(This is a sample cover image for this issue. The actual cover is not yet available at this time.)
This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
Author's personal copy Journal of Volcanology and Geothermal Research 219-220 (2012) 41–51
Contents lists available at SciVerse ScienceDirect
Journal of Volcanology and Geothermal Research journal homepage: www.elsevier.com/locate/jvolgeores
Tephra hazard assessment at Concepción Volcano, Nicaragua C. Scaini a,⁎, A. Folch a, M. Navarro b a b
CASE Department, Barcelona Supercomputing Center-Centro Nacional de Supercomputación (BSC-CNS), Barcelona, Spain Instituto Nicaragüense de Estudios Territoriales (INETER), Nicaragua
a r t i c l e
i n f o
Article history: Received 26 September 2011 Accepted 16 January 2012 Available online 28 January 2012 Keywords: Tephra fallout Probabilistic hazard maps Concepción volcano FALL3D model
a b s t r a c t Concepción volcano in Ometepe Island, Nicaragua, is a highly active volcano with a rich historical record of explosive eruptions. Tephra fallout from Concepción jeopardizes the surrounding populations, whereas volcanic ash clouds threat aerial navigation at a regional level. The assessment of these hazards is important for territorial planning and adoption of mitigation measures. Here we compute probabilistic hazard maps for Concepción volcano considering three different eruptive scenarios based on past reference events. Previous geological analysis is used to quantify the eruption parameters of the reference events. We account for uncertainties in the definition of the scenarios trough probability density functions. A representative meteorological dataset is created for each scenario by running the WRF-ARW mesoscale meteorological model over a typical meteorological year, defined in terms of wind speed and direction at a given atmospheric height. Tephra transport and deposition under different eruption and wind conditions is modelled using the FALL3D dispersion model. For each scenario, simulations are combined to build probabilistic hazard maps for critical values of tephra load and for threshold values of airborne ash concentration at relevant flight levels. Results are useful to identify the expected impacts for each eruption type and aim at improving the assessment and management of risk in the region. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Dispersion and fallout of tephra from explosive volcanic eruptions can cause multiple impacts on social and anthropic systems and account for a large fraction of the total economic loss caused by volcanic activity. Impacts, related to both weak and strong intensity explosive eruptions, can be caused by the presence of airborne ash, by tephra deposition at ground, or by post-eruption remobilization of deposited materials. Volcanic ash clouds disrupt air traffic at regional or even at continental scale (e.g. Guffanti et al., 2009). Fine ash particles in suspension, even if at low concentrations, threat engine turbines and are highly abrasive to fuselage, windscreens and mechanical components of airplanes (Casadevall, 1993). Substantial tephra deposition over buildings and infrastructures can lead to total or partial collapse of roofs (Spence et al., 2005a, 2005b). Electricity and energy lines can be interrupted by ash fallout, leaving the affected communities without power supply (Johnston and Becker, 2001). Regarding the water distribution systems, volcanic products can contaminate water bodies and cause damage and disruption of purification plants (Stewart et al., 2006). Tephra fallout also impacts heavily the transportation networks. Depositions of few mm are sufficient to cover road pavements impeding safe driving and also affect the operability of airports (Guffanti et al., 2009). Finally, tephra deposition at ground can also
⁎ Corresponding author. E-mail address:
[email protected] (C. Scaini). 0377-0273/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.jvolgeores.2012.01.007
cause major post-eruptive impacts. Mud flows and lahars, favoured at high-slope volcano flanks and after heavy rainfalls, have a tremendous destructive potential, particularly in tropical and rainy environments. Re-suspension of fine ash particles, a phenomenon typically occurring in arid and windy zones, may cause breathing difficulties and long-term respiratory diseases among exposed people (Baxter, 1999). For all these reasons, tephra hazard assessment is a relevant aspect in terms of managing policies and territorial planning. The Ometepe Island (Fig. 1), located in the Nicaragua Lake, is formed by two Quaternary stratovolcanoes, Maderas (11.4°N, 85.5°W, 1394 m a.s.l.) and Concepción (11.5°N, 85.6°W, 1700 m a.s.l.), connected by a narrow isthmus. Concepción is a highly hazardous volcano. The recent geological record reveals an intense explosive activity characterized by Strombolian to sub-Plinian events basaltic-to-dacitic in composition (Borgia and van Wyk de Vries, 2003). Due to the prevailing winds, past explosive eruptions have generated tephra deposits at the insular town of Moyogalpa (15,000 inhabitants) as well as at the western shore of the Nicaragua Lake (e.g. Rivas municipality, 45,000 inhabitants). One of the most copious fallout deposit is associated with the 1977 Concepción eruption (Delgado-Granados et al., 2006). Last eruptive event occurred in 2010 (Fig. 1b) and caused minor tephra deposition at the western part of the island. At present, the occurrence of intense tephra fallout would have a strong impact on the local communities, which are insufficiently prepared. Ometepe Island hosts multiple villages along its perimeter. Most buildings and structures have low-resistance covers and are poorly maintained. This makes the edification system highly vulnerable to tephra deposition because even moderate loads (~100 kg/m2) may
Author's personal copy 42
C. Scaini et al. / Journal of Volcanology and Geothermal Research 219-220 (2012) 41–51
Fig. 1. Left: view of Ometepe Island from the mainland. Right: Concepción volcano during the 15th April 2010 eruption. Photo taken by Ona Corominas at “La Flor” community, southern Ometepe.
