Desert STORMS: Nevada Mercury Flux Intercomparison Workshop

In September 1997, researchers from around the world gathered in Nevada to monitor natural mercury emissions from enriched desert soils. Tekran was there. In fact, virtually all of the research groups brought at least one Tekran Model 2537A analyzer to the study site. Some brought several. Representatives from Tekran were present as observers. We are always interested in what our customers are doing with our boxes, and meetings like this provide the best opportunities for customer feedback.

The study was given the name: Nevada STORMS (Desert Study and Tests Of the Release of Mercury from Soils), in honour of the unusual weather encountered.

Purpose

The purpose of this workshop was to give scientists the opportunity to compare different methods for measuring mercury fluxes. In addition to field studies performed during the day, evening sessions discussed the pros and cons of the major methods used to measure flux and how flux readings can be used to provide inputs to global scale mercury cycling models. The study was sponsored by EPRI, (the Electric Power Research Institute) and was organized and hosted by Dr. Mae Gustin, of the University of Nevada-Reno (UNR).

What are fluxes and why are they important?

Mercury fluxes are measurements of whether mercury is being emitted from, or absorbed onto, a particular surface. Those not specializing in mercury transport are often surprised to find out that large amounts of mercury in the atmosphere can be absorbed by, or released from, soils, vegetation, lakes and oceans. Since mercury is an element, it cannot be destroyed and once available to the biosphere will continue to cycle between these media for extended periods.

The problem of defining mercury sources and sinks on a global scale cannot begin to be studied until one has information on the exchange mechanisms at work between the major of earth's surfaces: land, vegetation and water.

Fluxes are expressed as an emission rate per unit of surface area, generally in units of ng/m2/hour. By convention positive values imply release to the atmosphere by the underlying medium and negative readings imply removal of atmospheric mercury to the surface.

How does one measure a flux?

There are two major approaches to measuring fluxes. There is considerable variation within each approach and part of the study's intent was to determine if this wealth of different designs gave equivalent results. The descriptions below (like everything else on this page) are simplifications of what is actually going on and are intended for the non specialist. The two approaches are not directly comparable: One measures the flux from a small area and is invasive, whereas the other determines the average flux over an area and is non invasive.

Background

Remember that in most locales, the fluxes being measured are extremely small. This requires great precision in the mercury measurements made. Prior to the advent of automated instrumentation, manual methods such as those pioneered by Dr. Steven Lindberg of Oak Ridge National Laboratory, typically required six replicates of each manual gold cartridge measurement to attain the required precision. Since a flux calculation requires a minimum of two measurements, at least twelve cartridges must be analyzed to obtain one flux measurement. It takes hours of sampling time to make the measurement and further hours in a laboratory to perform the analyses. The situation was made worse when it was realized that mercury fluxes can vary rapidly over time, being dependent such factors as amount of sunlight, temperature, and moisture.

The Model 2537A is capable of extremely rapid and precise measurements, making it ideal for flux monitoring. It came as a welcome relief to those working in the field and has turned out to be one of the major research applications for the analyzer.

Flux Chambers

A flux chamber is basically a box with an open bottom that is placed over the surface to be measured. The box has an inlet and an outlet. Ambient air is drawn into the chamber by a pump. The exit air is analyzed for mercury and compared with the inlet (ambient) air. If the exit is higher than ambient, mercury is being emitted by the soil. If it is less, then mercury is being absorbed into the surface. Simple ... in principle anyway. If you know the flow rate through the chamber and the surface area covered by it, the flux can be calculated.

Flux chamber techniques are very sensitive and can measure very small positive or negative fluxes. They have a defined footprint so the region being measured is always known. These advantages are counterbalanced by a few problems. Covering the surface in question is likely to interfere with the normal air exchange processes, resulting in biased or erroneous readings. Because the footprint is small, the method is not capable of measuring large areas of land since the number of sampling points required would be too large to be practical.

Ideally, the inlets and outlets should be measured continuously and simultaneously. In practise, many users of Tekran instruments use a sequential sampling arrangement such as the Model 1110 Two Port Synchronized Sampler. This monitors the inlet for a period of time and then switches over to monitor the outlet. The time scales are sufficiently short that the errors introduced are negligible.

Meteorological Techniques

There are several micrometeorological techniques that estimate mercury fluxes over an area (known as the 'fetch') that is upwind of the measurement point. Once again the principle is simple.

One tries to determine the mercury levels close to the ground and at a slightly higher elevation. If there is more mercury in the lowest ground air layer, the air is being enriched by mercury emissions from the surface. If the ground level concentration is lower, mercury is being depleted by absorption into the surface. Comprehensive monitoring of three dimensional wind fields around the sampling point is required to provide enough information to calculate the flux.

Micro-meteorological techniques have several advantages. They provide fluxes over a fairly wide area. They do not impact the site being measured.

