The health and environmental impacts of the June 2011 Puyehue-Cordón Caullé volcanic complex eruption

May 2012
Thomas Wilson1, Carol Stewart1, Heather Bickerton1, Peter Baxter2, Valeria Outes3, Gustavo Villarosa3, Elizabeth Rovere4
1: University of Canterbury, Christchurch New Zealand
2: University of Cambridge, Cambridge, United Kingdom
3: INIBIOMA (CONICET-Universidad Nacional del Comahue), Bariloche, Argentina
4: SEGEMAR, Buenos Aires, Argentina
EXECUTIVE SUMMARY
Effects of the June 2011 eruption of the Puyehue-Cordón Caullé volcanic complex (PCC-VC) on urban infrastructure, agriculture and public health in Argentina were investigated by a multidisciplinary team from New Zealand, the United Kingdom and Argentina. The team also studied the emergency management of the eruption crisis. Field work was carried out between 27 February and 16 March 2012. The team focused its efforts on the population centres that received the greatest ashfalls (Villa la Angostura, San Carlos de Bariloche and Ing. Jacobacci), and on agricultural regions affected by ashfall.
A notable feature of this eruption (and other eruptions in Patagonia) has been the extensive remobilisation of unconsolidated ash deposits by the strong prevailing winds in the region. Towns on the steppe such as Jacobacci have been particularly severely affected, and at the time of our visit, severe air quality problems were continuing to occur on a regular basis.
The public health response to the widespread ash fall has not been optimal and was marked by inadequate information being supplied to the population about possible short and long term health effects of exposure to the ash, the particle size and composition of the ash, and appropriate measurements of the concentration of ash in the air to assess risk. The most important single finding is the absence of crystalline silica (which can cause silicosis) in the respirable ash fraction (PM4), but further tests on a wider geographical range of samples are needed to fully confirm this result. However, a large fraction of PM2.5 is present in the ash and the ash deposits are very persistent in the arid areas leading to continuing, abnormally high exposure to PM2.5 and PM10 in the settlements in the steppe region, which is of great concern as it could carry long term risks to respiratory health, such as chronic obstructive pulmonary disease (COPD), and reduction in lung growth in children, as well as aggravation of asthma in adults and children.  A programme of background air pollution assessment and monitoring of ash levels inside houses and in individual groups of people at risk needs to be conducted in the steppe region as a matter of urgency, together with epidemiological research and respiratory disease surveillance.  Special measures to reduce the infiltration of ash into homes and buildings such as schools during the frequent episodes of strong winds continue to be needed and should be more widely adopted.  Methods to reduce the movement of ash by the winds and effective cleanup of ash from streets and around homes need to be actively pursued and strengthened. 
The eruption had severe effects on extensive livestock farming in the mallines in the Comallo to Jacobacci region, with stock losses of 40-60%. Causes of death included starvation, dehydration, rumen blockages and abrasion damage to teeth. The effects of the eruption were compounded by the pre-existing drought conditions in the region. Effects were comparatively less severe in farming areas closer to the volcano (with stock losses of around 25%), despite this region receiving much greater ashfalls. We have suggested some strategies for agricultural recovery based on our experiences of other eruptions in Patagonia.
Generally, impacts on electricity networks, water supplies, wastewater treatment, communications networks and road transport were as expected from our studies of other eruptions. Probably the most severe and ‘unexpected’ disruption was that Bariloche airport remained largely out of action well beyond its official closure period of one month for cleanup operations . This was due to caution exercised by the major airlines in the face of continued activity at PCC-VC, and uncertainty about safe operating thresholds. Flight disruptions are likely to have had substantial effects on the economy of the region although this was beyond the scope of our study to evaluate.
In terms of emergency management of the eruption crisis, it was evident to us that there is a dedicated and capable group of emergency and infrastructure managers who have worked hard to respond to and recover from the eruption crisis.  However these individual efforts were hampered by poor overall coordination between local, regional and national levels of civil protection and poor liaison with scientific agencies and with volcano observatories in Chile. Given the frequency of eruptions from Andean volcanoes, it is vital that the lessons learned be embedded to reduce the vulnerability of this area to future eruptions.