cause strong damage to edifices and inhabitants. A main circumambulation road and minor unpaved paths, often interrupted by landslides during the rainy season, interconnect Ometepe villages. There is no redundancy in the road network, a fact that increases the vulnerability of the system (large parts of the island can potentially become isolated by tephra fallout). A single boat route connects the major town of Moyogalpa with San Jorge, a little village in the western continental shore of the lake. Apart from the proximal effects, explosive activity can also impact on a wider area. Ash fallout could easily reach the heavily populated west coast of the Nicaragua Lake, as occurred during many past events. Moreover, the city of Managua, hosting the international airport and most of Nicaraguan infrastructures, lays at only 90 km NW from the volcano. An eruption under SE blowing winds could easily affect the operability of the Managua airport and several regional airplane routes passing over the area. Despite the high potential impacts of an explosive event at Concepción volcano, there is no official hazard map to be adopted as a reference by local municipalities and national civil protection authorities. Probabilistic hazard maps would give a strong support in the definition of expected impacts and in the prioritization of mitigation measures. Maps of ash concentration at different flight levels would help identifying the threat posed by volcanic ash clouds from Concepción to the Managua airport and regional airplane routes. Previous studies in this sense are very scarce. Delgado-Granados et al. (2006) assessed different volcanic hazards from Concepción volcano based on stratigraphic studies. In particular, the assessment of hazard from tephra fallout builds on the analysis of isopachs from few events combined with a statistical study of the prevailing regional winds. Here we extend the work of Delgado-Granados et al. (2006) and present a complete tephra hazard assessment for three different eruptive scenarios. Hazard evaluation builds on representative meteorological datasets (based on the concept of the typical meteorological year) and on numerical simulations that allow elaborating probabilistic hazard maps for different ground loads and airborne concentrations at relevant flight levels. The manuscript is arranged as follows. Firstly, we overview the geological setting and describe the methodology, including the definition of the eruptive scenarios, the strategy adopted to construct a synthetic meteorological dataset representative of the regional meteorology over the past 30 years, the set up of the tephra transport and dispersal model and, the construction of probabilistic hazard maps. Secondly, we present the results for the
different scenarios and investigate possible seasonal and daily dependencies. Finally, we conclude with a discussion on the social implications of the results and outline some necessary future developments. 2. Geological setting Nicaragua lies on the Central American volcanic arc, originated by the subduction of the Cocoa tectonic plate under the Caribbean plate (Kutterolf et al., 2007). The nation hosts 19 active volcanoes (Siebert and Simkin, 2002) including the morphologically twin stratovolcanoes of Maderas and Concepción that form Ometepe Island. Although highly active and dangerous, Concepción volcano has not been extensively studied. Borgia and Van Wyk de Vries (2003) provided the most detailed existing work on the morphology, stratigraphy, petrography, structure and deformation of the volcano. According to this work of reference, the volcano experimented a first evolution cycle characterized by a pre-volcanic environment, a building phase in which the volcano edifice grew up, and a final destructive phase. The “Tierra Blanca” Plinian eruption (2720 ± 60 B.P. according to radiogenic dating) marked the end of the first building phase (Borgia and van Wyk de Vries, 2003) and left a very characteristic stratigraphic record measuring up to 3 m at some parts of the island. After a period of apparent quiescence, the 1883 eruption marked the resuming of the eruptive activity and the entrance to a second evolution cycle. Several eruptions have occurred between 1883 and 2010, all associated with the growing of the present volcanic cone. The historic activity from 1883 up to 1977 was characterized by Strombolian or small sub-Plinian events (Borgia and van Wyk de Vries, 2003), with simultaneous occurrence of flank lava flows. The erupted material was basaltic to silicic-andesite in composition. The 1977 event was followed by an almost continuous series of small-intensity gas-rich ashproducing eruptions and, according to Borgia and van Wyk de Vries (2003), similar activity can be expected in the near future. Field studies (Borgia and Van Wyk de Vries, 2003; Delgado-Granados et al., 2006) clearly indicate that historical tephra deposits are more frequent at the Western sector of the volcano and that deposition has been negligible (or left no preserved record) in the rest of the Island. This is a direct consequence of the trade winds, which in this geographical area have a predominant westerly component. According to Borgia and Van Wyk de Vries (2003), the maximum short-term deposition thickness that can be expected in the island is of about 50 cm.
