Currently browsing category

stream temperature

Stormwater control measures modify event-based stream temperature dynamics in urbanized headwaters

Next week, the Watershed Hydrology Lab will be well represented at the CUAHSI 2014 Biennial Colloquium. We’ll be presenting four posters, so here come the abstracts…

Stormwater control measures modify event-based stream temperature dynamics in urbanized headwaters

Grace Garner1, Anne Jefferson2*, Sara McMillan3, Colin Bell4 and David M. Hannah1
1School of Geography, Earth and Environmental Sciences, University of Birmingham, Edgbaston, Birmingham, B15 2TT, UK.
2Department of Geology, Kent State University, Kent, OH, 44240, USA
3Department of Civil and Environmental Engineering, University of North Carolina at Charlotte, Charlotte, NC, 28223, USA
4Department of Infrastructure and Environmental System, University of North Carolina at Charlotte, Charlotte, NC, 28223, USA

Urbanization is a widespread and growing cause of hydrological changes and ecological impairment in headwater streams. Stream temperature is an important control on physical, chemical and ecological processes, and is an often neglected water quality variable, such that the effects of urban land use and stormwater management on stream temperature are poorly constrained. Our work aims to identify the influence of stormwater control measures (SCMs) of differing design and location within the watershed on the event-based temperature response of urban streams to precipitation in the North Carolina Piedmont, in order to improve prediction and management of urban impacts. Stream temperature was measured within SCMs, and upstream and downstream of them in two streams between June and September 2012 and 2013. Approximately 60 precipitation events occurred during that period. To unambiguously identify temperature increases resulting from precipitation, surges were identified as a rise in water temperature of ?0.2°C between the hours of 15:30 and 5:30, when the diurnal temperature cycle is either decreasing or static on days without precipitation. Surges up to 5°C were identified in response to precipitation events, with surges occurring both upstream and downstream of the SCM under some conditions. Surges were also recorded within the SCMs, confirming that temperature surges are the result of heated urban runoff. Classification tree modeling was used to evaluate the influence of hydrometeorological drivers on the generation and magnitude of temperature surges. In both streams, event precipitation, antecedent precipitation, and air temperature range were identified as the drivers of whether or not a surge was observed and how large the surge was, though the order and thresholds of these variables differed between the two sites. In a stream with an off-line, pond SCM, the presence of the pond in the lower 10% of the watershed did not affect the magnitude of temperature surges within the stream, but the pond itself had a wider range of surge magnitudes than did the stream. In a watershed with a large in-line pond, and a downstream contributing wetland SCM receiving flow from 40% of the watershed, the wetland increased both the frequency and magnitude of temperature surges observed in the stream. Our results suggest dynamic hydrometeorological conditions, SCM design, and position within a watershed all influence whether stormwater management reduces or enhances temperature surges observed within urban headwater streams, and that these factors should be considered in the recommendations for urban stormwater management systems.

Getting good stream temperature measurements without losing your probes

Tidbits temperature probeNote: I use stream temperature to understand groundwater-stream interactions and the response of streams to urbanization. Since ~2004, my stream temperature probe of choice has been the Tidbits temp probe, manufactured by Onset corporation. I like them because they are +/-0.2C and extremely durable, watertight, and reliable. Plus, I’ve had good customer service experiences with the manufacturer. What follows is my attempt to explain how I deploy them in the field, based on my cumulative experience and what I’ve learned from others. Please comment and add your own ideas and experiences, and I’ll amend the protocol as needed.

Getting ready for the field

  1. Obtain Tidbits temperature probes and the associated HoboWare Pro software. Read the documentation and learn how they work.
  2. Using the delayed start feature in Hoboware, set all of the temperature probes to start at the same time and at the same sampling interval. I like to set them to start evenly on the hour. It makes analysis easier later.
  3. You can’t change the calibration of the temperature probes within the software, and they should come pre-calibrated, but you should still check the calibration of your temperature probes relative to a certified thermometer and to each other. I recommend a 3 stage calibration check process, but you’ll want to do at least 2 temperatures that bracket the range of range conditions you expect to measure. You need to do each of these for a couple of hours, because while the response time in water is ~5 minutes, it is slower in air.
    • An ice bath (with stirring) or the refrigerator.
    • Room temperature, out of direct light, in a room with fairly stable temperatures for a couple of hours.
    • Depending on what temperature your streams are likely to be, you might want a temperature intermediate between the refrigerator and room temperature. (I’d love to hear your suggestions for an easy, good intermediate cool temperature.)
    • Or, if you are interested in summer headwater stream temperatures, you could use something like a consistenly shady area outside. I’ve also used my backpack, by putting all of the probes in the same container inside it, and then hiking around with them for several hours prior to installation.
  4. Download the Tidbits after the calibration check, and reset them for a simultaneous start on the day you’ll be deploying them in the field. I’ve used a 15 minute interval for projects where daily and seasonal fluctuations were of interest; but since we are now interested in storm response, I think we should set them to log at 5 minute intervals (in Celsius, please!).
  5. If there are extra tidbits available, I recommend deploying one in the air, in a shady area near the stream at each field site. I’ve hung them from a tree branch with fishing line, and a homebuilt radiation shield. My radiation shield was a gallon milk jug with the bottom cut off. The tidbit fit through the top opening, and then I screwed the top back on, so that the Tidbit hung freely within an area shaded on the top and sides by the milk jug.

