] [sensor 1 UUID] [sensor 2 UUID] … [yyyy-mm-dd hh:mm:ss] [data] [data] … … …
- Replace [sampling feature UUID] with the sampling feature retrieved from the Token and UUID List for your Site.
- Replace [sensor 1 UUID] (and subsequent sensors) with the UUID for that sensor.
- Replace [yyyy-mm-dd hh:mm:ss] with the date and time of each sample measurement in the format yyyy-mm-dd hh:mm:ss.
- Ensure that the [data] appears in the column for the correct sensor.
- Save the file as a comma-delimited file.
- Click “MANAGE SENSORS” in the “Sensor Observations at This Site” section.
- Click the paper clip icon to upload the comma-delimited file of site data.
- Click “BACK TO SITE DETAILS”; the data from the comma-delimited file should now be plotted as a sparkline plot on the site page.
Activating Cellular Service
If 2G (T-Mobile) cellular coverage is available at the location where the EnviroDIY Sensor Station is deployed, the Mayfly Data Logger will automatically send data via cellular signal to the Monitor My Watershed website. The cellular data plan will need to be activated and paid for by the owner of the sensor station. The cellular plans that are used for the EnviroDIY sensor stations are through Hologram (https://hologram.io/). It is recommended that cell plans be paid for on a yearly basis. If a monthly payment plan is used it requires regular payments and if these aren’t made within several days of the due date data transmission will be interrupted until payment is made. In order for the cell phone modules inside the Mayfly Data Loggers to access the internet, there must be an activated data plan for each of them. Activating the cellular network plan prior to installation is important to assess the functioning of the sensor station once installed. If the cell plan is activated, once the Mayfly Data Logger is turned on, data will start to transmit to Monitor My Watershed immediately on every five-minute interval. The company we use for hosting the data plan is called Hologram. Sign up for an account and activate the cellular network sim card at https://hologram.io.
Installing an EnviroDIY Sensor Station
The monitoring objectives and research goals should be the main consideration when preparing for the installation of an EnviroDIY Sensor Station. The parameters to be measured, frequency and timing of measurements, sensor station location, and placement of sensors within the stream all affect the data that will be collected. That data must appropriately characterize the environmental condition that is necessary to successfully answer a research question or achieve a monitoring goal.
Equipment needed for installing the sensor station and sensor bundle includes:
- 2 x 4 wood block (to hammer on)
- Wire cutters
- 7/16” wrench
- Power drill
- 5/32” drill bit
- 3/16” drill bit
- Two pipe wrenches
- PVC pipe cutter
- Black rebar spray painted orange: used to hammer a pilot hole in the stream bed and on the stream bank, avoid damage to the rebar and pipe.
- Tent stakes
- Extension rod for hammering the in-stream rebar in deep water where the rebar goes underneath the water (black rebar with PVC pipe attached).
- Nut driver (or screwdriver) that accommodates hose clamps
Choosing an Installation Location
Cellular Signal: Cellular signal can be variable and is subject to change during storm events and seasonal changes. Factors such as cloud cover and vegetation cover can affect the cellular signal in some cases. Cellular signal is also subject to change due to issues with the cellular network itself (i.e. temporary cellular network outages). Stroud Water Research Center uses a device, created by electrical engineer Shannon Hicks, to detect the strength of the cellular signal.
In the absence of signal detector, take your station out to the proposed installation location, turn it on, and see if data are easily transmitted to Monitor My Watershed on regular five-minute intervals before installing the station. If data do not appear in the online portal, search for a different location. A disadvantage of this signal testing method is that the test data collected while testing for cellular signal will be sent to Monitor My Watershed and and cannot be deleted. Test data points should be documented in an official Field Visit Data Form so they are not confused with subsequent stream measurements.
If there is not enough cellular signal in the location that needs to be monitored, the EnviroDIY Sensor Station can be used at that site but will need to be manually uploaded (see Programming and Activating an EnviroDIY Sensor Station: Registering a Station on MonitorMyWatershed.org).
Sunlight Availability: The sensor stations use a solar panel to recharge the lithium ion polymer battery. It is important for solar panels to be positioned so that adequate sunlight is received throughout the day. Vegetation can be cleared to help with sunlight exposure (Figure 8.1). However, keep in mind that as the seasons change, vegetation will grow back and can cover the solar panel (Figure 8.2).
It is also crucial to pay attention to the direction the solar panel is facing. Generally, in the northeastern U.S. a solar panel will receive the most sunlight if it is facing south (Figure 8.3). At times, solar panel orientation can be more important than the size of the panel. If the solar panel is facing in a direction that receives little sunlight, even with a large solar panel charging may be inadequate. The best position of a solar panel will vary depending on each specific station location. Heavy canopy cover, for instance, may dictate that the solar panel needs to be faced toward breaks in the canopy cover. Positioning a solar panel in a north orientation will rarely occur because it will receive substantially less direct sunlight. As the angle of the sun changes throughout the seasons it may be necessary to adjust the solar panel to obtain optimal sunlight.
River and Stream Morphological Features: Ideally, sensor stations are placed on straight and even flow sections of streams where debris will not accumulate on the sensors and where discharge measurements can be made for use in association with sensor data for developing rating curves. Placement within the stream channel should take into consideration stormwater flows and whether these flows will be detrimental to the sensor stability and exposure to scouring forces and movement of bedload. To this end, if storm flows are anticipated to be potentially damaging to sensors it may be necessary to position sensors in backwater areas, stream bends, or behind obstructions (e.g., boulders, abutments) that would provide protection from extreme stormwater flows.
Considering stream morphology, flow, and other physical characteristics when identifying sensor station installation locations is important for ensuring the collection of data suitable for satisfying project goals. Lateral (cross-channel) mixing may require a long distance of stream channel, so do not measure immediately downstream of tributaries or point sources that may not represent the mainstem of the channel (Figure 8.4). Turbulent streamflow may aid in mixing, but turbulence can create problems in monitoring field parameters, such as DO or turbidity. A location near the streambank may be more representative of local runoff or affected by point-source discharges upstream, whereas a location in the channel center may be more representative of areas farther upstream in the drainage basin. Large streams and rivers usually are monitored from the downstream side of bridge abutments, assuming that safety hazards and other difficulties can be reduced or overcome.
Cross-section variability and upstream influences are major factors for site selection. Cross-section surveys of field parameters must be made to determine the most representative location for monitor placement. Sufficient measurements must be made at the cross section to determine the degree of mixing at the prospective site under different flow conditions and to verify that cross-section variability at the site does not exceed that needed to meet data-quality objectives. Additional cross-section measurements must be made after equipment installation to ensure that the sensor station installation is representative of the stream during all seasons and hydrographic flow conditions (Wagner et al., 2006).
The best location for a monitoring site is often one that is best for measuring surface-water discharge. Although hydraulic factors in site location must be considered, it is more important to consider factors that affect water quality conditions. The same hydraulic factors that must be considered when selecting a specific site for measuring discharge in a channel also should be considered in selecting a water-quality monitoring location. Both purposes require a representative site that approaches uniform conditions across the entire width of the stream (in contrast with the condition shown in Figure 8.4).