Author's personal copy C. Scaini et al. / Journal of Volcanology and Geothermal Research 219-220 (2012) 41–51
3. Methodology Elaboration of probabilistic hazard maps involves three different steps: the definition of an eruptive scenario (with an associated range of uncertainty), the choice of a meteorological dataset covering a large period of time (decades) or a representative sampling of it, and simulations with a Tephra Transport and Dispersion Model (TTDM) considering different meteorological and volcanological inputs spanning the range of uncertainty. Simulations are then merged to produce hazard maps, giving the probability to exceed pre-defined thresholds of tephra ground load or airborne ash concentration. This section describes the strategy used in each of these steps. 3.1. Eruptive scenarios The definition of eruptive scenarios is usually based on the maximum expected or most probable event, although other less frequent options are possible. Following the work of Delgado-Granados et al. (2006), we consider three different eruptive scenarios for Concepción volcano based on three past reference events: • Low Magnitude Scenario (LMS) considers an eruption similar to the 2009–2011 events (Fig. 1b). The scenario is characterized by a short, very small intensity eruption (VEI ≈ 1) with gas ejection and few emission of tephra, as has been occurring since the 1977 eruption. This kind of event has the higher probability of short-term occurrence. The expected consequences of the LMS are low to moderate damages to proximal crops, minor damages to structures, and partial disruption of traffic in the island. • Medium Magnitude Scenario (MMS) considers an eruption similar to the 1977 event. This scenario is characterized by a short-duration Strombolian eruption of moderate intensity (VEI ≈ 2–3) with development of a sustained eruptive column and considerable emission of tephra. Impacts are expected not only in the island, but also across the continental shore of the Nicaragua Lake. • High Magnitude Scenario (HMS) considers an eruption similar to the 2720 B.P. “Tierra Blanca” event, which marked the end of the first building phase of the volcanic edifice (Borgia and van Wyk de Vries, 2003). This scenario is characterized by a sub-Plinian to Plinian eruption (VEI ≈ 4) with a vigorous sustained column lasting from several hours to few days. These three scenarios correspond to the most significant types of events expected for Concepción volcano and allow us to perform a general and coherent hazard assessment. Eruptive scenarios are characterized by the eruption column height, total erupted mass, eruption duration (or, equivalently, averaged mass eruption rate), and Total Grain Size Distribution (TGSD). These are the parameters that the FALL3D model needs in order to define the source term (e.g. Folch et al., 2009). Table 1 summarizes the characteristics of the scenarios. Table 1 Source parameters defining the eruptive scenarios. A Gaussian PDF peaking at the average value is assumed. Values of MER are obtained from column height using the Buoyant Plume Theory (Bursik, 2001). Parameter Column height (km)
MER (kg/s)
Duration (h)
Mean particle size (Φ)
LMS (2010) Min Average Max Min Average Max Min Average Max Min Average Max
2 2 2 104 104 104 1 1 1 1 1 1
MMS (1977) 3 5 7 105 0.5 × 106 106 0.5 1.5 2.5 2 1 0
For each scenario, the mean (more probable) values of the model input parameters are those of the respective reference eruption (2010, 1977 and 2720 B.P. respectively). However, and in order to deal with natural uncertainty, we consider a range of variation for each parameter assuming a Gaussian Probability Density Function (PDF) peaking at the reference value. LMS has an eruption column height of 2 km, whereas MMS and HMS have 5 ± 2 km and 20 ± 8 km (above the vent) respectively. The corresponding intervals of averaged mass eruption rates according to Mastin et al. (2009) and simulations performed using a 1D Buoyant Plume Theory model (Bursik, 2001) are of the order of 10 4 kg/s, 10 5 ÷ 10 6 kg/s, and 10 6 ÷ 10 8 kg/s respectively. Eruption durations for the 3 scenarios are of 1 h, 1.5 ± 1 h, and 30 ± 24 h respectively. The mean particle size is fixed to 1Φ (i.e. 0.5 mm) for the LMS, but varies between 2Φ and 0Φ (i.e. from 250 μm to 1 mm) for the HMS and MMS. We use the same range of mean particle size for these two scenarios although one could expect the HMS to produce a finer granulometric distribution because of its higher explosivity. We adopt this simplification because of the lack of detailed granulometric studies from Concepción deposits. The choice of the interval is based on studies of similar eruptions (e.g. the well-documented 1995–1999 eruptions of Soufriere-Hills volcano; Bonadonna et al., 2002). Finally, we do not take into account aggregation processes. It is important to mention that the three scenarios do not cover all possible eruptions from Concepción volcano. Some physically possible values of the source parameters are not included in the analysis and, moreover, the relative probability of occurrence of each scenario during a given time window is highly uncertain. For these reasons, each scenario has to be considered independently and their sum does not constitute a global hazard assessment. 3.2. Generation of a typical meteorological year using WRF-ARW Numerical TTDM require 4D meteorological data (mainly wind field and atmospheric temperature and density), typically furnished by global or mesoscale Numerical Weather Prediction Models (NWPM). In the context of hazard map elaboration, the meteorological dataset must be representative of the meteorology of the area over a sufficiently large period of time (typically years to decades). Moreover, for a fully numeric TTDM, the spatial resolution of the meteorological data and that of the TTDM should be comparable. Here we want to consider spatial resolutions of few km. Because the typical horizontal resolutions of global meteorological models are of ~ 1° (i.e. ~100 km at the equator or along a meridian), too coarse for our purposes, our analysis requires data from a mesoscale model. To this purpose, we use the Weather Research and Forecasting model (WRF-ARW), a mesoscale NWPM designed for both operational
Table 2 TMY obtained at Concepción for each scenario (pressure level) considering wind speed and direction as reference parameters with weighting factors of 0.3 and 0.7 respectively and using data from the NCEP/NCAR re-analysis 2 interpolated at the volcano coordinates (1979 to 2009).
