Research Spotlight: Erika Smull

Analysis of Rhodamine– WT and sodium chloride as stream tracers 

Conservative tracers help to understand solute transport processes such as advection, dispersion, and transient storage. Two popular stream tracers are Rhodamine- WT (RWT) and sodium chloride (NaCl). NaCl is considered to be conservative, yet it cannot be detected at very low concentrations. RWT, on the other hand, can be detected at very low concentrations (<0.1 ppb), yet it is known to sorb to organic materials and photo-degrade. I am studying the limitations and advantages of both RWT and NaCl as applied to large river slug injections. Our research team seeks to analyze the window of detection for each tracer (time of tracer arrival to time of tracer non-detection), as well as the late-time tailing behavior of each.

Dual slug injection

RWT visible downstream

For the past few weeks, I have been working to complete various combined slug additions of RWT and NaCl along a 1.5-kilometer reach on the Kuparuk River, an Arctic river located on Alaska’s North Slope. Both tracers are simply mixed together with river water in large buckets until the NaCl completely dissolves. Fluorescence and electrical conductivity (EC) are continuously sampled at the upstream and downstream ends of the reach prior, during, and after the injection, until fluorescence and EC return to ambient (pre-injection) levels. Various flow conditions have been characterized, and I plan to complete additional injections during the rest of my stay at Toolik Field Station. With a larger data set, the results will contribute to experimental design of conservative tracers in large rivers, and weigh the benefits of RWT versus NaCl.

Posted in Arctic research, Research Techniques, stream-goundwater research | Leave a comment

CSASN – July/August 2012


Oksrukuyik Creek

We are back at Toolik Field Station in Northern Alaska to continue work for the Changing Seasonality in Arctic Stream Networks (CSASN) project. Our plan for the next four weeks is to complete several solute (conservative and non-conservative) injections and complete a suite of ecological measurements on several rivers in the vicinity of Toolik Field Station. Check back for more detailed updates from the field over the next few weeks!

Posted in Uncategorized | Leave a comment

Measuring discharge with conservative tracers

Adam N. Wlostowski, Erika M. Smull, Michael N. Gooseff

A quick, simple, and effective way to measure discharge in small headwater streams is to use dilution-gauging methods. Measuring flow with standard wading rod methods can be difficult in small streams, especially under low-flow conditions. With a few kilograms of sodium chloride (NaCl), a bucket, and a conductivity data logger, you’ll be getting accurate discharge measurements in no time!

This post will describe the data collection and data analysis procedure used by our lab group to measure discharge in Alaskan headwater tundra streams. We will begin by outlining the mathematical concepts and methodology needed to measure discharge with conservative tracers. Then, we will illustrate step-by-step field procedures. Finally, we will provide access to an Excel spreadsheet used by our lab group to analyze raw injection data.

1. Background

Measuring the dilution of a known volume of conservative tracer is the basis of dilution-gauging methods. We use a simple “slug injection” of a known mass of dissolved NaCl tracer. A slug is a near instantaneous addition of a highly concentrated conservative tracer solution. As the tracer moves downstream, a concentration-time profile known as a break-through-curve (BTC) is recorded by sampling through time at a single downstream monitoring location.

The discharge measurement is made using the following equation.


where, Q is discharge [M3/T], M is the injectate mass [M], C is concentration [M/L3], and t is time [T].

The denominator term on the right hand side of equation 1 is a simple integration (area under the curve) of the BTC, commonly known as the zeroth moment. Looking only at the units of each term in equation 1, it is easy to see how we arrive at an estimate of discharge [L3/T] (equation 2).


Two large assumptions are made when using equation 1 to estimate discharge from a slug injection. First, we assume that all of the injected mass was recovered at the downstream sampling location. Second, the tracer is completely mixed across the channel when it passes the downstream sampling location. The nature of these two assumptions places emphasis on choosing an acceptable reach length (distance from injection location to sampling location). We don’t want the reach to be too long, where tracer might be removed by flow paths which exit the spatial extent of the reach completely; we also don’t want the reach length to be too short, such that the injection is not well mixed (both laterally and vertically) at the sampling location. We recommend trial and error field experimentation to best establish reach lengths under a particular discharge range.

A far more detailed explanation of the mathematical background can be found in [Kilpatrick and Cobb, 1985]

2. Field Methods

During June 2012, Erika Smull and I measured discharge at many locations along an arctic headwater stream located just south of Toolik Field Station on Alaska’s North Slope. Let’s go through our step-by-step field procedure.

1. Launch conductivity loggers to continuously sample at the downstram location. By continuously sampling conductivity, we can make a simple conversion to NaCl concentration using a laboratory-established relationship. Our group uses Onset HOBO U24-001 Conductivity Loggers, which can be easily launched and downloaded using HOBOware software installed on a field netbook. Because the reach lengths for the injections are relatively short, we use a fine sampling frequency (2 sec) to best characterize the BTC. It is very important to sync the time on the data logger to the time on your wristwatch, or vice-versa.

Erika launches HOBO EC loggers

2. After the logger is launched and logging, tie a bright piece of flagging to the logger (in this case, orange) and position it in the thalwag at the downstream sampling location. Position the logger such that measurements are made at approx 50% of the water depth. Record the river location of the downstream sampling site!