1.0    INTRODUCTION OF THE RESEARCH TEAM AND OBJECTIVES
A field team travelled to Chile and Argentina over the period 27 February to 16 March 2012, to investigate the impacts and consequences of the June 2011-present eruption of the Puyehue-Cordón Caullé volcanic complex (hereafter referred to as PCC-VC), Chile. Specific objectives of the visit were to investigate:
impacts of the eruption on critical infrastructure in urban areas;
impacts on agriculture including impacts on livestock health, evacuation, adaptations in farming practices and timescales of recovery and rehabilitation of farmland;
the emergency management of the eruption crisis and identify important lessons learned;
the phenomenon of remobilisation of ash deposits by wind and fluvial action; and
public health implications of the eruption.
Members of the field team were Dr Tom Wilson, Dr Carol Stewart and Heather Bickerton, of the University of Canterbury, New Zealand, and Dr Peter Baxter, of Cambridge University, UK, who joined the team between 27 February and 5 March. The team received a great deal of assistance from local collaborators from  INIBIOMA (CONICET and Universidad Nacional del Comahue), Bariloche (Dr Gustavo Villarosa and Lic. Valeria Outes). Dr Elizabeth Rovere of SEGEMAR organised Dr Peter Baxter’s schedule and identified many local contacts.  This work is supported by a wider group of colleagues in New Zealand, the United States and the UK, from the following organisations: GNS Science, Massey University, Durham University and the US Geological Survey. Review comments on this report were provided by Dr John Ewert, USGS; Professor David Johnston, Massey University/GNS Science, New Zealand; and Dr Richard Smith, Ministry of Civil Defence and Emergency Management, New Zealand.
The overall goal of our research is to reduce vulnerability to volcanic eruptions; specific aims are to improve understanding of the impacts of eruptions (and in particular, volcanic ashfall, which is the most widely distributed product of explosive eruptions) on human health, agriculture and critical infrastructure and to move towards identifying best practices for emergency management.
We have undertaken volcanic impact assessments in a range of countries worldwide, including previous work in Patagonia following the 1991 Hudson and 2008 Chaitén eruptions. We work within the auspices of the Volcanic Ash Impacts Working Group and the International Volcanic Health Hazard Network (IVHHN)1.
Contact details:
carol.stewart@canterbury.ac.nz
thomas.wilson@canterbury.ac.nz
heather.bickerton@pg.canterbury.ac.nz
pjb21@medschl.cam.ac.uk
valeria.outes@gmail.com
villarosag@comahue-conicet.gob.ar
eirovere@gmail.com

2.0    BRIEF OVERVIEW OF THE PUYEHUE-CORDÓN CAULLÉ ERUPTION
On 27 April 2011 a swarm of volcano-tectonic earthquakes was detected, centred on the Cordón-Caullé fissure zone2. These earthquakes continued to increase in magnitude and frequency until Saturday 4 June when the current eruption sequence began. A 5 km-wide ash and gas plume rose to 12.2 km height, accompanied by pyroclastic flows. By the early hours of Sunday 5 June, ash falls had occurred in the following population centres (Figure 1):
Approximately 15-17 cm of coarse ash fall was received in Villa la Angostura, 54 km ESE of the vent;
Approximately 3-4.5 cm of medium to coarse tephra (up to xx 6 mm-sized for particularly flat particles, 3-4 mm for the more spherical pyroclasts) was received in the city of San Carlos de Bariloche, located 100 km SE of the vent;
Approximately 5 cm of fine ash was received in Ing. Jacobacci, located 240 km ESE of the vent on the steppe.

Figure 1 Study area location showing ash isopachs (cm), population centres visited and sampling sites.