The measurement point in the vertical dimension also needs to be appropriate for the primary purpose of the monitoring installation. The vertical measurement point can be chosen for low-, base-, or high-flow conditions; if bed movement or sensor location during low flow is a problem, consideration should be given to moving the sensors along a bridge. For a medium to small stream with alternating pools and riffles, the best flow and mixing occurs in the riffle portion of the stream; however, if flooding changes the locations of shoals upstream of the monitoring site, the measurement point may no longer represent the overall water-quality characteristics of the water body. Streams subject to substantial bed movement can result in the sensors being lost or located out of water following a major streamflow event, or at a point no longer representative of the flow. A site may be ideal for monitoring high flow but not satisfactory during low flows.
The configuration and placement of water-quality monitoring sensors in cold regions require additional considerations in order to obtain data during periods of ice formation. Overall, a monitoring site should be safe and accessible, meet minimum depth requirements of the equipment, minimize the risk of vandalism, and be characteristic of the entire stream. It may be necessary to reconnoiter the site under several flow conditions before a determination is made.
Bank Conditions: When positioning logger stations on the stream bank, the stability and erosion potential of banks should be considered. Abstain from installing a station near areas on the stream bank that seem to be undercut. If a station is installed on an erosional surface, it is likely that the station will not survive long term.
The galvanized pipe that is used to mount the data logger box should be at a sufficient depth to withstand flood conditions without becoming dislodged: marshy areas with soft sediment will require at least 4’ of pipe installed into the ground, whereas rockier locations may only need 30-36” installed into the ground). If a location has sediment that is very sandy or has a lot of gravel, be aware that your pipe underground needs to be installed at a depth where the pipe will not rotate or spin loosely in the hole. If you try to spin or wiggle the pipe, it should hardly move. If the pipe continues to move, you will need to drive the pipe deeper into the ground until it no longer moves.
Bed Composition: The composition of the stream bed can determine what station parts are able to be used, particularly in regards to the in-stream sensor bundle. For example, if the bed composition is extremely rocky, or near bedrock, a shorter 2’ rebar will need to be used. If the stream bed is extremely soft (e.g., loose sand or mud) it maybe be necessary to use a longer rebar or other stake to ensure the sensors are stable within the channel. For this type of custom installation consult the www.envirodiy.org/forums/.
Distance from Sensors to Logger: For ease of access in conducting maintenance, the sensor bundle should be placed at a location out of the main flow path, near the stream bank, where the water is at least 30 cm but no more than one meter in deep. Sensor wires should never run across the stream bottom from one bank to another because this greatly increases the probability of damage to the sensor in high flow conditions. Always mount the sensor bundle near the same bank where the logger box is mounted. The OBS-3+ turbidity sensor needs at least 30 cm of water depth due to obtain good readings. The sensor wires are about five meters in length, however extension cables can be purchased if the location of the data logger box needs to be father than five meters for any reason. Historical hydrographs and first-hand observations describing overbank flooding should also be considered when positioning logger stations on the bank. Data logger boxes should be positioned well above any expected flood levels. In the event that a logger does become submerged, the waterproof casing will keep it protected.
Step-by-step videos are available on the Videos page.
Locate spot to deploy sensors
- Find location of sensor station in the stream.
- Keep in mind the flow pattern of the stream, the water depth, and take note of any built up debris from past storms and be sure to avoid these areas.
- When choosing a location for the sensors, keep in mind that there also needs to be a suitable location for the logger box on land within 15’.
- Hammer the test post into the stream bed to find a secure, stable spot.
- Set the anchor pole (solid steel rebar with a pointed end that will be the attachment point for the sensor bundle).
- This rebar is driven into the stream bottom using a hammer and a wood block to prevent damage to the top of the post.
- The length of the rebar should be driven far enough into the stream bottom to ensure that the sensor bundle and anchor pole will not move during flood events and does not twist within the substrate of the stream bottom.
- The rebar should be hammered so that the top of the rebar is flush with the water surface to minimize accumulation of debris on the sensor bundle.
- Additionally, the anchor pole should not be driven so deep into the stream bed that it cannot be removed when the study is complete or in response to changes in the site requiring station repositioning.
Prepare to install the logger box
- Find a location on the stream bank for station post/box with good sun exposure.
- Clear the area on the bank where the data logger station will be positioned by removing vegetation that may block sunlight to solar panels or that may cause problems during floods.
- Alternatively, if vandalism is a concern, leave some vegetation around the station to help camouflage the station.
- Two sections of ¾” diameter pipe are used to create the data logger mounting pole: a shorter length (typically 36” long) that will be hammered into the ground and a longer length (typically 60” long) that will mount the logger box above ground on the stream bank.
- Pound in a stake to create a pilot hole, then attach the ¾” coupler to the shorter pipe and hammer the pipe into the ground (place a 2 x 4 wood block on top of the coupling to prevent damage to the threads).
- Stop hammering the shorter pipe into the ground with a few inches between the ground surface and the coupler.
- Tighten the coupler onto the short pipe using pipe wrenches.
- Twist the longer pipe onto the coupler and tighten again using pipe wrenches; note that pipe lengths vary based on the specific needs of each station location.
Assemble logger box on pole.
- Attach data logger box to pole by 1) sliding two stainless steel hose clamps (⅞” to 1 ½” size) through the gap between the back of the logger box and mending plate, 2) tightening hose clamps around pole.
- To make it easier to secure, place the nuts of both hose clamps on the same side.
- Determine how high the data logger box should be on the pole by considering who will need to access this station: if children will need to access the station, place the logger box at eye level of the children, alternatively, if the volunteers are taller, place the logger box at a height that will be suitable for all needing access.
- Consider where the high flow marks are in the area and look for woody debris and other markers such as garbage to get an idea of how high the water usually gets during high flow events; place the logger box above these to prevent it from being submerged.
- Start with the bottom hose clamp; tighten it a bit, then do a final tighten at the end when solar panel is added.
Prepare the sensor bundle.
- Cut the ¾” PVC pipe to the appropriate length.
- Use the lettering that exists on the PVC pipe as a way to accurately measure the top of the black rebar with holes.
- Measure height of PVC where it lines up with the top of the rebar.
- NOTE: always cut at height of rebar, not the stream surface line.
- Do the final cut of PVC using the pipe cutters and fit on rebar to check and make sure the cut lines up properly. If the cut was done properly the top of the PVC pipe will be flush with the top of the rebar.
Attach sensor bundle to PVC.
- Unwind the sensor cables at the data logger box.
- Attach sensors to PVC using the larger hose clamp and the nut driver or screw driver. Things to note:
- The sensors need to be a certain height from bottom (about 3” if possible); this is site specific and depends on the amount of sediment and rocks on the stream bottom.
- The turbidity sensor needs to be facing into the stream channel not towards the stream bank. The four metal screws within the CTD sensor should be facing upstream so that fresh water is continually hitting it.