HMS (2720 BP) 12 20 28 106 107 108 6 30 52 2 1 0
43
January February March April May June July August September October November December
LMS 850 mb
MMS 600 mb
HMS 70 mb
2005 1979 1990 1986 1988 1993 1998 1979 2001 1996 2005 1995
1990 1988 1997 1982 2000 2006 1990 1987 1984 1994 1987 1995
2005 2004 1996 1985 1979 1981 1981 1986 1988 1992 1988 2003
Author's personal copy 44
C. Scaini et al. / Journal of Volcanology and Geothermal Research 219-220 (2012) 41–51
Table 3 Threshold values adopted in the elaboration of hazard maps at Concepción. Parameter
Threshold
Effect
Tephra ground load Tephra ground load Tephra ground load Airborne ash concentration Airborne ash concentration
1 kg/m2 50 kg/m2 100 kg/m2 0.2 mg/m3 2 mg/m3
Affectation and disruption of road traffic Collapse of weaker structures Collapse of stronger structures Restricted-fly zone No-fly zone
forecasting and atmospheric research (Michalakes et al., 2001). Mesoscale meteorological simulations over large periods (e.g. 30 years) involve large computing times and storage capacity. To circumvent these problems we adopted the Typical Meteorological Year (TMY) strategy. A TMY consists of 12 representative months selected from individual years of a time period and collated in a meteorological database (e.g. Finkelstein and Schafer, 1971; Janjai and Deeyai, 2009). The
concept of TMY is widely diffused in different fields like climatology (Kalogirou, 2003), solar power energy (Bulut, 2010), wind power analysis for generation plants (Yang et al., 2003), or energy simulations in building design (Lam et al., 1996), among others. The TMY is defined in terms of a meteorological variable of interest, in our case the wind field. Unfortunately, different definitions of the TMY exist in literature, and a single procedure to construct the TMY is not established. Here we apply the method of Filkestein–Schafer (FS) (Finkelstein and Schafer, 1971), which allows to define a TMY in terms of different meteorological parameters (in our case, wind speed and direction at a given location and altitude). We calculated three different TMY from the NCEP-NCAR re-analysis 2 data (1979 to 2009) interpolated at the coordinates of the volcano and at different pressure levels of 850, 600 and 70 mb, corresponding approximately to elevations of 1.5, 4.5 and 18 km above sea level. These pressure levels have been selected in conformity with the eruptive scenarios, and allow us to have representative meteorological data at different altitudes.
Fig. 2. Typical buildings found in Ometepe Island representing weaker (top) and stronger (bottom) constructions.
Author's personal copy C. Scaini et al. / Journal of Volcanology and Geothermal Research 219-220 (2012) 41–51
45
Fig. 3. Tephra ground load for a typical LMS run. Contours given in kg/m2.
The FS method bases on two criteria to select 12 months by minimizing the difference with the so-called long-term conditions (the adjective long-term refers here to the whole temporal series, from 1979 to 2009). The procedure to obtain the TMY can be summarized as follows: 1. For each month and parameter, calculate the long-term Cumulative Distribution Function (CDF) by grouping period data in bins and counting the number of values in each bin. It gives 12 long-term CDFs for each parameter. 2. For each month calculate N specific CDF, where N is the number of years (in our case, 30 specific CDFs for January, 30 for February, etc.). 3. As a first criterion, find the 5 years with minimum FS factor for each month, defined as the difference between long-term and specific CDFs, weighted for each parameter. We performed a weighted sum of the FS parameters (in our case only wind speed and direction) using weighting factors of 0.3 and 0.7 for wind speed and direction respectively. This choice is subjective, but can be justified because wind direction is more relevant to our purposes as it determines directional patterns of hazardous areas. The lower the weighted FS factor, the better a month of a year (e.g. February 1999) approximates the long-term CDF of that particular month (e.g. February).
4. As a second criterion, find which of the 5 years for each month has a minimum Root Mean Square Difference (RMSD) and finally select the 12 months conforming the TMY. Results for each pressure level (i.e. for each scenario) are shown in Table 2. Once the TMY's are defined, we run the WRF-ARW model to construct a yearly synthetic meteorological dataset for each scenario. WRF-ARW runs considered three nested domains centred respectively over Central America, Nicaragua, and the area around Ometepe Island. The computational domains had horizontal spatial resolutions of 32, 8, and 2 km respectively, and 18 vertical layers up to 10 hPa (i.e. approximately 30 km height). Initial and boundary conditions for WRF-ARW were furnished by the NCEP-NCAR re-analysis 2, consistent with the calculation of the TMYs. 3.3. The FALL3D TTDM To simulate the atmospheric dispersion of tephra we used the FALL3D dispersion model (Costa et al., 2006; Folch et al., 2009). The use of a fully numeric TTDM to elaborate hazard maps is uncommon because requires a large computational time. However, as in Folch
Fig. 4. Airborne ash concentration contours (mg/m3) at FL050 (top) and FL150 (bottom) for a typical LMS run. Left: results 3 h after the eruption start. Right: results 6 h after the eruption start.