EC logger (orange flagging) positioned in stream

3. Once the downstream site is all set, move upstream to the injection location and mix the injectate. Given the small nature of our stream, we dissolved approx 1 kg of NaCl into approximately 3 gallons of stream water. First, we partially fill a 5-gallon bucket with stream water. Then, we pour a pre-measured mass of NaCl into the bucket, being sure to record the NaCl mass! Using a small PVC stick, stir the bucket contents until all the NaCl is dissolved.

Erika mixes the injectate

4. Now we are ready to inject! Inject the dissolved tracer by quickly pouring the bucket’s contents across the width of the stream. Be sure to record the approximate time of injection.

Erika droppin’ a salt bomb

5. Immediately after pouring the injectate into the stream, rinse both the bucket and the mixing stick in the stream water. This will ensure all the NaCl mass makes it into the stream.

Rinse the mixing stick…

Rinse the bucket….

6. Wait for the injection to completely pass the downstream location and the stream concentration to return to ambient conditions before removing the data logger and downloading its data. Bringing a handheld conductivity logger into the field is an easy way to tell if the stream has returned to ambient conditions following the injection.

Once the data is off the data logger and safely saved, you are ready to return to the lab and process the data. Attached below, you will find an excel spreadsheet, which can be easily used to process raw data and calculate an approximate discharge value.



Kilpatrick, F. A., and E. D. Cobb (1985), Techniques of Water-Resources Investigations of the United States Geological Survey: MEASUREMENT OF DISCHARGE.

Posted in Arctic research, Research Techniques, stream-goundwater research | Leave a comment

Field Technique – installing water level loggers in shallow wells

After well has been established and cleared, one may choose to install a water level logger to record changes in water levels within the wells.  Here I describe a method for doing this that avoids several common issues and problems that are often associated with deploying loggers in wells.  Please feel free to contact me if you have any questions.  The site used below for the pictures that go along with this explanation is in Arctic Alaska with wells that are fully slotted through their length (even above ground surface).  NOTE that the rims of wells are not always flat and horizontal to the ground (the ground is not always uniform around a well for that matter).  Hence, we make a small black mark with a sharpie marker on the top of the well that is our reference point for measurements inside and outside of the well.  This should work fine as long as your well is not free to rotate at all.

Posted in Research Techniques | Leave a comment

CSASN – June 2012 Field Campaign

Elissa and Erika doing rock scrubs on the Kuparuk River near 5.5 km.

We have traveled to the Toolik Field Stationin northern Alaska to prepare our research reach for the Changing Seasonality in Arctic Stream Networks (CSASN) project.  Our entire project group (including collaborators from University of Vermont and University of New Hampshire) will be conducting nutrient uptake experiments in the Kuparuk River in late July-early August.

Posted in Arctic research | Leave a comment

Hands-On Workshop on Stream-Groundwater Techniques Completed

Flyer for our workshop that was circulated widely at the Fall Meeting of the AGU in 2011.

From 12-15 June, we held a workshop to demonstrate several different techniques used currently to quantify stream-groundwater interactions.  We had over 30 attendees, mostly graduate students.  Day 1 was focused on theory behind the techniques, Day 2 was a full field day in which we conducted a stream tracer experiment on Shavers Creek, in central PA; Day 3 was a data and sample analysis day; and Day 4 was devoted to data interpretation.  The group was enthusiastic and dedicated to the workshop.  Drs. Kamini Singha and Michael Gooseff developed this idea and organized much of it, and we had significant and terrific support from co-instructors: Dr. Roy Haggerty, Dr. Adam Ward, Dr. Judson Harvey, Dr. Olaf Cirpka, Ricardo Gonzalez, and Dr. Christine Hatch.  Most importantly, we had extensive and essential assistance from Kayla Berry and Jennifer Arrigio from CUAHSI.  The workshop was financially supported by a NSF grant to Drs. Singha and Gooseff as part of their outreach efforts, and by CUAHSI.

We will soon have posted online video clips of the presentations and other materials from the workshop.  Check back for updates!

All of the workshop participants and instructors at our field site at the end of day 2.

Posted in Research Techniques, stream-goundwater research | Leave a comment

Research Site on Shavers Creek, PA

Our riparian hydrogeophysics project has been focused on two streams running through the drained bed of Lake Perez in central Pennsylvania, about 20 minutes from the Penn State campus.  The lake was drained about 4 years ago.  We have installed on these two reaches over 100 shallow and deep groundwater wells to monitor water table dynamics in response to precipitation events, etc.  We have also deployed fiber optic cables to monitor streambed temperatures along these reaches.

Coring 10 ft deep holes for installation of wells at Shavers Creek.

In mid-June we hired a geoprobe rig and operator to install ~40 deep wells (relatively deep – 10 ft) at the experimental reach on the main stem of Shavers Cr.  We were able to extract ~5 ft cores in this process and we have collected and cataloged those so that we can characterize the stratigraphy of the soils and sediments adjacent to this stream.

We had a lot of help getting this work done and we are very grateful to Erik, Erika, Zach, Emily, Colin, and Ryan for their time, effort, and patience!

Dr. Singha and Zach install fiber optic cable on the streambed to be used with a distributed temperature sensor system to monitor streambed temperatures every meter along the cable at high temporal frequency.

Adam grabs a core from the geoprobe operator and delivers it to our processing team who will label and secure it for transport back to the lab

Posted in hydrogeophysics, stream-goundwater research | Leave a comment