By the 7th June the pumice and ashfall had moved further north to deposit up to 2 cm on San Martin de los Andes, 80 km to the NE. Buenos Aires city was covered by a thin layer of ash (Botto et al., 2012), since 5th June.  The Ezeiza international airport remained closed for more than 15 days, as well as local airports of more than 10 cities, including Uruguay and South Brazil. The diffuse ash plume circumnavigated the globe by 16 June. Ash and gas plumes continued to erupt from the fissure with heights up to 13 km, reducing to a few kilometres by early July, and causing continuing disruptions to flight schedules in New Zealand, Australia and South Africa over this period.
2.1 Variable ash characteristics
Variation in grainsize distributions from samples collected in Villa La Angostura during June 2011 reflects the variations of eruptive parameters, direction and intensity of winds. The coarse-grained samples collected during the first four days of the eruption are mainly composed of lapilli and coarse ash, the main fraction ranging between 500 µm-1,5 mm and less than 1 weight % <63 µm.  Fine-grained ash collected in June 8 presents 25 weight% of particles 500 µm in diameter and 30 weight % <63 µm. Results for the sample collected during the ashfall registered during June 13-14 indicate that almost 40 vol.% of the material is  <63 µm in diameter,  and ~12 vol. % corresponds to PM10 and 5vol.% of PM4.
For the coarse–grained samples from Bariloche collected during 4 June size diameter peaked between 375-750µm with less than 2% of material <63µm. Grain-size measured by laser diffraction in the sample from the ashfall occurred in June 11 ranges between 0,5-650µm, with ~53vol.% <63 µ in diameter, 12.56 vol% of PM10 and 5vol.% of PM4.
The amount of respirable material (PM4 and PM2,5) in fine-grained ash samples from medium to distal locations collected during June 2011, measured by laser diffraction (IADO CONICET-UNS) ranges from 3.5-7.4 vol.% PM4 and from 2 to 4.6 vol.% PM2.5. Exceptions to these ranges are reported for samples collected during the first days of the eruption in LS where contents of respirable fraction were substantially higher, varying between 12-30 vol.% PM4 and 7-16 vol.% PM2.5. Samples from these same locations collected during the last week of June show contents within the range reported for the rest of the sites.
2.2 Remobilisation of ash deposits
Unconsolidated ash deposits are at risk of remobilisation by wind and water processes. There has been extensive remobilisation of ash by the strong prevailing westerly winds in the region, with towns on the steppe such as Jacobacci particularly severely affected. Schools in the region were closed for almost four months after the eruption, and during windy conditions, townspeople have little choice but to remain indoors.
Fluvial remobilisation of thick ash deposits is likely to be highest in proximal areas with steep terrain. Infrastructure at risk may include roads, bridges and inhabited areas such as Villa la Angostura. Poor consolidation of ash in arid areas in the steppe also poses a risk of fluvial remobilisation. An intense rainfall event occurred on 8 March 2012 and caused extensive remobilisation of surface material in the form of mudflows that damaged roads and inundated farmland in the Comallo area. At the time of writing, it was not known to what extent the remobilised material consisted of recently deposited ash versus other surface material, but testing underway should resolve this matter.
3.0    OBSERVATIONS
The following sections describe our observations on impacts of the eruption on public health, pastoral farming and infrastructure. We also describe our observations on cleanup operations in the different population centres, and on the emergency management of the eruption crisis.
3.1    Public health consequences of the eruption
    3.1.1 Airborne ash hazards
Managing the health aspects of volcanic eruptions is an important part of the immediate emergency response and in the phase of restoring impacted areas to normality.  Public health measures regarding volcanic ash should be an integral part of crisis management and these are discussed in more detail in Appendix 1.
The measures should include the rapid dissemination of information on the nature of the ash and its potential for causing short- and long-term respiratory problems, as well as the types of preventive measures individuals and the community can take to reduce exposure.   In all explosive eruptions a significant proportion of the ash is very fine and capable of being breathed into the deep air sacs (alveoli) of the lungs, and it is these particles of fine ash which can be readily resuspended in the air from ground deposits during dry periods and remain in the air for hours or days afterwards. 