Selecting your field site

There are several very important things to consider when selecting your probe site. You are probably going to have to compromise somewhere in this list at some of your sites, but this is what to strive for.

  1. It meets your scientific objectives (i.e., is positioned appropriately relative to a stormwater BMP, restoration structure, tributary junction, or other field sampling/equipment site.)
  2. The probe will be under flowing water under a wide range of flow conditions. Good places include the channel thalweg or a pool that will not go stagnant (e.g., below a rock outcrop or structure that directs all streamflow into the pool).
  3. The probe will out of direct sunlight at all times of day, as best as possible. Deep shade, an overhanging bank, or an incised reach is good. Peak water temperatures occur in the mid- to late-afternoon, so this is the most important time to check and make sure your site is out of the sun. Adding a cobble on top of your probe, without completely burying the probe in the streambed, is another good way to keep the sun off of it (and to make it less likely to be discovered or banged up during high flow). If you think sun exposure is likely to be a problem (or your data suggest that it is), you should take measurements of shading with a densiometer. Measuring shading won’t fix the problem, but at least you’ll be able to discuss it.
  4. The probe placement is as geomorphically consistent with other probes in the project as possible.
  5. The probe can be discretely and securely attach the probe to something very stable. I’ve almost always used streamside trees, but a post holding other equipment would work too.
  6. The probe should be located somewhere it is possible to bring it up onto the bank while still cabled, so that the Tidbit can be downloaded into the laptop without having to balance the laptop in the middle of the stream.

Deploying the probe

  1. Loop steel cable through the hole on the Tidbit, and crimp the loop shut with a hand swager (like this one). I have cabling, crimps, a swager, and a cable cutter available in my lab.
  2. Measure out an appropriate length of cable to reach the secure attachment site, loop around it, and cut and crimp the cable. I like to give the cable enough room so that it can lie flush with the stream bed and bank and let the probe be in the thalweg, under a rock, but I try not to give it too much slack to get caught on things or let the probe go banging down the stream if it gets dislodged. And, of course, I never make a loop around a tree very tight
  3. Put the probe in the stream. If possible, place a cobble on top of it so that water flows under the cobble and the probe doesn’t get smooshed into the streambed.
  4. Mark your field site with (1) GPS coordinates, (2) discrete flagging or a stake, (3) write down really good field notes describing the site and how you got there, and (4) take photos of everything (like the ones below). Write your field notes so that your advisor(s) can find the site 2 years from now without your help. (Thanks!)

Note: We have tried a variety of methods for securely attaching the Tidbits temperature probes to a fixed object. Rope gets abraded, degraded, and eventually breaks in high turbulence and velocity flows. High test fishing line broke as well during a high flow in a first order stream. We have settled on steel cable, thin enough to thread through the hole of the Tidbits and secured by crimping, as shown below. Recently, we discovered that several of the cables that had been deployed for ~2 years had rusted and broken and that we’d lost the temperature probes at some point since we’d last downloaded their data. I’ve now heard that some people are using plastic coated steel wire. Maybe we should consider that as an alternative to the unocated cable.

I still believe that the steel cabling is a good attachment method, but our experience reminds me of the importance of regularly checking on field equipment. Even if the temperature probe can collect a year’s worth of data before its memory fills up, I’d recommend downloading the data at least once every 3 months (in a non-flashy stream) and doing a thorough check of the cable integrity each time. In urban streams, I now recommend downloading table and checking cable integrity every 2 weeks. Data from a lost probe can never be recovered.


Attachment for temperature probe at Deep Creek site DC 12, during the fishing line-era of installation. The flagging was also labeled with a project and site identifier.

The Tidbits is under the triangular rock at the center of the photo.

The Tidbits is under the triangular rock at the center of the photo. You might be able to see the line extending to it. Good points about this site: A rock protector with good water flow under and around it and good shading. Bad points: Not a lot of flow depth here, but we were at the seep initiating this stream. We had an air temperature probe at this site as well, and the water temperature was always significantly muted relative to the air temperature fluctuations.

Thanks to Sarah Lewis for adding her wise comments to via email. She taught me a lot of this stuff in the first place!