Using a proper monitoring system, those changes could be detected and interpreted to issue alerts before the eruption occurs. However, each volcano is different and each eruption is unique, making the task of forecasting them sometimes very tricky.
Nyiragongo volcano in Democratic Republic of Congo, close to the border with Rwanda, erupted in 1977 and 2002. Despite its proximity to densely populated areas undergoing rapid urbanization growth, it was not adequately monitored at that time, mostly due to the difficult environment (regional political context, security issues, lack of infrastructures, poor sanitary conditions, etc.).
Since 2005, a Belgo-Luxembourgian (BeLux) scientific consortium, funded on several research projects, helped to reinforce the local volcano observatory in Goma (GVO), carrying out more than 100 missions in the region between 2005 and 2021. These projects and initiatives allowed for the progressive installation of the KivuSNet (seismic and acoustic) and KivuGNet (GNSS) ground networks as well as the setup of (among others) real-time data acquisition, transmission and analysis tools, thus massively boosting the volcano monitoring capabilities in the Virunga region.
At the same time, the increase in remote sensing capabilities significantly reinforced the ability to monitor volcanoes from space. As a result, despite the difficulties inherent to the area, the monitoring capacities for Nyiragongo volcano became comparable to other densely monitored volcanoes in the world. On May 22, 10 seismic stations, 12 GNSS receivers and 2 infrasound arrays were telemetered in real time to GVO. Shortly before 7 P.M. (local time), residents of Goma raised the alarm. They witnessed lava flowing from a fissure on the flank of the volcano. This eruption took everyone by surprise. The resulting confusion as well as a lack of anticipation complicated the management of the first hours of the crisis.
No clear precursors had been identified in the weeks or days before the eruption that could have suggested an imminent eruption. The existing monitoring, contingency plan and emergency procedures were designed for a scenario similar to the only two known previous eruptions (1977 and 2002), where precursors were clearly identifiable at least a week before the eruption. The 2021 eruption utterly defied this perception. Several factors may explain this:
Nyiragongo's historic activity is poorly known. The 2021 eruption was the first one to be monitored with instruments. Moreover, the 1977 and 2002 eruptions had been preceded by strong earthquakes (M>5) felt a few weeks to a few days before the eruption started. Nothing similar happened in 2021, hence suggesting another triggering mechanism.
Nyiragongo’s crater has been permanently erupting since 2002, continuously producing signals (e.g. earthquakes, tremor, infrasound) whose intensity and partly also signature are varying. Detecting significant changes that could characterize an impending eruption is not an easy task in these conditions.
In retrospective analysis, seismic records show that the first events marking unusual activity detected by KivuSNet started less than 40 minutes before the first lava flows. This very short delay between the first earthquakes and the flank eruption indicates that the magma needed very little time to reach the surface because it was already stored at shallow depth. This is the first time that such a seismic sequence associated with a flank eruption at Nyiragongo has been recorded by a local sensors network. The detection of signals appearing only in the very short term before the eruption provides fundamental insights to be integrated in the monitoring methods by GVO, but also potentially for other volcanoes in the world with a large quantity of magma stored near the surface.
Unfortunately though, the story does not end here. In addition to the dramatic human losses and destruction of infrastructures, such a kind of crisis has different side effects, often driven by individual opportunistic initiatives, which can have long-term damaging consequences.
The observatory’s mission is to monitor the volcano and anticipate eruptions. Thus in May 2021, after GVO “failed” to anticipate this unpredictable crisis, there came the fear to be blamed for it. In search of escaping the anger of the population, or in the hope of gaining personal advantage, some people in charge of raising the alert, including some officials of GVO, started to propagate disinformation, claiming for instance that a lack of data availability had prevented them from anticipating the eruption. In view of the well-known sensitive context of this African region, these claims quickly resonated in the social media and, eventually, also in some international print media, ignoring the evidence provided to them. This undermined any rational debate, affected the credibility of the organizations in charge of the monitoring and the crisis management, and severely complicated GVO’s task to coordinate the efforts while some external offers of assistance were directed to individuals rather than to the authorities. This was a tough lesson learned from this eruption and highlighted everyone’s responsibility, including people apparently far from the affected region or from the (local) volcano monitoring community.
At the end, it is worth noting that while the eruption started without any clear precursory signal, the seismic crisis and rapid deformation that occurred during the following week were a clear manifestation of magma migrating at less than 500m depth below the city of Goma and, then, below Lake Kivu. This undeniable shallow magmatic intrusion implied the non-negligible possibility of potentially devastating secondary events that led the authorities to order a mandatory evacuation a few days after the end of the eruption. Indeed, this potential second eruptive stage may have led to lava flowing into the city, a phreatomagmatic eruption (if the magma comes into contact with underground water) or a limnic eruption of Lake Kivu, the latter being feared because of the huge amount of gas dissolved in its deep waters. Such possible events are potentially much more dangerous than the three eruptions known so far. Luckily, none of these catastrophic scenarios occurred but the very shallow depth at which magma migrated reveals that these scenarios may be of a higher probability than previously expected.