Particulate matter 10 µm and less in diameter is called PM10 and is routinely measured by air monitoring stations in cities in Europe and the USA to ensure that the concentrations of particulate matter (PM) in the ambient air do not exceed statutory limits for specific time periods, usually 24 hours.   After ash falls, the levels of PM10   can far exceed these limits until the rain clears the air, and then the levels will go up again whenever it is dry and windy.   The people who are most affected by this exposure to PM are those who already suffer from some type of chest problem, like asthma or chronic bronchitis, and they notice that their symptoms can get worse on windy days, especially when they go outside.  Serious acute problems in such patients are fortunately uncommon, even when ash is visible in the air and the levels of PM10 are high, but they may need to increase their medication. 
Much concern may arise over long term effects to health if the fine ash remains in the environment for long periods of time.  An important issue with volcanic ash from Andean volcanoes is the need to check for the presence of crystalline silica in the PM, as this mineral has the potential to cause silicosis, a form of fibrosing or scarring of the lungs, in previously healthy people.  Fortunately, ash analyses so far suggest that this mineral is not present in dangerous amounts but, as the ash has fallen over a wide area of Argentina in a prolonged eruption, more analyses should be undertaken to confirm this finding. 
Very high exposure to PM is presently happening in the Patagonian steppes and there will be concern there that such persistently high exposures could eventually impair the respiratory health of healthy children and adults.  The following research is recommended to monitor the risk to the health of people living there:
Routine data collection of mortality statistics and hospital/clinic attendances should be reviewed and monitored in the ash fall areas.  Tuberculosis and childhood pneumonia need to be included in the monitoring of poor rural areas.
A study of asthma and lung function in school children should be considered as a priority.  School children in Linea Sur settlements are a high exposure group and warrant long term follow up; low to medium exposure groups for   comparison could be chosen from the towns in the tourist areas where they received much more rain.
A study of respiratory symptoms and lung function in a group of adults, e.g., farmers, in the Linea Sur and a comparison group in rural areas where there is more rainfall.  A study cohort could be established for long-term follow-up.
Various objective variables for study could include inflammatory markers for systemic inflammation, as a sign of high exposure to fine ash and evidence of potential risk of cardio-vascular disease impact. Psychological stress (anxiety and depression) was reported to us across the ash fall out zone, often due to uncertainties about the evolution of this long lived eruption and the deep economic impacts.  It was an acute impact in some rural areas.  This also warrants further scientific investigation.
Eye symptoms have been common and in some cases severe enough to require hospital admission: further study is needed as ocular effects of this severity due to volcanic ash are very unusual.
The risks to health in high and low impacted areas needs to be regularly evaluated by monitoring background exposure in the community and personal exposure in individuals:
Particle air monitoring stations (as installed in major EU and USA cities) should be established in the study areas to record hourly and daily means of background PM10 and PM2.5..  These require a technician’s support and are costly.  The results can be automatically transmitted to a central office. 
Hand-held monitoring instruments like the DustTrak should be used to check PM levels in the outside air in towns and small settlements, and indoors in schools and houses to measure the effectiveness of sealing buildings against the ingress of ash.  They can be set to record for long periods and downloaded.
Exposures to PM in outdoor workers, or direct measures of exposure in other individuals, can be measured using instruments such as the Sidepak, which is a small device that works like the DustTrak, but fits on to a belt.
An exposure survey using the DustTrak and Sidepak instruments should be undertaken in the worst affected towns, such as Comallo and Jacobacci, to establish the current mean background levels of PM10 and PM2.5 in the ambient air under a range of representative weather conditions, and to define the much higher mean personal exposures of children and representative groups of adults.  See the methodology used on Montserrat by Alison Searl and others (Occup. Environ. Med. 2002; 59; 523-531).
