(Warning: this post contains a lot of words, and fewer scenic images.)
One of the primary missions of the Stream Team, and what consumes most of our effort, is running a network of seventeen stream gauges. The gauges are located in the Taylor Valley, the Wright Valley and the Miers Valley.
Our stream gauges measure stream flow and record the results using a datalogger (a specialized computer). More correctly, the gauge records the height of water in the stream, called the stage of the stream, and that stage value is later converted into streamflow using a mathematical relationship between stage and flow called a rating. A big part of our work is gathering data that are used to define and refine the relationship between stage and flow.
Streamflow is the rate at which water moves past a point, expressed as a volume over time. In the United States, streamflow is usually expressed in cubic feet per second, while in the rest of the world streamflow is usually expressed in the metric system, as liters per second or as cubic meters per second. You can measure the flow of water directly by keeping track of the time it takes to fill a bucket—a typical ½-inch garden hose will fill a five-gallon bucket in about 50 seconds. That is a flow of six gallons per minute, or 6 gpm. That is equivalent to about 0.4 liters per second (l/s) or about 0.014 cubic feet per second (CFS).
But buckets are not practical for measuring large flows, and even for low flows there is often no way to get the water to flow into the bucket (unless there is a convenient little waterfall.) So, streamflow is most often estimated by measuring the stage of the stream as it flows over or through some change in the channel, often a man-made ledge, called a control, or through a specially shaped flume (like the Baski, I wrote about earlier). The stage is related to flow by the rating, in the form of a chart, table or equation. Except for special types of controls and flumes, a rating table or curve cannot be developed on first principles, so a process of calibration is used, where spot measurements of flow using manual methods are combined with stage measurements from the gauge to construct the rating for that gauge. (I’m going to describe the calibration process in a separate post.)
So, estimating flow involves four elements: First you find or construct a control or install a flume; then you maintain an ongoing series of measurements of stage; at the same time you conduct an ongoing program of calibration to construct and refine a rating; and finally you use the rating to calculate estimates of flow from your measurements of stage. I’ll describe the process of calibration in another post. Below I am going to describe the physical infrastructure of the control and measurement equipment, which we refer to collectively as a gauge.
Our longest, and one of our oldest controls is at the F3 Gauge on Lost Seal Creek, shown at the top of this post. Lost Seal Creek is so named because of a mummified seal that rests just a couple hundred meters from the F3 Gauge
The F3 control, like all of our controls, is essentially a dam made out of sandbags (at some locations we pile up rocks and cover them with nylon tarp). The black thing in the middle is a Parshall flume, which is intended to measure the lowest flows.
To the left of the Parshall flume is a lower section in the control, just a space where the top layer of sandbags was left off. This section facilitates measuring a middle range of flows. Water starts to flow through it when the level of water backing up behind the flume reaches the lowest level of the lowered section.
As water rises higher (at higher flow rates) it will begin to flow over other areas of the control. Because the control is not perfectly level water may flow through several areas. Eventually, at very high flows, water will pass over broad areas of the control. In the photo below water is flowing through all three parts of the control, the flume, the low-flow section and the broad crest of the weir to the right of the flume. In fact, though it is not apparent in this photo, water is flowing to the right of the control. Note that the depth of water is almost two feet higher than when there is no flow or very low flow. It is this change in stage that is measured by the devices in the gauge box.
The essential function of the control is to provide a reliable and consistent relationship between the elevation of the water behind the control, the stage, and the amount of flow over the control, i.e. a stable rating. Unless the control is damaged, modified, distorted by heaving or settling, or silted in, the relationship between the depth of water behind the control and the flow over the control is reliable and repeatable. One of our tasks at the start and end of the season is to survey the elevation of critical parts of the control and the measuring equipment to detect and allow for correction of changes due to freezing and thawing of the soil and permafrost foundation of the control
The Lost Seal Gauge was constructed in 1990, along with several other gauges on streams emptying into Lakes Fryxell, Bonney and Hoare. Over the years a few gauges have been abandoned due to difficulties with their maintenance and new gauges have been installed. Currently, there are 17 gauges in our network, including some in the Wright Valley and the Miers Valley.
The equipment that measures and records stage at our gauges is mounted in a plywood box located close to the control. That is the orange box in the foreground of the Lost Seal control in the photo below.
Inside the box is a steel cylinder of high-pressure nitrogen connected through a regulator to a conoflow, a device that meters out that nitrogen at a very slow rate. The rate of flow of nitrogen coming out of the conoflow amounts to a bubble the size of a peppercorn every second or so. From the conoflow a plastic tube runs from the box to the bottom of the stream, just upstream of the control and at an elevation just below the lowest point on the control, called the point of zero flow, or PZF. As the nitrogen bubbles out of the tube, through an orifice, it is resisted by the pressure of the water above it. That resisting pressure is proportional to the stage. It is this pressure that is converted into an electrical signal by a sensitive pressure sensor called a transducer and recorded by the datalogger, both located in the orange box.
Our work proceeds in three phases as the season progresses. First we open the gauges to get them ready for flow. We do this at the start of the season, while everything is frozen and before any flow occurs. Then, while the streams are flowing, we take measurements to calibrate the sensors in the box and the rating curve for the gauge. Finally, as the season draws to an end, we will close the gauges and prepare them for the winter.
When we open a gauge we do three things: we install a new bottle of compressed nitrogen and check the gas system for leaks; we pull the storage module (where the stage data are recorded) off the datalogger and install a new one; and we survey the reference points at the control. These gauges are located on top of the active zone, soil that freezes in the winter and thaws in the summer. That freezing and thawing can cause the control and the orifice to shift relative to each other, which will change the rating. If the control has heaved or settled the relationship between depth and flow rate will change. We survey the elevation of specific reference points on the control and at the elevation of the orifice in order to know if the structure has shifted.
The last part of the flow measurement system is the datalogger, an electronic apparatus that records the electrical signals from the pressure sensor on what is essentially computer memory. The technology used in our dataloggers is decades old, and appears archaic in today’s world of super-computing cell phones with 32 GB micro-SD chips, but it works, and it works reliably at all sorts of temperatures. At the start of the season we make sure that the datalogger is working properly, and that its calendar and clock are correct, and then we replace the data storage module. We take the “old” storage module back to F6 and download its contents onto the computer, and from there back to Boulder over the internet. (Some gauges have telemetry equipment that also transmits data back to Boulder as they come in.)
The old storage module contains data starting at the end of last season, when the gauge was last visited and closed by last year’s team in late January. The data generally do not extend too long after the gauge is closed since the nitrogen will run out and the datalogger battery may run down when the sun goes down (the batteries are recharged by solar cells, which can be seen on the boxes in some of the photos) but this does not matter because flow stops and the streams freeze solid in February or perhaps March.
At the end of the season we make a last visit to each gauge. At that time we pull the storage module we installed at the start of the season and install a new one. That newly installed storage module will be the “old” storage mogule pulled when the team that comes down in the following November opens the gauge. We put a super-strong cargo strap around the box and cinch it tight, so that there is no possibility that the wind could catch one of the doors and open it up. Finally, we do one more survey of the control reference points and the orifice. Then we say goodbye to the gauge and leave it on its own for the winter, for the next team to open in November.