- The hose clamp should be tightened over electrical tape around the turbidity sensor to provide extra protection, and also above center/middle of CTD sensor to avoid damage to the pressure transducer.
- Tighten to the extent that sensors don’t move up/down, but not tight enough to cause damage to the sensors.
Attach sensor bundle to rebar.
- Place sensor cables from sensor box (on land) to the rebar mounted in stream.
- Consider bank stability and debris pile ups.
- Lay out all cables first, then bundle what you don’t need at sensor box.
- Wrap cables under roots along the bank that are stable.
- Do not attach to vines or loose materials that will break during storm events.
- Secure the sensor bundle using the metal retaining pin.
- Line up holes of PVC and post.
- Insert pin from upstream side so clamping is on downstream side (to prevent debris from building up).
- In the stream: make sure angle of turbidity sensor is facing into channel and there are no obstructions within one foot.
- If window is facing the wrong way, the retention pin can be used as a handle to twist metal rebar until turbidity window is facing the correct direction.
- Work from the sensor bundle back to the data logger box when securing the cables using tent stakes and zip ties.
- Secure the sensor cables along bank keeping these things in mind:
- Use zip ties, tent stakes, hammer, and wire clippers.
- Use stable roots.
- Leave a little give between stakes (stabilization spots).
- Two zip ties might be needed if a root is wide.
- Zip tie tips:
- There are two sizes: 11” (long) and 8” (short).
- Don’t pull too tight in the cold as they are brittle and will break.
- Don’t crimp zip ties along sensor cables.
- Use black zip ties with UV protection for outdoor use, not white zip ties (these will break very quickly if being used outdoors).
- Cut zip ties at a perpendicular angle and as short as possible.
Attach solar panel to post.
- Add solar panel.
- Solar panel orientation should be facing south or west. Keep in mind obstructions to sunlight such as leaves, and account for this when choosing a direction for your solar panel.
- Adjust as necessary for height.
- Attach with U-bolt and tighten by hand first, then with wrench (7/16”); don’t tighten too much, as the plate is aluminum and may bend.
- Connect solar panel wire: there are two notches on side of wire connect to panel. Push wire from logger box all the way past notches (this makes it waterproof).
- Once connected, an orange light will appear on the Mayfly Data Logger inside of the logger box.
- This display will show whether or not the lithium ion polymer battery is being charged.
- If the LED light on the Mayfly Data Logger does not come on, this means that either the solar panel wire is not connected, or that the battery is fully charged.
- Secure solar panel wire with zip ties.
- Secure all wires and trim with cutters (making a horizontal cut as this makes the cut less sharp).
Attach sensor wires to the post.
- Bring sensor cables to the logger box and loop them neatly around box.
- Secure the sensor cables with zip ties: do not pinch or crimp the CTD cable (particularly the white vented area of the cable) because this could damage the vented tube inside the cable and affect pressure (depth) measurements.
- Trim zip ties on the pole, but do not trim zip ties along wire length on the ground so you can find the wire if needed at a later time.
Make final adjustments.
- At the top of the sensor station post, add plastic ¾” cap to keep out critters such as bees, ants, or spiders.
- Turn on sensor station and check for functionality by looking at the light pattern as well as plugging in USB to Android phone to check sensor readings.
- Place lock on box.
- The specific size padlock that fits on the Pelican case of the EnviroDIY logger box is a AB410 CCL Sesamee Resettable Combination Padlock, and can be found at the following link: www.padlockoutlet.com/ab410-sesamee-resettable-combination-padlock.html. Using a combination padlock, instead of a padlock with a key, is beneficial if the station is going to have multiple people maintaining the station and needing access to the logger box. The best location for locking the logger box is the bottom hole on the logger box door.
- A laminated sign, sticker, or business card should be attached to the logger box (Figure 8.23). Information on this signage should include at minimum contact information of the person/organization managing the station. Additional information may include descriptions of station function, monitoring goals, project descriptions, and educational intentions.
Sensor Station Management
Management of the EnviroDIY Sensor Station includes quality control practices, sensor cleaning, station maintenance, power issues, data curation, and other logistical issues associated with keeping a sensor station functioning properly. Important considerations in overall station management are:
- For quality control, sensor station data should be cross-checked by conducting on-site measurements with calibrated handheld probes.
- For quality control, all sensor station data should be backed up on a secure server or hard drive.
- Sensor cleaning should be done according to what the data dictate (i.e., if data indicate fouling, sensors should be cleaned). The Campbell OBS-3+ turbidity sensor is particularly vulnerable to fouling and generally needs to be cleaned on a weekly basis.
- Battery levels below 3.7 V indicate that sensor station function is at risk and alternative power options should be explored.
- For quality control, maintenance and sampling should be documented using the Field Visit Data Form or in waterproof field notebook. Copies of field data sheets and/or notebooks should be made on a regular basis and stored in a secure location.
Field Visit Documentation
Any time a visit is made to a sensor station it is recommended that sensor cleaning activities, staff gauge height, photos of site and station conditions, and any other maintenance and/or monitoring activities described in section 9.2, as well as supplemental data collection activities (e.g, discharge measurements and grab sample collection; see Supplemental Sampling, Rating Curves, Loads) be documented using the Field Visit Data Form (Table 9.1) or in a field notebook. Back up copies should be made of data sheets or field notebooks on a regular basis. In addition to basic quality control, documentation of all activities associated with sensor station management will allow identification of issues and support improvements to the management process.
When applicable, each individual grab sample (see Supplemental Sampling, Rating Curves, Loads) collected should be documented in the Grab Sample Information section of the Field Visit Data Form on page 2. Depending on the context of the project, grab samples may be processed internally or shipped to an external lab (see Supplemental Sampling, Rating Curves, Loads).
A Stream Discharge Data Form should be completed every time a distinct set of discharge measurements are made (see Supplemental Sampling, Rating Curves, Loads and Appendix G for details on measuring discharge). These data should then be entered and saved into the Discharge Rating Curve Calculator for that site (see Supplemental Sampling, Rating Curves, Loads).
To ensure integrity of the data, it is recommended that data sheets be printed on weather resistant paper (e.g., Amazon: “Rite in the Rain All-Weather Copier Paper, 8 ½” x 11″, 20# White, 200 Sheet Pack (No. 8511)”. It is also recommended that data sheets be completed using a graphite pencil — ink pens may not be effective when used on the weatherproof paper. A waterproof field notebook (e.g., Amazon: “Rite in the Rain All-Weather Hard Cover Notebook, 4 3/4″ x 7 ½”, Yellow Cover, Journal Pattern (No. 390)”) can be used instead of field data sheets, however, it should be noted that without a field sheet guide it can be easy to forget key pieces of information.