Disadvantages are that these methods are inherently about an order of magnitude less sensitive than flux chamber techniques. The actual area being monitored varies with the prevailing winds and is difficult to determine. The method only works under some types of weather and wind conditions.

Conduct of Survey

The survey was conducted from September 1st to September 4th, 1997 on a site south of Reno, Nevada, known locally as Steamboat Springs. The surrounding region is famous for valuable mineral deposits. Those famous wild-west mining towns, Virginia City and Carson City are both within 30 miles of the site. The location was chosen because the soils were known to be enriched in naturally occurring mercury.

The flowery descriptions above do not prepare one for the reality of field work in the Nevada desert. Conditions are not exactly ideal. Neither researchers nor equipment are designed to function under an environment that includes nighttime temperatures of less than 5 degrees Celsius, daytime highs of well over 30 degrees in the shade, lots of direct sun, and liberal amounts of blowing sand and dust. Fixed shelters and air conditioning were not available and electrical power was supplied by generators.

The good thing about a desert is that it almost never rains, so simple open tents are acceptable as shelter from the elements. It hadn't rained in over two months before the study began. Naturally, it rained during the study. (More on this later.)

Lest one think that the study consisted of leisurely daytime monitoring activities followed by evening fun and games (this was Nevada after all), Mae made certain that we all had evening seminars and presentations to attend.

Study Participants

Mercury researchers from several countries availed themselves of this unique opportunity to compare their measurement methodologies with those of their colleagues. The participants are listed by the tents they occupied during the survey. (There were usually several teams per tent.)

Organization & Participants	Flux Chamber?	Micromet Method?	No. of Tekran Analyzers Used
GKSS, Germany -Ralf Ebinghaus, Hans Kock	Yes	No	2
Chalmers University, Sweden -Zifan Xiao, Jonas Sommar	Yes	No	-
Frontier Geosciences -Ralph Turner, Dirk Walschlager, Jacqueline London	Yes	No	1
Atmospheric Environment Service, Toronto, Geological Survey of Canada, -Pat Rasmussen University of Guelph, Ontario -Grant Edwards, Jeff Kemp, Colleen Fitzgerald-Patrick	Yes	Yes	3
Atmospheric Environment Service, Quebec -Laurier Poissant, Alain Casimir, Martin Pilot	Yes	Yes	2
Oak Ridge National Labs -Steve Lindberg, Jim Owens	Yes	No	1
University of Michigan -Alan Vette, Frank Marsik	Yes	Yes	2*
University of Nevada, Reno -Mae Gustin United States Geological Survey -Mark Hunerlack, Mark Majewski	No	Yes	1*

Observers:

Electric Power Research Institute
-Don Porcella, Mary Ann Allan

Atmospheric & Environmental Research Inc.
-Prasad Pai

Atmospheric Environment Service, Atlantic Region
-Steve Beauchamp, Rob Tordon, Faisal Boudala

Yale University
-Xuhui Lee

Tekran Inc.
-Frank Schaedlich, Dan Schneeberger

Minerals Exploration & Environmental Geochemistry
-Shea Clark Smith

* The University of Michigan received its two analyzers midway through the study. So much for guaranteed two day delivery by a major courier who shall remain nameless. A loaner instrument destined for the University of Nevada did not make it to the survey site at all due to problems with shipping.

Results

Full results will be presented in an upcoming special issue of a major, peer reviewed journal. In the meantime, the following general findings can be divulged.

1) There is lots of naturally occurring mercury at this particular site. This was not particularly surprising since this is why the site was chosen.

2) Fluxes were high, particularly during daylight hours. At night, ambient air concentrations usually rose markedly, but fluxes declined.

3) When the improbable happened and it rained during the survey, it got incredibly cold and windy. It rained hard. Real hard. We would have taken pictures, but we were all too busy hanging onto the tents to prevent them from being blown away. Although the storm was not high on anyone's list of pleasant events, it was serendipitous since some rather surprising things occurred. Mercury emitted from the sand skyrocketed when it got wet. Higher than most of us expected. The fluxes skyrocketed and stayed high for hours.

4) All of the Tekran analyzers performed excellently during the study. (It's hard to be objective here.) As manufacturers, we cannot recommend operation under these conditions on a regular basis, but it's comforting for researchers to know that the Model 2537A works under virtually all conditions. Several groups used the Model 1110 Two Port Synchronized Sampler as a switcher for their flux chambers. It worked well and saved the cost of a second analyzer.

Some preliminary results are available. Our thanks to Oak Ridge National Labs and to the University of Michigan for having draft data available so soon.

Acknowledgements

Tekran would like to thank Don and Mary Ann from EPRI for their support in organizing this intercomparison. In particular, Tekran and all of the participants want to thank Mae Gustin and her students for their tireless work in organizing every aspect of this undertaking. They went to great lengths in planning the event and in resolving any difficulties that arose. (I mean, how often have you received a menu for your choice of box lunch months before you went on survey in the desert?) The rainstorm was about the only unplanned event and I'm starting to wonder ...