Chinese (Simplified)EnglishFrenchItalianPortugueseSpanish

Ocean Optics
Worldwide Headquarters
Largo, Florida, USA

+1 727-733-2447

[email protected]

Sales, Service
& Support Facility
Duiven, The Netherlands

+31 26-319-0500
+33 442-386-588

[email protected]

Ocean Optics GmbH Sales,
Service & Support Facility
Ostfildern, Germany

+49 711-34-16-96-0

[email protected]

Sales Support
for the
United Kingdom

+44 1865-819922

[email protected]

Sales, Service
& Support Facilities
Shanghai, PRC – Beijing, PRC

+86 21-6295-6600

[email protected]


The Science Behind the Trail by Fire Expedition

DOAS Measurements of SO2 Flux from Volcanoes in the South American Andes

Last fall we told you about a pioneering expedition by a group of geoscientists to measure volatile gases emitted by 15+ high-altitude active volcanoes along the South American Andes. Over a period of four months, these intrepid researchers took their mobile volcano observatory (a converted Land Rover Defender) up mountaineering routes in the first-ever comprehensive study of the Nazca Subduction Zone. By taking both ground-based and aerial measurements of volcanic emissions using state of the art gas monitoring equipment and correlating the results to satellite-based sulfur dioxide (SO2) data, they expect to provide the first estimates of gas composition and flux at a large number of Andean volcanoes, in turn improving knowledge about the impact of these volcanoes on our atmosphere and climate.

Measurement of the flux of SO2 from the volcanoes using a variant of UV absorbance spectroscopy called scanning DOAS (Differential Optical Absorption Spectroscopy) was key to the study, for which Ocean Optics contributed three custom-configured Flame spectrometers. Curious to see how the spectrometers performed and learn more about the science behind the expedition, we caught up with the lead on the project, Dr. Yves Moussallam, not long after his return home to University of California, San Diego.

The ‘Trail by Fire’ team (L-R): Dr. Philipson Bani, Dr. Ian Schipper, Aaron Curtis, Dr. Talfan Barnie, Dr. Nial Peters, and Dr. Yves Moussallam

The ‘Trail by Fire’ team (L-R): Dr. Philipson Bani, Dr. Ian Schipper, Aaron Curtis, Dr. Talfan Barnie, Dr. Nial Peters, and Dr. Yves Moussallam

Ocean Optics: First of all, can you tell us why you wanted to make this expedition?

Moussallam: The atmosphere that allowed our planet to spark and sustain life formed as a result of gases emitted by volcanoes early in the Earth’s history. These volatile elements, mainly water and carbon dioxide, are constantly recycled back into the deep earth at subduction zones – points where tectonic plates sink into the mantle. During this process, the sinking plate is subjected to increasing heat and pressure, releasing volatiles. When added to the mantle, volatiles induce melting and fuel volcanic explosions, completing the cycle. We’ve had this picture of the earth’s system for quite a while, yet the actual flux of volatiles in and out of the deep earth still remains poorly quantified. One of the outstanding questions with profound implications for the Earth system regards how much of the subducted volatile budget is released back to the atmosphere, and how much remains trapped at depth.

Ocean Optics: Why choose the South American Andes?

Moussallam: Our objective was to provide the first accurate estimate of the flux of volatile species (H2O, CO2, SO2, H2, CO, HCl, HF, H2S and OCS) emitted by volcanoes along the entire length of the Nazca plate subduction zone (~6000 km), from the southern tip of Chile all the way to the equator. With nearly 200 active volcanoes (four currently erupting and 18 in a state of unrest), the South American Andes is one of the world’s most tectonically and volcanically active regions. The remote locations, high elevations (up to 6,893 m) and lack of established trails to the summit of most of these volcanoes means that few scientific studies have been done. There’s been very little data on this region in the global database of volcanic gas emissions, and we wanted to correct that.

Ocean Optics: What method did you use to measure SO2 in volcanic plumes, and how did our spectrometers fit in?

