Lava from a volcano

Volcanic eruption of Mt. Nyiragongo

12 July 2004 12:00 UTC

Lava from a volcano
Lava from a volcano

On Monday 12 July 2004, at 05:30 UTC, Meteosat-8 observed an eruption of the Nyiragongo volcano.

Last Updated

24 May 2022

Published on

12 July 2004

The Nyiragongo volcano (Democratic Republic of the Congo, geographical location 1.52 S, 29.25 E, summit elevation 3,469m) is one of Africa's most well-known volcanoes because of recent catastrophic eruptions. In January 2002 lava flowed from the volcano into the city of Goma, displacing 500,000 people.

On 12 July 2004, a thin ash plume rising to about 5,400m above the volcano was observed. One week later, on Sunday 18 July 2004, the Nyiragongo volcano erupted again. Both eruptions were observed in satellite images.

The eruptions continued in August 2004 causing health problems for about 60,000 people. About 30,000 square kilometres of land west of the volcano was destroyed by volcanic fallout, including 5,000 square kilometres of land in the nearby Virunga National Park, endangering chimpanzees and other wildlife. People in the area were contracting stomach illnesses as well as respiratory and bone diseases. Fluoride overdoses caused a variety of sickness and turning people's teeth transparent. The emissions also produced acid rain.

The eruption of 12 July 2004 is best visible in the Meteosat-8 High-Resolution Visible (HRV) images (see below). The animation actually shows two plumes coming from two locations close to each other: a faint plume extending southwest of the volcano; a thick plume extending to the southeast, and an arc of ash stretching over Lake Kivu between the two plumes.

The explanation for the different directions of the plumes could be that the plumes come from two sources located at different altitudes: the faint plume is probably related to a crater at a lower elevation where winds are more from the northeast, whereas the main; thick plume could be related to ash at a higher elevation coming from the main crater.

Unfortunately, there is no picture from the eruption to verify this hypothesis. It was, however, reported that the Nyamuragira volcano, located 14km northwest of the Nyiragongo volcano, also erupted during the July–August 2004 period.

Looking at one single image (e.g. the MODIS image below), one could have thought that the thin plume was something like the remnants of the plume from an earlier eruption and that a wind shift had then occurred. Thus, it is clear that for effective real-time monitoring of volcanic activity, images from geostationary satellites are essential and provide valuable information that is not contained in single images from polar-orbiting satellites.

The thicker plume extending to the southeast can also be detected in the infrared channels of MSG (for the thinner plume the resolution is obviously not good enough). As in the case of the sulphur plant explosion over northern Iraq (see case study), the plume is best visible in the brightness temperature difference between the IR10.8 and the IR8.7 channels.

Nyiragongo is indeed known for its high emission of sulphur dioxide (SO2) and Meteosat-8 has two channels, the WV7.3 and the IR8.7 channels, that are principally sensitive to high SO2 concentrations (see figure , courtesy CIMSS, University of Wisconsin and CSIRO, Melbourne). As shown in the images and the animation below, the brightness temperature difference between the IR13.4 and WV7.3 channels is less useful to detect the plume. This difference is strongly influenced by surface temperature variations and by changes in the water vapour content so that the signal from the SO2 plume is only visible at certain times (e.g. the night-time plume on 11 July between 00:00 and 04:00 UTC).

From experience with NOAA AVHRR and GOES imagery, it is also known that under certain conditions — semi-transparent ash clouds; small ash particles; large temperature difference between ash cloud and underlying surface and low water content in ash cloud — volcanic ash clouds can be detected in the split window channels (IR10.8 and IR12.0). Therefore, NOAA is currently producing an experimental GOES volcanic ash product that is based on the split window channels and the IR3.9 channel.

According to US experts, volcanic ash at high altitudes normally has a distinctive positive (IR12.0–IR10.8) temperature difference of more than 2K. However, in the case of the Nyiragongo eruption, the plume is not visible in the brightness temperature difference between these channels proving that, as known from other cases, the detection of volcanic ash is more difficult in the tropical atmosphere.

Met-8, 12 July 2004, 08:15 UTC
Figure 1: Meteosat-8 Channel 07 (IR8.7), 12 July 2004, 08:15 UTC. Larger Area
Met-8, 12 July 2004, 08:15 UTC
Figure 2: Meteosat-8 Channel 09 (IR10.8), 12 July 2004, 08:15 UTC. Larger Area
Met-8, 12 July 2004, 08:15 UTC
Figure 3: Meteosat-8 Difference Image IR10.8–IR8.7, 12 July 2004, 08:15 UTC. Larger Area
Animation 1 (06:00-10:00 UTC) Animation 2 (10-12 July 2004)
Met-8, 12 July 2004, 08:15 UTC
Figure 4: Meteosat-8 Channel 12 (HRV), 12 July 2004, 08:15 UTC. Larger Area. RGB Composite HRV, IR10.8–IR8.7. Animation (06:00–10:00 UTC)
Met-8, 12 July 2004, 08:15 UTC
Figure 5: Meteosat-8 Difference Image IR13.4–WV7.3, 12 July 2004, 08:15 UTC. Larger Area.
Animation (10–12 July 2004)
Met-8, 12 July 2004, 08:15 UTC
Figure 6: Meteosat-8 Difference Image IR12.0–IR10.8, 12 July 2004, 08:15 UTC. Larger Area
Animation (10–12 July 2004)
Met-8, 12 July 2004, 08:15 UTC
Figure 7: Meteosat-8 RGB Composite VIS0.8,IR10.8–IR8.7,IR12.0–IR8.7, 12 July 2004, 08:15 UTC
Met-8, 12 July 2004, 08:15
Figure 8: Meteosat-8 RGB Composite NIR1.6, VIS0.8, VIS0.6, 12 July 2004, 08:15

The MODIS RGB composite image (true colour) below shows the shifting ash plume blowing approximately 75km southeast into Rwanda. The red boxes mark the pixels where MODIS has detected a thermal anomaly, i.e. places where the temperature is much higher than the surrounding region. The boxes typically mark active fires, though in this case, one marker appears to be over the top of the volcano.

MODIS on Terra, 12 July 2004, 08:30
Figure 9: Terra MODIS RGB Composite (pixel size 250m), 12 July 2004, 08:30 UTC. Credit: NASA