|
GNV – Nyiragongo Project In the frame of the international aid operations following the January 2002 eruption
1.Physico-chemical investigation of the volcanic products erupted during the January 2002 eruption, and simulation of possible lava flow paths, aimed at volcanic hazard mitigation
Coordinator: Paolo Papale i. Istituto Nazionale di Geofisica e Vulcanologia - Sede di Pisa via della Faggiola 32 I - 56126 Pisa - Italy phone
+39 050 83 119 31
2.Background Lava flows which covered the Southern flank of Nyiragongo and devastated the city of Goma during the January 17th, 2002 eruption were produced from several locations along a huge system of fractures extending over more than 10 km from the volcano summit to the suburbs of Goma (UN-OCHA report, March 2002). Lava flows from high altitude vents traveled at a high velocity estimated on the order of tens of km per hour, while lava flows from low altitude vents had an average velocity of the order of hundreds of m per hour. Lava flow velocities and field evidence concur to suggest that the magma discharged from high altitude was degassed and very fluid, while magma discharged from low altitude, and which devastated the city of Goma, was gas-rich and viscous. Accordingly, the kind of hazard associated with the two types of lava flows was different. Lava flows from high altitude vents posed a severe threat to the inhabitants of the villages on the volcano flank, since many of them, especially children, had no time to escape and were run over by the fast flowing lava. Conversely, several trees, even of relatively little size, did not fell due to extremely fluid behavior of lava flows. Different from the above, slowly moving lava flows from low altitude vents did not represent a direct menace for the inhabitants of Goma, as they could easily escape from them, but had a high destructive capability resulting in the devastation of a significant portion (about 15%) of the town including the city and business center. It seems likely that such lava flows were feeded by magma initially at higher pressure with respect to the magma that feeded the highermost lava flows. However, it is to be understood whether such high pressure magma came from the inner portion of the shallow volcano system, or from deeper levels within the crust. The implications for the eruption dynamics and the possible near-future scenarios are quite different, since in the latter case - deep provenance of magma feeding lava flows which destroyed the city of Goma - an efficient mechanism leading to magma uprise should be involved, suggesting that a continuation of the rifting process in the area might involve new magma uprise and the formation of new highly destructive lava flows. Points 1)-3) above represent the objectives of the present research proposal. Lava flow paths largely depended on vent location and morphology, with the several ancient tuff cones and the old Shaheru crater, located along the Southern slope of the volcano, playing a substantial role in diverting the flowing lava, and resulting in a lateral expansion of the overall lava field. Local morphology and ground slope, together with vent location, determined the areas of Goma that were destroyed by lava flows, and those that were safe.
3.Objective The above picture rises the necessity of understanding 1) which are the factors that determined such a large viscosity difference between the two lava flow types, 2) is there a different origin of the two lava flow types, 3) what is the depth of provenance of the lava flows which devastated the city of Goma, and 4) what are possible lava flow paths in case of future eruption, or, in other words, what is the distribution of vent locations from which lava flows could be able to affect the densely inhabitated towns of Goma (DRC) and Gyiseni (Rwanda). Points 1 to 4 above represent the objectives of the present project. Such objectives will be accomplished as described in the following section.
a.Methods
We propose:
4.to undertake a detailed chemical and mineralogical analysis of lava samples from the high and low altitude lava flows, including major and trace element determinations, as well as isotopic determinations of volatile abundance in glass inclusions within crystals and in the groundmass, and textural determinations of the different kinds of lava discharged from different vents;. This objective encompasses tasks 1 to 3 below;
5.to perform laboratory determinations of the viscosity of melt samples obtained from the upper and lower lava flows (task 4);
6.to realize a map of possible lava flow paths in the Nyamuragira-Nyiragongo-lake Kivu area. The map will immediately give the areas that can be invaded by lava flows from a vent opening somewhere in the above region, and will serve for the early warning operations, for the medium-long term reconstruction and/or expansion of the towns of Goma and Gyiseni, for selection of safest areas for refugee camps, and for the preparation of contingency plans in case of future eruptions (task 5).
Below a more detailed description of the tasks in the project is reported.
a.Description of tasks
Task 1 - Chemical and isotopic analyses of lava samples. Responsible: A. Basu (University of Rochester). About 10-15 samples from the topographically high and low portions of the lava flow of the January 2002 eruption will be analysed. Analyses will include major and trace elements by XRF, and another 25 elements including rare earths by ICPS method. Isotopic analyses will include the 3He/4He ratios in minerals and glass inclusions, by crushing them and then measuring the volatile gases by mass spectrometry. Additionally, the Sr and Nd isotopes will be analysed using thermal ionization mass spectrometry.
Task 2 - Mineralogical and petrochemical analysis of lava samples. Responsibles: M. Pompilio (INGV Catania) and P. Landi (INGV Pisa). The same samples as in task 1 will be analysed to determine the mineral abundance and chemistry, and the equilibrium-nonequilibrium relationships between different minerals and groundmass. Analyses will be done by Electron Microprobe ??.
Task 3 - Textural analysis of lava samples. Responsible: M. Polacci (INGV Pisa). Gas-rich and gas-poor samples from lava flows produced at different locations along the fracture system will be analysed in order to determine the textural parameters characterizing vesicle distributions. These data will give constraints to the modalities of degassing from the different portions of the active fracture system. In turn, this will help understanding the different eruption modalities for the lava erupted at different altitudes along the volcano slopes.
Task 4 - Determination of melt viscosities. Responsible: D.B. Dingwell (IMPG Munich). Viscosities of melts representative of the groundmass portion of lava flows from different altitudes along the lava field will be determined in the laboratory, as a function of dissolved water content and temperature. Different techniques appropriate to the different investigated viscosity ranges will be used, namely, concentric cylinder viscometry at low viscosity, and micropenetration at high viscosity. The data will be used for the calibration of semi-empirical viscosity laws, like the TVF equation and its more recent modifications, which will allow calculation of viscosity as a function of temperature and dissolved amount of water.