Moussallam: We use Ocean Optics spectrometers to perform a technique called ultraviolet scanning differential optical absorption spectroscopy (DOAS). It uses UV light from the sun, comparing the absorbance of a band specific to SO2 in the ultraviolet for light transmitted through the plume of gases emitted from the volcano to light that is not (305–335 nm is our typical monitoring range). This gives us the concentration of SO2. Then we factor in the speed of the plume movement to get the SO2 flux in kg/s or tons per day. SO2 is the only species we can do this with, as it’s the only one that absorbs strongly in the ultraviolet. Most of a volcanic plume is water, but since that’s already present in high concentrations in the atmosphere, it’s better to use SO2 as an indicator of volcanic activity. We then get closer to the volcano’s crater and correlate our remote measurements of SO2 using DOAS to local measurements of all the different species. The remote sensing gives us the flux of SO2 and only that. To complement that, we make direct measurements of the complete chemistry of the volcanic emissions (H2O, CO2, SO2, H2, CO, HCl, HF, H2S and OCS) using open-path FTIR and electrochemical and non-dispersive infrared (NDIR) “multi-gas” sensors. From these local measurements, we get a ratio of each species to SO2, then use those ratios in combination with the SO2 flux to determine the flux of the other gases.

Absorbance of volcanic gases

Absorbance of volcanic gases

Ocean Optics: How long have you been using Ocean Optics spectrometers?

Moussallam: The group I came from at the University of Cambridge has been using DOAS to measure the SO2 flux from volcanoes using Ocean Optics spectrometers for about 10 years. It’s very easy to do, as SO2 is found in low concentrations in the atmosphere, but in high concentrations in the gases emitted from volcanoes. The advantage of using UV light for the measurement is that scattered UV light from sun is quite uniform.

Ocean Optics: You’ve used USB2000+ spectrometers for DOAS in the past, but this was your first experience making measurements with the Flame spectrometer. How did you find the conversion process?

Moussallam: The conversion was pretty direct. All the scripts we had for the USB2000+ worked very easily for the Flame, and we found the spectrometers really similar. We had no issues with the transition – the three Flame spectrometers worked really well for what we needed.

Flame Spectrometer with Villarrica in the background

Flame Spectrometer with Villarrica in the background

Ocean Optics: How did Flame perform as compared to the USB2000+ spectrometers you’ve used in the past?

Moussallam: They were very similar. The spectral features were very clear, and noise was quite low. Results were just as good as before – slightly better. The Flame is better thermally controlled, which was very helpful in the field. Most of our measurements were made at higher temps, 15-40 °C. We used the spectrometers mostly in the morning, as you start getting good UV light around 9 a.m. in that region; we always had clouds in the afternoon.

Ocean Optics: Can you tell us about where the Flame spectrometers were mounted?

Moussallam: We measure the speed of the plume movement by doing a transection of the plume, and there are a few ways to do that. One is to put the spectrometer with telescope for collecting light on the Land Rover, and then transect the plume. But it is often more practical to scan the plume (that’s where we get the term scanning DOAS). We fix the spectrometer and use a step motor to change the angle of the telescope accurately, scanning it through the plume slowly but accurately while we collect spectra continuously. We can do this horizontally with the plume rising vertically, or if the plume is being blown horizontally by the wind, you can transect vertically. On this expedition, we usually got close enough to look at rising plumes (horizontal transection). That makes the geometry more accurate, as it’s hard to know the height of the plume when transecting a plume moving horizontally.

Horizontal vs vertical plume scanning

Horizontal vs vertical plume scanning

Ocean Optics: Is that the only way you used the Flame spectrometers?

Moussallam: No – another thing we do is to use the spectrometer to calibrate a UV camera. It’s a camera that uses filters to look at the light in 2 wavelength regions of the UV (310 nm for SO2, and 330 nm where SO2 doesn’t absorb). The ratio of these two images gives you an image of the SO2 in the plume in two dimensions. This is a technique that has recently been developed. It allows you to get an image of the plume and see how quickly the plume is moving in space. We don’t believe this method is very spectroscopically accurate, so we point the spectrometer to one point in the image and use it as a calibration point for the image. When we compare this method to our transection of the plume, we get good agreement. The camera gives a very good estimate of the speed of the plume, and minimizes the error in the flux measurement from the transected plume.

DOAS methods

Field of view (FOV) of various monitoring instruments: static DOAS, scanning DOAS and UV camera. The upper right inset shows an image take by the UV camera (the image is a ratio of the 310 to 330 nm) and the location of the static DOAS FOV within that image. The static DOAS is then used to calibrate the image in absolute terms.