Author's personal copy 46
C. Scaini et al. / Journal of Volcanology and Geothermal Research 219-220 (2012) 41–51
Fig. 5. Tephra ground load hazard maps for MMS. Contours give the probability to achieve values of 1 (top), 50 (middle), and 100 (bottom) kg/m2 respectively.
and Sulpizio (2010), a fully numeric TTDM is necessary because we want to compute airborne ash concentration maps in addition to the classical ground deposition maps. FALL3D computes dispersion and deposition of tephra over a 3D computational domain given meteorological variables (in our case WRF-ARW outputs for the different TMYs) and model input parameters (given by the scenarios). For each run, model input parameters (column height, eruption duration, and mean particle size) were sampled within the considered range (Table 1) using a stratified sampling technique (e.g. Rao and Krishnaiah, 1994). Stratified sampling is a sampling technique that can be applied when previous knowledge about the shape of a population of values is known or assumed (a Gaussian PDF in our case), and consists of a random sampling but limiting the number of members within pre-defined bins. The advantage with respect to a purely random sampling is that stratified sampling gives a similar sampling accuracy with, at least, one order of magnitude less members. To model the source term we used the Buoyant Plume Theory option of FALL3D (e.g. Bursik, 2001; Folch et al., 2009), which
proportionate the Mass Eruption Rate (MER) and the vertical distribution of mass given the height of the eruptive column and the conditions at the vent. Regarding the TGSD, we assumed a Gaussian distribution with 10 particle-size bins from − 4Φ to 5Φ and a standard linear dependency of particle density with diameter (values of 1200 and 2500 kg/m 3 for the two end-members). Particle sphericity was considered constant and equal to 0.9. The TGSD varied for each simulation depending on the sampled value of the mean particle size. For LMS and MMS, the FALL3D computational domain was similar to the WRF-ARW inner nested domain, with a vertical resolution of 250 m and extending upwards up to 8 and 12 km above terrain for the LMS and MMS respectively. For HMS, the computational domain was similar to the intermediate WRF-ARW domain, with a vertical resolution of 1 km and 30 vertical layers. 3.4. Elaboration of hazard maps Probabilistic hazard maps give the probability that, at each spatial point, a threshold value is exceeded during an interval of time (e.g.
Author's personal copy C. Scaini et al. / Journal of Volcanology and Geothermal Research 219-220 (2012) 41–51
47
Fig. 6. Airborne ash concentration hazard maps for MMS at FL050 (left) and FL150 (right). Contours give the probability to achieve values of 0.2 (top) and 2 (bottom) mg/m3 respectively.
Renschler, 2005). The temporal component is not considered here because the frequency of the eruptive scenarios is uncertain (obtained probabilities should therefore be scaled by the eruption occurrence probability). Concerning thresholds, we considered tephra load values (in kg/m 2) and airborne ash concentrations at critical flight levels (see Table 3). Usually, thresholds are chosen in order to represent a critical condition, e.g. road disruption, collapse of a building, etc. We defined critical load values following the building classification proposed by Spence et al. (2005a, 2005b), who provided empirical vulnerability curves for several types of building typologies. A simple survey on the island reveals that a high percentage of buildings and structures have roofs of weak resistance and are poorly maintained (Fig. 2). For this reason, we adopted a precautionary approach for the choice of load thresholds and consider that values as low as 50 and 100 kg/m2 could lead to total or partial collapse of a fraction of weaker and stronger structures respectively. Concerning values for road traffic disruption, we considered a value of 1 kg/m 2, corresponding roughly to 1 mm of deposit thickness. Finally, for airborne ash concentration, we assumed thresholds of 0.2 and 2 mg/ m 3, which correspond to the restricted and no-fly zones defined in Europe during the 2010 Eyjafjallajökull eruption in Iceland. Hazard maps for civil aviation purposes were computed at flight levels (FL) of FL050, FL150, and FL300 (for HMS only), corresponding to 5000, 15,000, and 30,000 feet of nominal pressure. These FL correspond to typical landing/taking-off and cruise altitudes. Finally, we also focused on particular localities like the towns of Moyogalpa, Altagracia, and San Jose (in the Ometepe Island), and Rivas and Granada, the
largest two cities of the region. For MMS and HMS, hazard maps were computed performing two simulations per day (eruptions assumed to start at 06:00 and 18:00 LT) during the whole TMY, i.e. 2x365 = 720 simulations for each scenario. In contrast, for the LMS we did not perform the full hazard assessment. Few simulations for this scenario already showed that the consequences of this event are always constrained to the volcano flanks and cause minor effects in the nearby communities. Hazard assessment for the LMS is discussed in the next section on a semi-quantitative basis. 4. Results 4.1. Hazard assessment for LMS Figs. 3 and 4 show typical results for the LMS. Critical ground load thresholds (Fig. 3) are reached only at the upper flanks of the volcano, with a maximum load of 1 kg/m 2 (i.e. 1 mm of deposit thickness roughly). Some km further, around the village of San José del Sur, ground load values decrease to 0.1 kg/m 2 or less. Clearly, such low accumulations do not threat any structure or building. On the other hand, airborne ash concentration thresholds are exceeded only at the proximities of the volcano and decrease rapidly in few hours (Fig. 4). For example, at FL050, the peak of concentration >2 mg/m 3 is very localized and of short duration, of about 3 h only. At higher atmospheric levels (e.g. FL150), concentration values are always below 2 mg/m 3 and remain confined within a relatively small area. All the simulations performed for the LMS lead to very similar quantitative
Table 4 Probability (in %) to exceed thresholds of tephra ground load and airborne ash concentrations at relevant locations (MMS scenario). Ground load
San Jose del Sur Esquipulas Moyogalpa Rivas Granada Managua Airport
Concentration at FL050
Concentration at FL150
1 kg/m2
50 kg/m2
100 kg/m2
0.2 mg/m3
2 mg/m3
0.2 mg/m3
2 mg/m3
30 73 52 26 ~0 ~0
1 3 2 ~0 ~0 ~0
~0 ~0 ~0 ~0 ~0 ~0
64 87 91 66 10 8
50 80 80 50 5 3
52 61 76 45 16 12
45 51 63 33 8 5
Author's personal copy 48
C. Scaini et al. / Journal of Volcanology and Geothermal Research 219-220 (2012) 41–51
Fig. 7. Tephra ground load hazard maps for HMS. Contours give the probability to achieve values of 1 (top), 50 (middle), and 100 (bottom) kg/m2 respectively.