Tasks 1 to 4 above will give us a detailed picture of the similarities and differences in the physico-chemical characteristics of magma erupted from the uppermost and lowermost portions of the fracture system during the January 2002 eruption. Such physico-chemical determinations will make us understand the origin and nature of the source region of magma, the possible occurrence of mixing between magma already stored in the crust and new magma from mantle depths, the processes of differentiation and crystallization of magmas, the possibly different depths of provenance of magma erupted at different altitudes, and the factors that determined the different rheological behaviors. The final research products should allow us to formulate hypotheses on the dynamics of the January 2002 eruption (whether it was determined by new uprise of magma from what depth, etc?), and therefore suggest a picture for the present state of the volcano, which will represent a basis for the interpretation of monitoring signals and for forecasting the volcanic hazard.
Task 5 - Realization of a map of possible lava flow paths. Responsible: A. Neri and P. Papale (INGV Pisa). Due to the present lack of a DEM in the area of interest, in light of the activities proposed by a French group of researchers involving P. Briole and J.-C- Komorowski, and within the frame of the urgent activities which are being carried on under the coordination and financial support of UN-OCHA, the present task will be carried out in successive steps which are described below. Such steps represent successive refinements of the proposed map, and allow a less accurate, preliminary map to be obtained in a shorter time, and a more accurate, detailed map to be obtained subsequently. During the first step we will obtain an existing map with resolution 1:50000 (Thonnard RLG, Denayer ME, Antun P, 1965. Carte Volcanologique des Virunga (1/50000), Afrique Centrale, Feuille No. 1, published by the Centre National de Volcanologie (Belgique), Missions Geologique et Geophysique aux Virunga, Ministere de l´Education et de la Culture, Bruxelles, Belgium), digitalize the steep sloping regions around the central craters of Nyiragongo and Nyamuragira volcano (up to 4-6 km distance), and use the digitalized map as a basis for preliminary assessment of possible lava flow paths in the upper volcano areas. This will allow an evaluation of possible flow directions from vents located close to or within the central craters. In the second step we will use as a digital topographic base the DEM produced by the above mentioned French group, who will process inages from the IKONOS satellite system and obtain a digital map with vertical precision close to 1 m and horizontal resolution of 3-5 m in a limited area between Nyiragongo and the town of Goma, and a broader map with resolution better than 10 m including the Nyamuragira-Nyiragongo-lake Kivu area. With this basis, a much more detailed and confident map of possible lava flow paths over the entire region of interest will be produced. In a third step we will merge our results with additional ones from the French group, who will also realize a detailed map of lava flows from the January 2002 eruption. Real lava flows will be reproduced using the DEM and vent positions from the French group, and the simulated paths will be compared with the mapped ones in order to address the confidence of the method employed for the kind of lava erupted at Nyiragongo. Possible lava flow paths will be obtained using a probabilistic lava flow model based on the assumption that lava flows follow the steepest path downward from the erupting vent. The model does not employ physical modeling equations, however, the spreading of lava front and the capability to overflow small obstacles is taken into account by using stochastic variations of the topography over which lava moves. Applications to the 2001 Mt. Etna eruption have successfully predicted the observed lava flow paths.
7.Milestones and project length
The project will last 6 months, in order to respond to the urgency represented by the on-going unrest at Nyiragongo volcano. A project meeting will be held at the 4th month of project.
2nd month: detailed chemical, mineralogical, isotopic, textural analyses, and viscosity measurements of lava samples from the lowermost vents; 4th month: detailed chemical, mineralogical, isotopic, textural analyses, and viscosity measurements of lava samples from the uppermost vents; simulations of lava flow paths based on digitalization of 1:50000 map; possibly (depending on project outcomes from the French group), first simulations of lava flow paths with new DEM; 6th month: reconstruction of the dynamics of the January 17th, 2002 eruption, and implications for present state and future scenarios; production of a map of hazard by lava flow invasion in the Nyiragongo-Lake Kivu and, possibly, Nyamuragira area.
8.Research personnel
Paolo Papale, INGV Pisa (Coordinator) Asish Basu, Rochester Univ. NY Augusto Neri, CNR Pisa and INGV Pisa Massimo Pompilio, INGV Catania Sonia Calvari, INGV Catania Donald B. Dingwell, IMPG Univ. Munich Margherita Polacci, INGV Pisa Patrizia Landi, INGV Pisa Massimiliano Favalli, CNR Pisa Piergiorgio Scarlato, Carmela Freda, INGV Roma Dario Tedesco, Univ. Napoli 2; Orlando Vaselli, Univ. Firenze (funded by UN-OCHA) Jean-Christophe Komorowski, Pierre Briole, IPG Paris (funded by UN-OCHA) Research contract 1 (6 months fully dedicated for viscosity determinations) Research contract 2 (6 months fully dedicated for map digitalization and lava flow path simulation runs)
Costs (k euro)
Analytical costs (lab time, consumables): 18 Personnel: 13 Travel expenses: 8 Acquisition of a plotter system: 8 Consumables: 1
1.TOTAL: 48
a. Fund Partition
The funds (in keuro) will be partitioned as follows:
4 (analytical) + 2 (travel) = 6 to the INGV Catania; 2 (analytical) + 13 (personnel) + 2 (travel) + 8 (plotter acquisition) + 1 (coordination costs and consumables) = 26 to the INGV Pisa; 6 (analytical) + 2 (travel) = 8 to the IMPG, University of Munich; 6 (analytical) + 2 (travel) = 8 to the Univ. of Rochester. |