Remote Sensing of Real-Time Volcanic Activity Part 1: Webcams and Thermal Cameras

Thermal camera view of Stromboli volcano, Italy (real-time: http://193.206.127.31/snap.jpg). A – 00:10 UTC on March 15, 2013.

Often, volcanoes are difficult to reach and it is not possible to have observers with a line of sight view of the volcano. Remote sensing of volcanoes and their products, pre-, syn- and post- eruption, provides a unique method to systematically analyze how they change with time and with the aim to detect possible precursors to significant events.

So the first question is “What is remote sensing?” Well, it is:

‘the science and art of obtaining information about an object, area or phenomena through the analysis of data acquired by a device that is not in contact with the object, area or phenomena under investigation”

Remote sensing data can come from ground-based equipment, airborne surveys and spaceborne satellites and their sensors. Here, in this article, I report on some of the capabilities from ground-based equipment, such as webcams and thermal cameras, that can detect thermal signals and detect/measure volcanic plumes and clouds in real-time. The data from this equipment can then be used for systematically detecting changes in volcanic activity.

Ground-based remote sensing equipment

While webcams and thermal cameras are not the only instruments and equipment that can be used to detect and analyze real-time volcanic activity, they have become some of the most available and used tools for remote sensing of a volcano, from the ground.

The number of webcams viewing a volcano is now rapidly increasing, with some volcanoes having multiple cameras, either viewing the volcano from different angles or with different zooms. For example, Shiveluch volcano in Kamchatka.  This volcano is monitored by the Kamchatka Volcanic Eruption Response Team (KVERT, http://www.kscnet.ru/ivs/kvert/index_eng.php).

Online, there are three webcams that provide real-time access views of the volcano, see Figure 1. Here, the three views all are of the same side of the volcano but at very different scales. Figure 1A shows the camera with lowest zoom and hence greatest view, with the volcano between the trees.

Figure 1C is has the most focused view of the volcano, with the dome in the centre of the image. Also, in Figure 1, the three views are all cloudy, for the time on March 18, 2014 (04:14 – 0:16 UTC). Also, note that these are daytime views of the volcano and as such, often webcams, are unable to capture volcanic signals are night, but this is not always the case. Some webcams can capture night-time activity.

Figure 1. Three webcam views of Shiveluch volcano on March 18, 2014. A – 04:15 UTC from Shiv1 – http://data.emsd.iks.ru/videokry/svl.jpg; B – 04:16 UTC from Shiv2 – http://volcano.febras.net/archive/latest_She2.jpg; C – 04:14 UTC from Shiv3 – http://volcano.febras.net/archive/latest_She.jpg Click to enlarge image.

To be able to capture the thermal signals from a volcano in a webcam at night, there are several items that need to happen: 1) the target is hot enough to elevate the webcam pixel data so it is brighter than surrounding ones; 2) webcam wavelength is designed to be susceptible to thermally hot targets; and 3) the weather conditions are such that no local cloud obscures the signal.

Let’s take the three webcams for Shiveluch again from Figure 1. Now a night-time view is provided, March 15, 2014, Figure 2. Here, the images are within 5 minutes of each other. Note that the Shiv2 webcam, Figure 2B, shows a strong thermal signal from volcanic activity at the dome on Shiveluch volcano. The other two show no thermal signals from changes in brightness of the webcam pixels. Here, Figure 2B’s webcam is more susceptible to the thermal signals occurring at the volcano.

Figure 2. Three webcam views of Shiveluch volcano on March 15, 2014. A – 10:55 UTC from Shiv1 – http://data.emsd.iks.ru/videokry/svl.jpg; B – 11:00 UTC from Shiv2 – http://volcano.febras.net/archive/latest_She2.jpg; C – 10:59 UTC from Shiv3 – http://volcano.febras.net/archive/latest_She.jpg Click on image to enlarge.

Additional examples, from my blog, on thermal signals can be found at the following links:

• Shiveluch volcano, Russia, webcam overnight, March 12 – 13, 2014 at 10 fps – http://volcanodetect.blogspot.com/2014/03/shiveluch-volcano-webcam-overnight_13.html
• Popocatepetl volcano, Mexico, webcam overnight, March 5 – 6, 2014 at 10 fps – http://volcanodetect.blogspot.com/2014/03/24-hr-webcam-view-of-popocatepetl.html
• Mount Etna volcano, Sicily, Italy, webcam 6 – 8 pm UTC, Jan 23 2014 – http://volcanodetect.blogspot.com/2014/01/general-interest-etna-as-night-falls-in.html

Algorithms can be designed to detect for these changes in thermal signals and provide an event detection system for changes in volcanic activity. Lessons can be learnt from the satellite remote sensing community in terms of change detection that can be adapted to webcam imagery. Then the webcam data can be used to detect volcanic activity in real-time. The next stage would then to be able to calibrate the data from brightness values to potential elevated temperatures.