Details of the instruments that will directly read out and record PM levels can be supplied on request.
3.1.2 Drinking water hazards
Following a volcanic eruption it is very common for there to be a high level of concern among both the public and civil authorities about chemical contamination of water supplies. These concerns may be partially allayed by characterisation of the soluble salt content of the ash to enable an assessment of hazards from leachable elements such as fluorine, for drinking water supplies. A sample of PCC-VC ash collected on 4 June in Bariloche was submitted for leachate testing at the INVAP laboratories. However, unfortunately the method used was a non-standard one for leachate characterisation, thus the results are of limited value for hazard assessment. Advice on recommended methods for volcanic ash leachate characterisation can be found on the website www.ivhhn.org. Specific protocols are being developed for this purpose by an international working group.
According to one interviewee, over 500 surface water analyses were carried out following the eruption. No results of concern were found, either for human health or aquatic life. The most notable feature was high levels of suspended solids.
In most cases the physical presence of ashfall in water treatment systems will overwhelm any chemical contamination problems. Ashfall can enter treatment plants and is likely to damage pumping equipment (particularly impellers and motors) and block sand filters, requiring a greatly increased maintenance schedule (see Section 3.2.2 for summary of eruption impacts on municipal water supplies in the study areas).
Of particular concern for public health is that suspended ash (turbidity) can provide a growth substrate for microorganisms, and can protect them from disinfection, thus increasing the risk that the population may be exposed to waterborne diseases. It is also important to note that unless a high quality water source is used, such as secure groundwater, standard disinfection methods alone are unlikely to guarantee that drinking water is safe from protozoa such as Giardia and Cryptosporidium (which are relatively resistant to chlorination and which require additional removal steps such as filtration).
We received many word-of-mouth reports about a high incidence of gastrointestinal disease in both Bariloche and Villa la Angostura. The provincial health authorities are well aware of an ‘endemic corridor’ of acute diarrhoea in Neuquen province that has occurred for several years. Causes are likely to be multiple and complex, including poor sanitation, contaminated food or drinking water and underlying health issues. A 2008 report from the Departamento de Epidemiología-Dirección de Epidemiología y Estadística, Subsecretaría de Salud-Provincia del Neuquén identified the bacterial agent Shigella as a common causative agent; Shigella is an indicator of deficiencies in environmental sanitation and the presence of untreated sewage.
Our observations were that some of the water supply systems for small communities have an insufficient level of treatment to have confidence that the water is microbiologically safe, particularly for stream-fed systems. In the town of Villa la Angostura, four out of eight drinking (tap) water samples collected on 11 July 2011 by staff of the Bromatologia Municipal were assessed as unfit for human consumption (see Table 1 in Appendix 2). However it should be noted that these problems are likely to predate the eruption; local staff have been well aware of deficiencies in individual treatment systems since March 2009, and have made suggestions for improvements.
Recommendations
Urgent improvements are needed to water supply systems for some of the smaller communities in the area, particularly stream-fed systems, to ensure the safety of drinking water. In particular, improved filtration systems are necessary.
Monitoring of network water supplies should include, at a minimum, turbidity, pH, E.coli and chlorine residuals. Laboratories should be provided with the resources to carry out this work satisfactorily.

3.2    Pastoral farming
3.2.1    Impacts of the eruption
Pastoral farming style and production techniques vary widely in the depositional area of the ashfall, (Figure 1), from small, dispersed operations in parklands of Parque Nacional Nahuel Huapi, to extensive production models on the arid steppe. Thus the impacts of the ashfall, recovery paths and mitigation options are also variable. 