Ocean Optics: You studied over 15 volcanoes in the South American Andes on this expedition. How long did you study each volcano or crater?

Moussallam: We measured both locally and remotely over the period of a few days, as both can change. Remote measurements can only be done when there is a clear sky, while local measurements can be made anytime. We did both over a few days to get a good average. Both the flux and composition of volcanic plumes can vary, which can indicate that something is changing at depth, but the fluctuations are not usually very large over these time scales. Once we knew the composition of the plume at a particular volcano, we could monitor the flux remotely over time with reasonably good accuracy. By looking at multiple volcanoes along a specific range, we can determine which ones contribute the most of a particular species to the atmosphere.

Ocean Optics: You weren’t the only ones looking at these volcanoes, were you?

Moussallam: That’s right. We’re also comparing results to satellite measurements of SO2 taken from space, though satellite-based measurements have much lower spatial accuracy than ground-based measurements. In fact, we coordinated with NASA to make satellite measurements concurrently with our ground-based measurements, with the goal of using the ground-based measurements to improve the calibration of satellite measurements.

Ocean Optics: Were any of the emissions different from what you expected?

Volcano cross section

Volcano cross section

Moussallam: The volcanoes we were looking at were all along the Nazca Subduction Zone. Many of the volcanoes had never been measured before, because they’re quite remote. No one really knew how much volatile gases were coming out of the volcanoes, and because they had never been measured before, we had quite a few surprises. Volcanoes that we thought were very small emitters of volcanic gases ended up being very high emitters of high temperature gases. There are two types of degassing volcanos. Some degas through low temperature hydrothermal systems (like the ones at Yellowstone National Park, which have fumaroles, are low temperature, and emit mostly rainwater that’s recycled through the crust). The other type of degassing is from gas coming directly out of solution from the magma, and those have a very different signature. The composition of the gases in these types of volcanoes tells you something about the magma at depth. Some volcanoes that were thought to have a few low-temperature fumaroles actually showed magmatic gases. Based on what we found, we now plan to do specific case studies at some different volcanoes that surprised us.

Ocean Optics: What do you hope to learn once you analyze the data collected?

Moussallam: What we did was take a snapshot for each volcano of its volatile gas flux. The larger project in question is to determine how volatile elements are cycled at the scale of the planet. To really understand the process of volcanic gas cycling, we want to extend our observations through time. Calibration to satellite measurements facilitates this and permits longer studies to be more accurate. Our calibration of the satellite data also makes it possible to go back over the satellite record for the last decade and reanalyze it with greater accuracy.

Ocean Optics: But this is part of a bigger picture still, isn’t it?

Moussallam: Yes, definitely. The larger question is to find out if the earth is ingassing or outgassing – whether volatile elements (water, carbon dioxide, etc.) are dominantly being trapped in the Earth mantle or released to the atmosphere. This is important, because we want to know how our atmosphere was created. Before life, the chemistry was regulated by the chemistry of volcanic gases. Now it is dominated by life itself (through respiration and photosynthesis). There are a lot of arguments going on to understand how our atmosphere was formed. It is believed that very early in the earth’s history all of our water and other volatile elements were lost to space and that all the water we currently have on earth was brought back at a later stage through an intense bombardment of comets or meteorites.

Ocean Optics: So what does all this mean for the average person?

Moussallam: Volcanoes contribute a number of gases to our atmosphere, so quantifying the flux of these gases at a large scale is of prime importance if we want to model this atmosphere and predict its fluctuations. The biggest uncertainty in current atmospheric models comes from not knowing how much gases volcanos contribute to our atmosphere. Many believe that these models are not incorporating volcanic emissions accurately and that this is the cause of the large discrepancy observed between models’ predictions (of atmospheric warming, for instance) and measurements. We hope our work will contribute a significant step towards a more accurate estimate, so that we can all better understand how our atmosphere, and therefore our climate, might evolve over time.


The Trail by Fire expedition was chiefly funded by Land Rover via a grant administered by the Royal Geographical Society of London. The scientific instruments were sourced or financed by various universities (Univ. of Cambridge, New Mexico Tech, UC San Diego, and others), research institutes (CNRS, INGV) and research councils (NERC and NASA). Ocean Optics contributed three custom-configured Flame spectrometers to the expedition.