Webcams can also be used to detect and measure volcanic plume height, as long as the altitude of the top of the webcam image is known. Also, for this the plume/cloud’s direction has to be assumed. Many researchers have used the plume height of the eruption plume to measure eruption rates (e.g. Sparks [1997] and Mastin et al. [2009]). If the plume is vertical in the plane of the camera, then the eruption rate method will be most accurate. If away from the camera, then the rate method will be an under-estimate and if towards the camera then the method will be an overestimate of the rate. Figure 3 shows some example of plumes captured by webcams, highlighting the issue of webcams being too close and the plume leaving the field of view.

Figure 3. Webcam views of Popocatepetl volcano, Mexico (real-time view: http://www.cenapred.unam.mx/images/popoTlamacas2.jpg). A – 00:10 UTC on April 18, 2013; and B –19:41 UTC on May 7, 2013. Click on image to enlarge.

Additional examples, from my blog, on volcanic plumes/clouds can be found at the following links:

• Shiveluch volcano, Russia, ash plume and cloud, January 20 2014, 22:57 UTC – http://volcanodetect.blogspot.com/2014/01/shiveluch-ash-plume-and-cloud-january.html
• Shiveluch volcano Russia, volcanic event, October 18, 2013 – http://volcanodetect.blogspot.com/2013/10/shiveluch-volcanic-event-october-18-2013.html

So webcams, while sometimes limited, can still be very useful for remote sensing of active volcanoes. They are usually designed for their most optimal use in the day-time, but some of them can be useful at night, but this depends on wavelength, thermal signal strength of the active volcano and local weather conditions.

There are some camera systems, set up at active volcanoes, which can measure in the thermal infrared and hence are very useful, day or night. Figure 4 shows a couple of examples of the thermal camera set up at Stromboli Volcano by the University of Florence. Figure 4A shows an image captured at 00:10 UTC (local nighttime) on Mar 15, 2013, where a new explosive event from the volcano has occurred. This image has been captured at a most opportune time, as Figure 4B shows that at 11:50 UTC on March 5, 2013, an explosive event has just happened as the slopes of the volcanic summit are at elevated temperature than the surrounding region.

Figure 4. Thermal camera views of Stromboli volcano, Italy (real-time: http://193.206.127.31/snap.jpg). A – 00:10 UTC on March 15, 2013 and B – 11:50 UTC on March 5, 2013. Click on image to enlarge.

These two images in Figure 4 show the timing of the data captured is critical for understanding the signals measured and the repeatability of the volcanic signals. Also, the data is only as a jpeg, and not raw data, so the colors seen would need to be converted to approximate temperatures using the scale bar as a guide. Spampinato et al. (2011) provide a detailed review of the different data processing, levels of accuracy, applications and methodology of data collection for volcanic surveillance with infrared cameras.

Thermal cameras provide an excellent dataset for remotely measuring a volcanoes activity. They are often expensive and as such only a few exist for real-time analysis and they are usually used in short term campaigns, either from the ground or in overflights. However, satellites and their sensors can provide a spaceborne view of what a volcano is doing, in terms of thermal signals, volcanic ash release, sulfur dioxide flux and deformation signals.

Author

Dr. Peter Webley is an Assistant Research Professor at the Geophysical Institute, University of Alaska Fairbanks. Dr. Webley’s focuses upon using remote sensing data to analyze natural hazards, such as volcanic events, forest fires, landslides and coastal erosion. Dr. Webley has designed new mechanisms to visualize the development of volcanic ash clouds. He has taken the three-dimensional dispersion model simulations that used to be visualized on two-dimensional maps and displayed them in their original three-dimensional form. Dr. Webley has been the guest editor for two special issues of the Journal of Volcanology and Geothermal Research (JVGR) in 2009 and 2013. His paper collaborating on eruption source parameters, Mastin et al. (2009) and listed in his selected publications is the highest cited publications in JVGR since 2008 with over 95 citations.

Recently in 2013, Dr. Webley, along with Dr. Jon Dehn, an Associate Research Professor at the Geophysical Institute, University of Alaska Fairbanks, formed a company called V-ADAPT, Inc [Volcanic Ash Detection Avoidance and Preparedness for Transportation], (www.vadapt.net), from their research at the University in analysis of volcanic activity and dispersion modeling of volcanic ash clouds. V-ADAPT, Inc. provides data, tools, analysis, and risk assessment of volcanic ash for the aviation and other transportation industries. They offer a comprehensive system to help in planning and response to volcanic eruptions for its clients. It is based on over 20 years of the founders’ experience in mitigating hundreds of eruptions in the North Pacific. The company focuses on volcanic hazard assessment and scenario planning through research and development, consultancy and service-orientated web-based tools.

References

Spampinato, L., Calvari, S., Oppenheimer, C., and Boschi, E., 2011. Volcano surveillance using infrared cameras. Earth-Science Reviews, 106 (1), 63-91.