results, showing that this event poses little threat to communities and regional air traffic. 4.2. Hazard assessment for MMS Fig. 5 shows the tephra ground load hazard maps obtained for thresholds of 1, 50, and 100 kg/m 2. Ground loads equal or larger than 1 kg/m 2 have a probability equal or lower than 50% to occur in the Western coast of Nicaragua Lake and less than 5% in the Pacific coast. The probability to reach critical values for road disruption is substantial in the area around Rivas municipality (≈25%). Ground loads equal or larger than 50 kg/m 2 are constrained at the Ometepe Island and have a probability b5% to occur in the main villages located in the western shore. Finally, the probability to accumulate >100 kg/ m 2 is substantial only in the central area of the island and negligible in the most populated coastal areas. Fig. 6 shows the probability of having critical ash concentration values at FL050 and FL150. According to results, the restricted-fly zone (0.2 mg/m 3) has a probability of 70% at FL050 and of 30% at FL150 to cover the whole island and
the Western rive of the lake. The probability of having a restrictedfly zone at the Managua international airport is of around 10%. Concerning the no-fly zone criteria (2 mg/m 3), probabilities are also very high (70%) at Ometepe and at the Pacific coast. However, the no-fly zone probability at the region of Managua is less than 5% for all fly levels. Table 4 reports the probabilities of exceeding the threshold values at some relevant locations. 4.3. Hazard assessment for HMS Probabilistic ground load hazard maps for thresholds of 1, 50, and 100 kg/m 2 are shown in Fig. 7. Tephra load values are likely to reach critical values for road disruption over a wide area, including the Pacific coast (>50% probability) and the cities of Granada (>30% probability) and Managua (≈25% probability). Load values of 50 and 100 kg/m 2 are very likely (>50%) in the whole Ometepe Island and along the Western shore of the lake and, even if low, are not zero in Granada (≈5% probability) and Managua (≈1% probability). Hazard maps at fly levels FL050 and FL150 are shown in Fig. 8. In
Author's personal copy C. Scaini et al. / Journal of Volcanology and Geothermal Research 219-220 (2012) 41–51
49
Table 5 Probability (in %) to exceed thresholds of tephra ground load and airborne ash concentrations at relevant locations (HMS scenario). Ground load
San Jose del Sur Esquipulas Moyogalpa Rivas Granada Managua Airport
Concentration at FL050
Concentration at FL150
1 kg/m2
50 kg/m2
100 kg/m2
0.2 mg/m3
2 mg/m3
0.2 mg/m3
2 mg/m3
100 99 99 97 40 26
99 98 91 56 7 3
98 91 79 46 4 1
100 100 100 99 77 71
100 100 100 99 65 57
100 95 97 86 74 65
100 95 93 75 55 50
this scenario, the affectation is dramatic at regional level. The whole central part of the country, including the capital, has a very high probability (around 70%) to become a restricted-fly zone. The extent of the no-fly zone (2 mg/m 3) is also considerable for the 50% probability contour. The probabilities of exceeding the threshold values at some relevant locations for scenario 3 are shown in Table 5.
4.4. Seasonal influence Hazard maps computed over a TMY provide a global picture and allow to assess the expected impacts of a given scenario. However, it is important to consider also what may happen during specific seasons. Climate in Nicaragua is characterized by two distinct seasons, the rainy season (from May to November) and the dry season (the rest of the year). Typically, these two periods present very different upper atmosphere wind regimes in terms of both wind speed and predominant directions. When probabilistic maps are computed over the whole TMY, this seasonal variability is masked. On the other hand, tephra fallout during the rainy season becomes more hazardous because water increases the deposit load (Macedonio and Costa, submitted for publication) and the probability of having landslides and lahars at the volcano flanks. Fig. 9 shows seasonal tephra ground load hazard maps for MMS and HMS. For MMS, differences are not substantial. This is because winds in the lower atmosphere are quite uniform during the whole year. However, the situation is completely different for HMS, reflecting the very distinctive
predominant wind directions in the upper atmosphere during wet and dry seasons. 5. Summary and discussion We computed probabilistic hazard maps for Concepción volcano considering three different scenarios. Mean values of the parameters characterizing each scenario are based on past reference events (2010, 1977 and 2720 B.P.). We also accounted for uncertainty assuming a PDF for each parameter. Values for volcanological inputs of the dispersal model were sampled using a stratified sampling technique and meteorological inputs were furnished during TMY defined for each scenario. We run the WRF-ARW mesoscale meteorological model to generate the meteorological dataset over the TMYs and the FALL3D dispersal model to simulate tephra dispersion and fallout. Results of 720 simulations have been combined to produce probabilistic hazard maps for threshold values of ground load and airborne ash concentrations. Main results can be summarized as follows: • LMS (2 km mean column height, VEI ≈ 1) poses little threat on communities and air traffic. Expected impacts of fallout limit to minor road traffic disruption in areas close to the volcano. No substantial tephra accumulation is expected in the main villages of Ometepe. Critical airborne ash concentration values are also very localized and have a short duration. • For the MMS (5 km mean column height, VEI ≈ 2–3) results indicate that road traffic disruption can be expected over the whole
Fig. 8. Airborne ash concentration hazard maps for HMS scenario at FL050 (left) and FL150 (right). Contours give the probability to achieve values of 0.2 (top) and 2 (bottom) mg/ m3 respectively.