|Table 9.1. Description of Field Visit Data Form|
|Name(s)||Names of individuals conducting the work|
|Site ID||ID assigned if a sensor station was deployed in association with the Delaware River Watershed Initiative|
|Stream Name||Name of stream|
|GPS (Lat/Long)||Site coordinates in Decimal Degrees|
|Photos (Yes/No)||General record of whether photos were taken — a repository for these photos may be developed|
|LoggerID||Running tally of all EnviroDIY sensor stations that have been built and deployed. Each individual sensor station has a specific SL number, e.g., SL###|
|Location||Simple description of general site location, e.g., nearby bridge crossing or confluence with another stream|
|Date, Time, AM/PM, EST/EDT||Date on which site visited; Time at which form was completed; AM/PM; Eastern Standard Time = fall/winter, Eastern Daylight Time = spring/summer|
|General Notes/Photo Descriptions||Space to record information about the specific visit to the site and/or to describe photos that were taken|
|Staff Gauge Height (m), Time, AM/PM, EST/EDT||Water level on staff gauge measured in meters; Time at which measurement recorded; AM/PM; Eastern Standard Time = fall/winter, Eastern Daylight Time = spring/summer|
|Sensor-Reported Water Depth (mm), Time, AM/PM, EST/EDT||Depth reading from CTD sensor in millimeters; Time at which measurement recorded; AM/PM; Eastern Standard Time = fall/winter, Eastern Daylight Time = spring/summer|
|Precipitation||Amount of precipitation|
|Water Clarity||Coarse assessment of water clarity|
|SENSOR STATION MAINTENANCE|
|Sensors Submerged?||If sensors are not submerged or partially submerged it may be necessary to consider repositioning the sensors|
|Location of Sensors Changed?||Documentation of any re-location of sensors. If rating curves are being developed moving sensors may require assistance from Stroud Water Research Center|
|Cleaned Sensors? Exact time||Important to document exact time when sensors were cleaned – allows reference to specific spot in time-series data|
|Retrieved Memory Card?||Memory card needs to be removed to supplement online data if gaps exist or if online data are not available|
|Changed Batteries?||Batteries may need to be changed if solar charging is inadequate — record of this will help characterize site conditions|
|Cleaned Solar Panel?||Solar panel may collect dust or debris; record of this will help characterize site conditions|
|Other sensor station maintenance?||Describe any other situational issues|
|GRAB SAMPLE INFORMATION|
|Grab Sample Taken?||Record of collection of sample|
|Sample Number||Specific number listed on grab sample bottle — unique number for every bottle — this is an important detail to record|
|Bottle Type||“Square Nalgene” 1 L or 500 mL|
|Lab Sent To||Lab that will analyze sample, usually this is Stroud Water Research Center|
|Time||Exact time (to the minute) that the grab sample was collect — this is an extremely important detail to record because it will allow the grab sample data to be matched with specific sensor measurements at the time grab was collected|
|Volume||Grab samples will either be 500 mL or 1 L|
|Date Shipped||Date when sample was delivered to FedEx for shipment|
|Chain of Custody #||Specific number from Chain of Custody form provided in grab sample shipping kits|
|Field Meter Brand/Model/Serial#||Record of the instrument used to collect the data|
|Was meter calibrated? Standard, Calibration Result||Record of calibration|
|OTHER PARAMETERS (e.g., NITRATE, PHOSPHATE, CHLORIDE, pH, DO)|
|Parameter, Brand/Model, Result||Record of any other measurements using kits, meters, or other methods|
|Field duplicate Taken of Grab Sample?||Sometimes grab samples may be collected in duplicate for quality control purposes|
|Performed Cross-Section Survey?||Cross-section survey is done when sensor station is installed — mapping of channel profile for predicting cross-sectional wetted area for use in calculating discharge|
|Flow Measurement w/ Flow Meter?||Discharge measurements collected using a flow meter|
|Flow Measurement w/ Neutrally Buoyant Object||Discharge measurements collected using a timed float of a neutrally buoyant object|
|Flow Measurement w/ another method?||Discharge measurements collected using any number of methods including timed fill, salt-dilution, etc.|
Although data can be stored on the microSD card for >800 years and the online data portal (MonitorMyWatershed.org) also has long-term data storing capacity, data should be backed up to a secure hard drive or server on a regular basis. Recommended data back-up frequency is every six to eight weeks if data are transmitted to the data portal and every two weeks if data are only stored on the microSD card. This process ensures that long term data records are secure even if damage to Mayfly Data Logger, microSD card, or a website malfunction occurs. Downloaded data files should be filed chronologically and each file title should include a unique identifier for the particular station and the date range or download date of the file.
Some sensor stations have intermittent cell coverage which could cause data transmission to the Monitor My Watershed website to stop. When this happens, sensor data will continue to be stored on the microSD card and missing data can be retrieved from it. Regardless of whether a sensor station is online, sensor data will always be stored on the microSD card; therefore, if data stop transmitting to the website due to cell coverage issues, the complete data set can still be accessed via the microSD card. To download data from the microSD card:
- Open the logger box.
- Turn off the Mayfly Data Logger.
- Remove microSD card from board.
- Insert blank microSD card (so that data can continue being recorded).
- Turn the Mayfly Data Logger back on.
- Insert microSD card (the one you just removed) into a standard SD card adapter.
- Insert adapter into appropriate port on computer.
- Save data file to secure hard drive or server with the recommended format: SL#_mm-dd-yy.
- Delete data from microSD card for return to Mayfly Data Logger.
- Open data in Excel and graph as needed.
The Meter Hydros 21 CTD and Campbell OBS-3+ turbidity sensors are both factory calibrated and should not need recalibration (https://manuals.decagon.com/Manuals/13869_CTD_Web.pdf; https://campbellsci.com/documents/us/manuals/obs-3+.pdf). However, it is highly recommended that calibrated handheld sensors be used to cross-check sensor station data (see Appendix C for a list of portable handheld sensors). All quality control measurements should be recorded in the “In-situ measurements” section of the Field Visit Data Form. Any discrepancies between sensor station data and data collected using hand held sensors should be further investigated. If sensor station data are identified to be inaccurate, consult the sensor user manual or the manufacturer of the sensor.
These cross checks of sensors should be done on a quarterly basis and more frequently if accuracy of sensor station data are questionable. A certain level of error is inherent in the sensor station sensors and in all handheld sensors. There is no defined limit specified here for what is acceptable in terms of variability in these measurements. More importantly, measurements should always be documented and evaluated in the context of the specific project intentions for data usage.
To ensure reliability of sensor station data, data trends and patterns should be regularly evaluated. Graphical displays of data and assessment of the data in the context of the specific site conditions and ecological expectations should be used as a quality control measure to ensure data reliability. Issues and problems to look for in data-quality assessment are extensive but there are specific common issues that can be considered when troubleshooting for problems in sensor data (see Appendix J and K). If sensor station data appear suspicious (outside of normal range for the site, erratic, etc.) cross-checks with other calibrated sensors should be done immediately. If sensor station data appear normal, cross-checks with handheld sensors should be done quarterly as a general quality control method.
To confirm the accuracy of depth measurements from the CTD sensor, check the sensor depth using a metric ruler or meter stick to measure depth from the window of CTD sensor (where pressure transducer is located) to the water surface. This hand-measured depth should be similar to the CTD depth sensor reading; however, the CTD sensor is not designed for measuring absolute depth with high precision. Instead, it is designed to measure changes in depth, which is how it is being used here (i.e., for tracking changes in discharge). Therefore, the hand measurement of sensor depth should be used as a guide in tracking sensor function; the specific depth it reads should be consistent and predictable in accordance with changes in water levels. If this is found to not be the case then the sensor may not be functioning correctly and assistance should be sought via www.envirodiy.org/forums/. Additionally, the offset between the staff gauge and the sensor depth can be used as a frame of reference for evaluating any changes to the sensors or the staff gauge.
Sensor Station Maintenance
To ensure data from the sensor station are accurate and continuous it is necessary to:
- Keep sensors clean and area around sensors free of debris.
- Keep area around logger box clear of vegetation, debris, and insects.
- Keep the solar panel clean and exposed to as much sunlight as possible.
- Maintain battery level and cellular transmission of data by ensuring the battery is > 3.7 V.
See Appendix E for a checklist of sensor station maintenance activities and timing of these activities. The key issues in maintaining a sensor station are:
- Logger box and Mayfly data logger
- Functionality of the sensor station is suspect if battery level falls below 3.7 V.
- Any moisture inside the logger box can cause logger board malfunction.
- Cycling the power (turn board off, pause 10 seconds, turn board back on) as done with a computer is a common fix for loss of cellular transmission and other miscellaneous issues.
- Turbidity sensor
- Turbidity sensor is particularly vulnerable to fouling: any debris, sticks, leaves that attach to it or algae that grows on it will cause false high readings and degrade data quality.
- Data from turbidity sensor can be affected by objects in its field of vision (38 cm); this can include debris and sediment piled up around or on the sensor and algae growing on the sensor.
- Because of the extended field of vision of the turbidity sensor, turbidity data are suspect when water level is less than 30 cm.
- CTD sensor
- The pressure transducer that measures water depth can be damaged if ice forms on the sensor.
- As long as the CTD sensor is submerged, data from the sensor should be accurate.
- The CTD sensor is not as prone to fouling as the turbidity sensor; however excessive debris accumulation on the conductivity Wenner array (four screw heads) will cause inaccurate conductivity readings.
- In rare cases the Meter Hydros 21 CTD sensors have had factory production issues that affect data accuracy. Quality control checks with calibrated handheld sensors (see Sensor Station Management) should be used when data are suspect.
Cleaning Sensors In-stream
Most of the time sensors can be cleaned without removing the sensor bundle from the stream. Depending on the site and stream conditions, and depending on weather and any other environmental influences or unforeseen technical problems with the stations, sensor cleaning frequency may range from weekly to monthly.
The Meter Hyros 21 CTD sensor cleaning involves scrubbing the semi-enclosed parts of the sensor (four point Wenner array shown in Figure 9.1) with a scrub brush to remove any accumulated dirt and debris. Cleaning the CTD sensor may not be necessary on a regular basis and is dependent on whether debris accumulates inside the partially enclosed part of the sensor.
In contrast, the Campbell OBS-3+ turbidity sensor usually needs regular cleaning as it will tend to accumulate algal growth, leaves, sticks, and debris, all of which will affect data accuracy. Unlike the CTD sensor, which is not highly affected by this type of debris accumulation, the turbidity sensor is affected by any debris within its field of vision, which extends to 38 cm from the sensor window. The rate at which this accumulation occurs is variable, so maintenance schedules should be planned on a site-specific basis (see When to Clean Sensors).
To clean the sensors in-stream:
- Clear all accumulated debris on the stream bottom away from the sensors using hands and feet or a tool if necessary.
- *Note: The turbidity sensor can read 38 cm away from the sensor window so make sure debris near the sensor is cleared away accordingly. If anything is in the sensor’s field of vision (e.g., rocks, accumulating sediment, stream bank, woody debris) the turbidity readings will be inaccurate.
- Remove all large pieces of material (leaves, sticks) from the sensor bundle by hand or with the long bristles of the sensor brush.
- Use the longer bristles of the sensor brush to gently clean the side slots on the CTD sensor (Figure 9.1).
- If there are any sticks or hard objects wedged in the slot make sure to remove them carefully so as to not damage the pressure transducer (the white disc in the CTD sensor).
- Use the long bristles to clean the four screw heads inside the side slot (these are the points at which conductivity is measured).
- Use the short stiff bristles of the sensor brush to clean the signal window on the turbidity sensor (Figure 9.1).
- If a sensor brush is not available, using any other plastic bristle brush or your fingers is acceptable for cleaning the sensors.
Do not remove the sensor bundle (Figure 9.2) from the stream unless cleaning cannot be done while sensors are in the stream. If removing the sensor bundle from the stream is absolutely necessary, follow these steps:
- Record the orientation of the sensors in the stream and note the position of the PVC conduit with regard to the top of the steel mounting stake. Also note the position and orientation of the mounting pin. When you return the sensor bundle to the stream you will need to position the sensor bundle with the same orientation and at the same level to ensure consistent data.
- If necessary, remove tent stakes and zip ties securing the sensor cables to stream bottom and tree roots. This loosens the sensor cable so the sensors can be brought to the stream bank.
- Remove the mounting pin; it functions like a standard safety pin.
- Slide the sensor bundle off of the steel stake.
To remount the sensor bundle (Figure 9.2) on the mounting stake:
- Make sure orientation of sensor bundle is the same as it was before you removed the sensor bundle.
- Slide sensor bundle (via PVC conduit) back onto the steel stake.
- Match holes in PVC conduit up with holes in steel stake at the same vertical level as before.
- Slide mounting pin through both sets of holes and lock mounting pin.
Clearing Around the Station
Similar to the sensors, the logger box, solar panel, and associated components will require regular maintenance. All debris and vegetation should be cleared from around and above the logger station.
The logger box is waterproof but it should be opened periodically to confirm that no moisture has entered and that the Mayfly Data Logger and all other internal components are intact. Vegetation should be cut back from around the logger box and solar panel. Although the logger box is fully sealed from the external environment, vegetation can get in the way when opening the box and can possibly introduce water, insects, and debris. Anytime the logger box is opened it is important when closing the box to ensure that there is no grass, leaves, stems, or other debris breaching the seal: this debris will allow moisture into the box and damage the electronics.
Vegetation covering the solar panel can reduce solar exposure and can cause additional debris and dust accumulation on the solar panel. Vegetation should be cut back on a regular basis from beneath and around the logger box and solar panel using any number of different tools (Figure 9.3). Clean dust and debris from solar panel with your hand or with a soft cloth. The canopy above the solar panel should be kept open enough to ensure consistent exposure to sunlight throughout the day.
When to Clean Sensors
Generally, sensors need to be cleaned on a weekly basis and sometimes more frequently during the fall or during storms when leaves and debris are abundant.
If the four-point Wenner Array (screw heads) in the CTD sensor becomes covered in silt, algae or other debris the data will be affected. This type of fouling does not usually occur quickly so the CTD sensor will usually produce accurate data for weeks (or even months in some cases) without needing to be cleaned.