Author's personal copy 50
C. Scaini et al. / Journal of Volcanology and Geothermal Research 219-220 (2012) 41–51
Fig. 9. Seasonal tephra ground load hazard maps. Contours give the probability to achieve values of 1 kg/m2 for dry (left) and wet (right) seasons. Results are for MMS (top) and HMS (bottom) scenario. Note how the seasonal influence is negligible for MMS and, in contrast, very relevant for HMS. See text for details.
Ometepe Island and in the mainland, with effects extending across the provinces of Rivas and Granada. However, the probability of tephra fallout is very low at Managua and at the international airport (b5%). Tephra deposition could damage crops and pastures and affect the mainland province of Rivas, where agriculture is the main economical activity. Results suggest that collapse of buildings and structures would be very localized in the volcano flanks, affecting mostly sparse buildings. However, it is worth to mention that MMS threats large parts of the island, not because of copious tephra fallout, but by remobilization of deposits emplaced at the volcano flanks during eventual strong rain episodes. For this scenario, we found very weak seasonal influence, as shown in Fig. 9. • HMS (20 km mean column height, VEI ≈4) is likely to produce generalized disruption of ground transportation systems in the whole central part of Nicaragua. This situation could easily lead to a north– south partition of the country, with consequent damages to the main economical activities, including agriculture and commerce. Emergency response operations should be prepared to cope with this situation. Collapse of buildings and structures would be massive, especially in the island of Ometepe and in the Rivas province and, to a lesser extent, also in Granada and Managua. For this scenario there is a strong seasonal influence caused by the variability of the prevailing winds in the upper atmosphere, as shown in Fig. 9. Managua International airport is the only international airport in Nicaragua and one of the busiest in Central America. We have assessed the expected impacts of each scenario over air traffic (Tables 4 and 5). The probability of declaring a no-fly zone (airborne ash concentration of 0.2 mg/m 3) over most of the Nicaraguan airspace is not negligible for MMS (up to 5%) and very high for HMS (up to 57%). For MMS, the airspace above Rivas and Granada should probably be closed and the operability of the Managua airport may be reduced due to ash at low FL impeding landing and take-off operations. However, ash clouds are typically dispersed in hours and, in practical terms, the crisis could be managed with flight re-routing and reallocation. In contrast, the airspace above Managua International Airport presents a high probability of affectation during the HMS, with a critical situation lasting from days to weeks, causing serious disruption of the air transportation system. Ash clouds can also
spread over the Pacific Ocean during the dry season and over the Caribbean during the wet season (results not shown). The scenarios we defined are characterized by different magnitudes of the eruption parameters and the implications for hazard assessment and risk management depend on such differences. In any case, hazard maps developed for Concepción volcano underline the importance of adopting specific mitigation measures for vulnerability reduction. This is important for the definition of a strategic plan for sustainable land development and to improve medium and longterm risk assessment in this region. Acknowledgements This work has been funded by Spanish Research Project “Atmospheric transport models and massive parallelism: applications to volcanic ash clouds and dispersion of pollutants at an urban micro-scale” (ATMOST, CGL2009-10244) and the CYTED thematic network "Red Iberoamericana para el monitoreo y modelizacion de cenizas y aerosoles volcanicos y su impacto en infraestructuras y calidad del aire” (CYTED, 410RT0392). Simulations have been done at the Barcelona Supercomputing Center (BSC-CNS) facilities using the MareNostrum supercomputer. We acknowledge the constructive reviews from A. Costa and an anonymous reviewer. We also thank H. Delgado for his important suggestions that have improved the manuscript and O. Corominas for providing the picture of the April 2010 Concepción eruption. References Baxter, P.J., 1999. Impacts of eruption on human health. In: Sigurdsson, H. (Ed.), Encyclopedia of Volcanoes. Academic Press, New York, pp. 1035–1043. Bonadonna, C., Mayberry, G.C., Calder, E.S., Sparks, R.S.J., Choux, C., Jackson, P., Lejeune, A.M., Loughlin, S.C., Norton, G.E., Rose, W.I., Ryan, G., Young, S.R., 2002. Tephra fallout in the eruption of Soufriere Hills Volcano, Montserrat. Geological Society, London, Memoirs 21, 483–516. Borgia, A., Van Wyk de Vries, B., 2003. The volcano-tectonic evolution of Concepción, Nicaragua. Bulletin of Volcanology 65, 248–266. Bulut, H., 2010. Generation of representative solar radiation data for Aegean Region of Turkey. International Journal of the Physical Sciences 5, 1124–1131. Bursik, M., 2001. Effect of wind on the rise height of volcanic plumes. Geophysical Research Letters 28, 3621–3624. Casadevall, T.J., 1993. Volcanic hazards and aviation safety: lessons of the past decade. Flight Safety Foundation – Flight Safety Digest, 1–9 May 1993.