The Campbell OBS-3+ turbidity sensor generally requires much more frequent cleaning than the CTD sensor. Because the turbidity sensor functions by sending light out into the water column and then detecting reflection of this light off of suspended material, any leaves, grass, sticks, or other material attached to or near the turbidity sensor will cause false high turbidity readings. Therefore, it is important to clean the turbidity sensor as soon as any debris attaches, particularly during storm events when it is typically most important to acquire accurate turbidity data. Unless debris detaches naturally (i.e., carried away by the current), turbidity readings will remain inaccurate and unusable until the debris is removed. During high water conditions it may be necessary to attach a brush to a long pole in order to reach the turbidity sensor and free any attached debris (see Figure 9.1).
For developing the TSS/turbidity rating curve (see Supplemental Sampling, Rating Curves, Loads) for a site it may be necessary to clean the turbidity sensor on an hourly basis during storms. The sensor can foul more frequently than normal during storms, but this is also the time when accurate turbidity data are most important. Therefore, when attempting to develop the TSS/turbidity rating curve it may be most effective to clean the sensor hourly and collect grab samples over this same time period.
If sensor data are online, these real-time data can be used to inform timing of sensor cleaning. Normal turbidity readings for cloudy and muddy water (in the eastern U.S.) are generally < 300 NTU (although this can vary); therefore, if online turbidity readings show NTU levels well above this (e.g., > 1000 NTU) this is a strong indication that the sensor is fouled. Furthermore, if turbidity readings suddenly spike and/or increase dramatically and are not associated with changes in depth (i.e., increased stream flow due to precipitation) this is also a reliable indication that fouling has occurred (Figure 9.4; see Appendix J). False-high turbidity readings will also occur with algae growth on the sensor. In these cases turbidity data may increase gradually over multiple days and may be in the normal range for water conditions (i.e., <400 NTU; Figure 9.4 and see Appendix J). Understanding the natural turbidity ranges and stream response to precipitation will help in determining when turbidity readings are suspect. In all cases of false-high turbidity readings removal of debris and cleaning of the sensor with the sensor brush should be done as soon as possible.
The issue of highest concern during winter is damage to the pressure transducer in the Hydros 21 CTD sensor. The pressure transducer, as the name implies, is sensitive to pressure changes, so when water expands during freezing the pressure of the ice directly against the pressure transducer can damage it. If freezing does occur around the CTD and turbidity sensors all data will be inaccurate during this period time and should not be used. Turbidity data should return to normal after the ice melts. Depth data will be suspect and should be closely checked after the ice melts to determine if the sensor was damaged. Temperature and conductivity data should return to normal but should also be checked to ensure accuracy after the ice thaws, as these can be affected if the pressure transducer is damaged.
The shallower the water in which sensors are positioned, the higher the risk of freezing. Monitoring air and water temperatures during the winter can be important, especially in cases where sensors are located within the part of the water column where freezing may occur. Freezing risk is generally highest in small streams where there are not deep locations available for sensor placement, but freezing may also be a risk for deeper sensors during severe and extended low air temperatures when ice layers get thicker than normal. Ice that forms at a depth above the sensors is not a high risk to sensor integrity, but this can make it difficult to access sensors. If the ice has to be broken with a hammer or chisel there is risk of damage via direct contact of the hammer/chisel with the sensors and/or sensor cables. There is also a risk of damage to sensors through shifting pieces of ice as they are broken and through possible damage from the shock of the hammer or chisel. If freezing of the water surface is an issue and cleaning sensors during this time period is important it is recommended that ice layers be removed on a frequent enough basis so as to not require a hammer or chisel for breaking the ice (i.e., remove ice when it can be broken by hand). Clearing ice at this frequency will ensure that the sensors are accessible for cleaning and ensuring accurate data.
The target voltage level for the battery is 3.7 V or higher. Below this level the station may lose functionality and data may stop transmitting to the web data portals. The typical battery voltage pattern is for the battery to charge during the day and lose power during the night (see Figure 9.5). In certain scenarios where sunlight is restricted or diminished due to canopy coverage, cloudy days, or seasonality, battery level may decline over multiple days due to inadequate solar exposure and incomplete charging (Figure 9.5 and Appendix J). This gradual decline may also happen when cell coverage is intermittent, in which case the Mayfly Data Logger may make repeated unsuccessful attempts to send data to the website using extra power each time (Appendix J).
In cases where solar charging is not adequate for keeping the battery above 3.7 V it may be necessary to assist in powering the stations with a second fully charged replacement battery (Figure 9.5). In these cases it will be necessary to acquire a backup battery, charger, and adapter and cycle batteries according to the requirements of the specific situation (Figure 9.5).
In some cases, solar charging may vary throughout the year with seasonal canopy cover changes as well as other canopy cover changes (e.g., fallen trees, growth of vegetation). Reposition the solar panel to receive better sun exposure. In extreme cases DIY methods can be used to place solar panels and/or the logger box itself in locations that will provide adequate solar coverage (e.g., placing solar panel high above vegetation and separate from the logger box).
The staff gauge is intended to serve as the “master” reference for stream depth at the EnviroDIY Sensor Station location. The staff gauge will almost always measure a different depth than the CTD sensor simply because they are positioned at different locations in the stream. For general reference and to confirm sensor or staff gauge data consistency, it is important to know the offset between the staff gauge and the sensor depth, i.e., the difference is between the water level as measured by the staff gauge and the water level as measured by the CTD sensor. This offset between staff gauge water depth and sensor water depth is also a key piece of information for use in developing the hydrologic (depth/discharge) rating curve. Staff gauges installed in association with EnviroDIY Sensor Stations are currently considered to be semi-permanent. They are set in the stream on ½” pipes which are generally stable and resilient; however, high flows in larger streams and rivers can damage the gauge by bending or breaking the pipe on which the gauge is mounted. If a staff gauge is bent or damaged it should be reset or replaced to the exact same depth using the the sensor/staff offset as the reference point.
Supplemental Sampling, Rating Curves, Loads
To further understand and explain the ecology of a site and lend insight into EnviroDIY Sensor Station data patterns, a variety of additional measurements and sample collection methods can be employed. Handheld sensors, dilution kits, and other point-in-time sampling methods (see Appendix C), as well as grab samples for laboratory analysis, can be used to gather additional explanatory information. Site specific monitoring questions will dictate which parameters are most important to investigate. Furthermore, understanding landscape conditions in a watershed including factors such as point sources, commercial, industrial and agricultural land uses, wastewater treatment plants, and other potential sources of pollution can help in understanding and explaining EnviroDIY Sensor Station data patterns (see www.wikiwatershed.org for a toolkit of resources to investigate these types of landscape-level data).
Described here is a process for collecting data that can be used to enhance the value of time series data. The “supplemental sampling” that is described is required when monitoring intentions dictate additional data are needed for monitoring of discharge patterns and concentrations and amounts (loads) of total suspended solids (TSS) and chloride (Cl–). Specifically, this supplemental sampling focuses on:
- Performing discharge measurements at a range of water levels (i.e., storm flows) (Section 10.2.1).
- Collecting samples of water (“grab samples”) across the range of turbidity levels observed at a site (Section 10.2.2).
- Collecting grab samples across the range of conductivity levels observed at a site (Section 10.2.2).
These discharge measurements and grab samples are used to develop rating curves (Figure 10.2) that describe the following relationships: 1) sensor depth (mm) to measured discharge (m3/s), 2) sensor turbidity (NTU) to laboratory TSS (mg/L), and 3) sensor conductivity (μS/cm) to laboratory Cl– (mg/L). Once developed, regression equations from these rating curves allow the time-series Sensor Station depth, turbidity, and conductivity data to be transformed to the more quantitative and ecologically informative estimates of stream discharge (m3/s), total suspended solids (TSS)(mg/L), and chloride (mg/L), respectively (see Load Calculator spreadsheet). It is imperative that the sensors are producing accurate data when grab samples and discharge measurements are collected, so that sensor data can be matched up with the supplemental data for developing the rating curves. This means that sensors should be cleaned prior to collecting grab samples. In high water (when grab samples are often collected) an extension for the sensor brush may be needed so that sensors can be reached for cleaning. Once discharge, TSS, and chloride data are generated via application of rating curve equations, these data can then be used to calculate amounts of water (discharge) and material (TSS load and Chloride load) moving in a stream over time (e.g., stormflow flashiness, sediment and chloride loads, identifying criteria exceedances).
The following list describes key elements of supplemental sampling and data processing associated with EnviroDIY Sensor Stations:
- Supplemental sampling consists of:
- Discharge measurements over a wide range of discharge to develop hydrologic rating curve.
- Grab samples over a wide range of turbidity to develop TSS versus turbidity rating curve.
- Grab samples over a wide range of conductivity.
- Grab samples for answering site-specific questions.
- Supplemental sampling allows development of rating curves for depth versus discharge, TSS versus turbidity, and Cl– versus conductivity
- Once developed, a rating curve will allow you to:
- Infer stream discharge (m3/s) based on depth data
- Infer Total Suspended Sediment (TSS, mg/L) concentrations based on turbidity data
- Infer Chloride (Cl–, mg/L) based on conductivity data
- Once you can infer this information you essentially have continuous discharge, TSS, and Chloride data, meaning you can then calculate amounts of material moving in the stream for specific time periods (i.e., loads), e.g.:
- “During this storm event discharge quadrupled in half an hour.”
- “During this storm event there was # kg of sediment transported by the stream.”
- “During this snow melt event # kg of road salt (Cl–) entered the stream.”
- “During this salt flush event anything living in the stream was exposed to # mg/L chloride for # hours”.
Supplemental Sampling and Rating Curves
Instructions for developing the following three rating curves are presented here. These rating curves can then be used together and with the continuous data to calculate quantities of water, sediment (loads), and chloride (loads) moving in the stream over specific periods of time (Section 10.3 and Appendix I).
- A discharge rating curve that relates sensor depth (mm, measured by CTD sensor) to discharge (m3/s; Figure 10.3).
- A TSS versus turbidity rating curve (Figure 10.4).
- A Cl– versus conductivity rating curve (Figure 10.5).
Discharge Rating Curve- Measuring Discharge
Discharge (Q, m3/s) is calculated as flow velocity (V, m/s) multiplied by wetted cross-sectional area (distance x depth = Area, m2): Q = V x A. In wadeable conditions, measurements of flow velocity and wetted area are collected using a flow meter or timed neutrally buoyant object float (velocity), meter tape (distance), and survey rod (depth). Detailed descriptions of these methods are in Appendix G.
Discharge data should be recorded on the Stream Discharge Data Form, field notebook, or other data form, and entered into the Discharge Rating Curve Calculator spreadsheet to calculate final discharge (Appendix H). These discharge data are matched with sensor depth data (same date and time) to develop the discharge rating curve.
In unwadeable conditions full cross-channel discharge measurements and wetted cross-sectional area measurements are not possible without expensive equipment. In these cases cross section wetted area can be estimated by mapping the shape/dimensions of the confined stream channel at baseflow and then using this information along with staff gauge and/or sensor depth measurements to predict cross section wetted area during unwadeable flow conditions. A method for delineating the channel cross section profile is provided in Appendix G. Once the channel cross section has been delineated, the staff gauge water level or sensor depth can be entered into the Stage To Area Predictor spreadsheet to generate a Predicted Wetted Cross-Section Area (Appendix G). This Predicted Wetted Cross-Section Area can then be entered into the Discharge Rating Curve Calculator spreadsheet (Appendix G “Predicting Discharge” and Appendix H) with an estimate of velocity (measured with a neutral buoyant float or velocity meter) to estimate discharge.
Discharge Rating Curve- Generating the Discharge Rating Curve
Final discharge along with sensor depth (at the time discharge was measured) can then be incorporated into a site-specific hydrologic rating curve using the Discharge Rating Curve Calculator (e.g., Figure 10.3; see Appendix H and Instructions in the Discharge Rating Curve Calculator itself for detailed instructions). To develop a preliminary hydrologic rating curve it is recommended that at least five discharge measurements be made during stream conditions ranging from baseflow to the extreme stormflow. Multiple measurements can be taken during a single storm as flow rises and falls. However, it will be necessary to measure discharge in multiple storms as measuring the full range of flows within the confined channel is usually not possible during a single storm. There is no defined endpoint for when a rating curve is complete although statistical measures of the correlation coefficient and significance value can be used as a guide. Because timing of these measurements is crucial (i.e., measurements need to be taken across a range of high flow conditions, which are often short-lived) the real-time sensor data should be used along with weather predictions and other tools (e.g., nearby USGS gauge data, first-hand knowledge) to plan for visits to conduct discharge measurements.
To develop the discharge rating curve it is necessary to measure discharge near the sensor station over a range of flows (i.e., from baseflow to stormflow; Figures 10.4 and 10.5). Each time discharge is measured at a site, data should be recorded on the Stream Discharge Data Form or in a field notebook. Staff gauge height and sensor depth at the start and at the end of when measurements were taken should always be recorded. Staff gauge depth is considered the master reference and sensor depth is then related to this (sensor offset, Figure 10.6).
Discharge measurements (flow velocity, distance, and water depth), as well as staff gauge level, sensor depth and other descriptive information for the Stream Discharge Data Form are then entered into the Discharge Rating Curve Calculator, where these data are automatically incorporated into a rating curve. The equation that represents the relationship of sensor depth to discharge (i.e., the finished rating curve) can then be used along with continuous data to calculate loads using the Load Calculator spreadsheet (Appendix G).
Collecting Grab Samples for Rating Curves
A “grab sample” is a quantity of water generally collected in a field-safe plastic bottle that is then analyzed in a lab for any number of parameters. Grab samples can be collected for any number of situational monitoring intentions. For purposes of TSS versus turbidity and Cl versus conductivity rating curve development, grab samples should be collected across the range of observed turbidity and conductivity values. For TSS versus turbidity this will mean collecting samples from during storms and low flow periods. For conductivity/Cl this will mean collecting samples when conductivity increases due to snow melt and/or rain events that carry salt into a stream, i.e., grab sample collection is timed to coincide with higher than normal conductivity levels.