Author's personal copy C. Scaini et al. / Journal of Volcanology and Geothermal Research 219-220 (2012) 41–51 Costa, A., Macedonio, G., Folch, A., 2006. A three-dimensional Eulerian model for transport and deposition of volcanic ashes. Earth and Planetary Science Letters 241, 634–647. Delgado-Granados, H., Navarro, M., Farraz, I., Alatorre Ibargüengoitia, M.A., Hurst, A.W., 2006. Hazard Map of Concepción Volcano (Nicaragua). Fourth Conference Cities on Volcanoes IAVCEI, 23–27 January, Quito, Ecuador. Finkelstein, J.M., Schafer, R.E., 1971. Improved goodness of fit tests. Biometrika 58, 641–645. Folch, A., Sulpizio, R., 2010. Evaluating long-range volcanic ash hazard using supercomputing facilities. Application to Somma-Vesuvius (Italy), and consequences on civil aviation over the Central Mediterranean Area. Bulletin of Volcanology 72, 1039–1059. doi:10.1007/s00445-010-0386-3. Folch, A., Costa, A., Macedonio, G., 2009. FALL3D: a computational model for transport and deposition of volcanic ash. Computers & Geosciences 35, 1334–1342. Guffanti, M., Mayberry, G.C., Casadevall, T.J., Wunderman, R., 2009. Volcanic hazards to airports. Natural Hazards 51, 287–302. Janjai, S., Deeyai, P., 2009. Comparison of methods for generating typical meteorological year using meteorological data from a tropical environment. Applied Energy 86, 528–537. Johnston, D., Becker, J., 2001. Volcanic ash review - Part 1: impacts on lifelines services and collection/disposal issues. Auckland Regional Council Technical Publication, No. 144. Kalogirou, S.A., 2003. Generation of typical meteorological year (TMY-2) for Nicosia, Cyprus. Renewable Energy 28, 2317–2334. Kutterolf, S., Freundt, A., Pérez, W., Wehrmann, H., Schmincke, H.U., 2007. Late Pleistocene to Holocene temporal succession and magnitudes of highly-explosive eruptions in west-central Nicaragua. Journal of Volcanology and Geothermal Research 163, 55–82. Lam, J.C., Hui, S.C.M., Chan, A.L.S., 1996. A statistical approach to the development of a Typical Meteorological Year for Hong Kong. Architectural Science Review 39, 201–209. Macedonio, G., Costa, A., submitted for publication. Brief Communication: Rain effect on the load of tephra deposits. Natural Hazards and Earth System Sciences.
51
Mastin, L.G., Guffanti, M., Servranckx, R., Webley, P., Barsotti, S., Dean, K., Durant, A., Ewert, J.W., Neri, A., Rose, W.I., Schneider, D., Siebert, L., Stunder, B., Swanson, G., Tupper, A., Volentik, A., Waythomas, C.F., 2009. A multidisciplinary effort to assign realistic source parameters to models of volcanic ash-cloud transport and dispersion during eruptions. Journal of Volcanology and Geothermal Research 186, 10–21. doi:10.1016/j.volgeores.2009.01.008. Michalakes, J., Chen, S., Dudhia, J., Hart, L., Klemp, J., Middlecoff, J., Skamarock, W., 2001. Development of a Next Generation Regional Weather Research and Forecast Model. In: Zwieflhofer, Walter, Kreitz, Norbert (Eds.), Developments in Teracomputing: Proceedings of the Ninth ECMWF Workshop on the Use of High Performance Computing in Meteorology. World Scientific, Singapore, pp. 269–276. Rao, C.R., Krishnaiah, P.R., 1994. Handbook of Statistics. Volume 6: Sampling. In: Rao, C.R., Krishnaiah, P.R. (Eds.), Elsevier/North-Holland, Amsterdam, pp. 125–145. Renschler, C.S., 2005. Scales and uncertainties in using models and GIS for volcano hazard prediction. Journal of Volcanology and Geothermal Research 139, 73–87. Siebert, L., Simkin, T., 2002. Volcanoes of the world: an illustrated catalog of Holocene volcanoes and their eruptions. Smithsonian Institution, Global Volcanism Program Digital Information Series, GVP-3http://www.volcano.si.edu/world/. Spence, R.J.S., Kelman, I., Baxter, P.J., Zuccaro, G., Petrazzuoli, S., 2005a. Residential building and occupant vulnerability to tephra fall. Natural Hazards and Earth System Sciences 5, 477–494. Spence, R.J.S., Kelman, I., Calogero, E., Toyos, G., Baxter, P.J., Komorowski, J.C., 2005b. Modelling expected physical impacts and human casualties from explosive volcanic eruptions. Natural Hazards and Earth System Sciences 5, 1003–1015. Stewart, C., Johnston, D.M., Leonard, G.S., Horwell, C.J., Thordarson, T., Cronin, S.J., 2006. Contamination of water supplies by volcanic ashfall: a literature review and simple impact modelling. Journal of Volcanology and Geothermal Research 158, 296–306. Yang, H.X., Lu, L., Burnett, J., 2003. Weather data and probability analysis of hybrid photovoltaic-wind power generation systems in Hong Kong. Renewable Energy 28, 1813–1824.