Grab samples can then be processed by any number of labs for TSS and Cl, as well as many other parameters. Lab certifications, sample processing time, etc. should be assessed prior to hiring a lab to perform the required analyses to ensure resulting data are of the desired level of accuracy and precision. Examples of grab sample materials and shipping protocols are provided in Appendix L.
The basic process for collecting and shipping grab samples is:
- Clean the sensors. It is imperative that the sensors are producing accurate data when grab samples are collected, so that sensor data can be matched up with the grab sample data for developing the rating curves. In high water (when grab samples are often collected) an extension for the sensor brush may be needed so that sensors can be reached for cleaning.
- Collect grab sample (see below).
- Complete grab sample, label with date and time. Make sure that time represents exact time when sample was collected.
- Complete Field Visit Data Form. Make sure to complete the Grab Sample Information section on the back side of the data form.
- Make sure that date and time are exactly the same as listed on the grab sample label.
- Place sample on ice/ice-pack or in refrigerator.
- Prepare sample for shipping/delivery.
- Complete any lab-required chain-of-custody forms.
- Chill and package sample(s) according to lab protocols.
- Ship or deliver sample to lab.
- Note that some labs do not accept shipments on weekends. Sample collection and/or shipping should be adjusted accordingly.
Steps may vary depending on the analyte and the lab but the basic process for collecting a grab sample to be analyzed for TSS and Chloride is as follows:
- Rinse grab sample bottle (Figure 10.7) three times with typical stream water before collecting sample.
- Fill bottle with water that is characteristic of current conditions (i.e., make sure it is from the main flow and that it does not contain material churned up from people walking in the stream).
- Facing into the current, invert the bottle and submerge it to ⅔ total depth of the water (i.e., mouth of bottle should be facing down and bottle should not fill as you submerge it).
- Turn the bottle right side up and begin raising the bottle through the water column at an even pace so that when the bottle reaches the surface it is filled.
- Cap the bottle making sure that the cap is clean (rinse it with stream water if need be).
- Fill in any header information on the grab sample label (e.g., site ID, stream name, date, and time).
- Complete the grab sample bottle label with “Date” and “Time.” Record exact date/time (to the minute) that grab sample bottle was filled with water. This information is key when grab sample data are being used for rating curve development; the exact time is necessary so that sample data can be matched to the exact sensor station readings from the same time.
- Place sample immediately on ice (or ice pack) in a cooler until shipping.
- If Field Visit Data Form is being used, complete the Grab Sample Information section on the back side of this form, Figure 10.8). Make sure that date and time are the same as on the grab sample bottle label.
Generating the TSS and Chloride Rating Curves
In the context here grab sample data are used for developing rating curves for TSS versus turbidity and Cl versus conductivity (Figures 10.9 and 10.10). As with the hydrologic rating curve it is recommended that at least five samples be collected to develop each of the rating curves; however, additional samples will likely be necessary to fully articulate the rating curves.
Depending on the intention for the EnviroDIY Sensor Station data, it may make sense to develop both the TSS versus turbidity and the chloride versus conductivity curves or just one of them (e.g., if road salt input is not an issue, developing the chloride/conductivity curve would not make sense). Turbidity and conductivity ranges will be specific to the stream, so an understanding of these ranges should be established prior to collecting the grab samples. Using historic EnviroDIY Sensor Station data, the high and low points can be identified for focusing grab sample efforts across the range.
As with the hydrologic rating curve, the grab samples can be collected from a single storm or more likely from multiple storms but, as mentioned, should be collected from across the range of observed low to high turbidity and conductivity values (Figures 10.11 and 10.12). For developing the Cl– versus conductivity rating curve, samples will likely need to be taken during winter time salt flushes from rain or snow melt events (Figure 10.13). These values are then plotted against sensor data at the time grab samples were collected and the linear regression equation is generated (Figures 10.8 and 10.9) for use in final calculation of material loads (see Section 10.3 and Appendix I).
To develop the final rating curve for TSS/turbidity or chloride/conductivity grab sample data should be graphed in a 1:1 scatterplot (e.g., Figures 10.9 and 10.10). To do this:
- Acquire grab sample lab analysis data.
- Match grab sample data points up with sensor data points by choosing sensor data points that are closest in time (exact minute) to when each grab sample was collected (exact minute).
- Add grab sample and sensor data points to Excel spreadsheet; sensor data in one column, grab sample data in another column (Tables 10.1 and 10.2)
- Plot the data in a scatterplot, fit a linear regression line to the scatterplot, and display regression equation (Figures 10.9 and 10.10).
- Use this equation along with the continuous sensor data in the Load Calculator spreadsheet.
Once a rating curve is developed, the equations describing the relationships between sensor depth and discharge, turbidity and TSS, and/or conductivity and chloride should be entered into the Load Calculator spreadsheet. The continuous sensor data for the time period of investigation (e.g., a single storm, a season, etc.) should also be imported into the Load Calculator. Once the rating curve equations and continuous sensor data (from a specific time range) are entered, the Load Calculator will automatically apply the rating curve equations to the sensor data and generate continuous data for the parameter that is not being directly monitored by the EnviroDIY Sensor Station (i.e., discharge, TSS, and chloride). Once estimated continuous discharge, TSS, and/or chloride data are generated, the Load Calculator will automatically calculate fluxes of the material moving in the stream over time and sum these amounts to arrive at final loads.
Visit https://www.envirodiy.org/mayfly-sensor-station-manual/appendices/ to view, download, and/or print the appendices.
Decagon Devices, Inc. 2016. Decagon CTD-10 sensor manual (Meter Hydros 21, as of 2018): https://manuals.decagon.com/Manuals/13869_CTD_Web.pdf. Version: October 25, 2016
Campbell Scientific, Inc. 2017. OBS-3+ and OBS300 Suspended Solids and Turbidity Monitors. Revision: 4/17. Campbel OBS-3+ Turbidity sensor manual: https://s.campbellsci.com/documents/us/manuals/obs-3+.pdf
USGS. 1982. Rantz, S.E., U.S. Geological Survey, Water Supply Paper 2175: Measurement and Computation of Streamflow. Available: https://pubs.usgs.gov/wsp/wsp2175/
This manual was written by David Bressler, Sara Damiano, Scott Ensign, Shannon Hicks, Rachel Johnson, and Tara Muenz at Stroud Water Research Center.
This publication was developed under Assistance Agreement NE – 83675001 awarded by the U.S. Environmental Protection Agency. It has not been formally reviewed by EPA. The views expressed in this document are solely those of Stroud Water Research Center and EPA does not endorse any products or commercial services mentioned in this publication.
The opinions expressed in this report are those of the author(s) and do not necessarily reflect the views of the William Penn Foundation, grant number 158-15.
Special thanks to Alliance for Aquatic Resource Monitoring (ALLARM) at Dickinson College for helpful comments on an early draft of this manual.