Agricultural land between Jacobacci and Bariloche received up to 5 cm of fine ash. Extensive, low-intensity sheep, cattle, horse and goat farming is concentrated in the mallines (grassland valleys) as the rest of the landscape is too dry. Preceding the ashfall there had been six years of drought with very low average rainfall (<100 mm/yr). This compounded the impacts of the ash and undoubtedly increased the livestock losses. Jacobacci municipality staff estimated that livestock losses after the ash fall were around 40-60% for a total regional herd of 225,000 sheep and 60,000 goats, due to starvation, dehydration, rumen blockages and tooth abrasion. Tooth abrasion led to further issues with foraging and grazing, causing further reliance on supplementary feed and premature aging of the animals. Fleece prices are also low in the region due to ash collecting in wool, and usable wool has dropped from 50-55% per fleece to 25-30%. Birth rates were also down from 60% to around 10-30% as mothers were malnourished and stressed. Continuing wind remobilisation of ash deposits is prolonging these impacts.
All of these factors have contributed to reduced income for farmers leading to most becoming reliant on supplementary feed supplied by federal government, municipalities and INTA (Instituto Nacional de Technología Agropecuaria). There have also been many cases of livestock being sold and moved to Buenos Aires, La Pampa and Chubut provinces (estimated at up to 100,000 animals).  The Jacobacci - Bariloche region has also been severely affected by continuing wind remobilisation of ash deposits, which is acting to prolong the impacts of the original ashfall.
Farming in Parque Nacional Nahuel Huapi is different, as it takes place on national park land and is dominantly cattle farming. Livestock numbers are limited; however the open style of farming means that farmers do not always know exact numbers as they may only see each animal once a year. The animal losses in this region were much lower despite the closer proximity to the volcano and greater ashfall depth, and were comparable to those experienced after a severe winter (around 25%). As in Jacobacci, the losses were sustained due to starvation, rumen blockages and tooth abrasion causing foraging difficulties, rather than any known chemical toxicity issues. The impact of the ashfall was lessened by the fact that in the early stages many farmers realised that there would be problems with access to feed, thus they commenced slaughtering their animals for their households, or selling animals before their condition worsened. Evacuation options in this region were very limited, because there is no road access to many of the farms, and boat access was limited by thick pumice rafts which formed on Lago Nahuel Huapi, and which made navigation on the lake difficult or impossible for up to 11 days.
For many farmers in this region, there are likely to be significant difficulties both in determining the best course of action in managing the effects of the eruption, and in finding the necessary resources to act.  However, we observed that local farming advisors and scientists are acutely aware of the situation on farms.  Impact assessment and research programmes are active in the area, aimed at quantifying where recovery efforts should be focused and what strategies might be most effective.  In general, efforts have been aimed at encouraging diversification of pastoral production systems to boost production and build resilience. 
3.2.2 Strategies for recovery
Due to the brief nature of our field visit to such a diverse region, it would be inappropriate to make sweeping recommendations for recovery.  Instead we have chosen to highlight successful strategies we observed in the affected area, and also from our investigations into other recent eruptions in Patagonia (see Appendix 3 for further details):
The provision of refuges for livestock to shelter from direct ash fall or remobilised ash is likely to be a successful strategy.  Some farmers in the Jacobacci area had already commenced building refuges prior to the eruption (typically small structures walled along the western edge to provide protection from the prevailing wind direction) and these proved to be a valuable asset. Many other farmers are now adopting the same approach.
In areas that received less than 15 cm ash fall, pasture recovery and stabilisation of the ash deposit will be aided by full cultivation of the topsoil (ideally to 20-25 cm depth).    However, farmers in the steppe region are not equipped for large scale cultivation or fertilisation. A possible solution could be the acquisition of shared community resources for cultivation.
Diversification into greenhouse cultivation, and the construction of windbreaks, were essential components of recovery for areas such as Puerto Ingeniero Ibañez, Santa Cruz province, which were subjected to wind-remobilised ash for years after the 1991 Hudson eruption. These were supported by government development subsidies. In these areas, investment in irrigation schemes was also instrumental in developing agricultural production and improved vegetation growth provided additional shelter from windblown ash.
There are lessons from previous eruptions in Patagonia which can help inform recovery strategies.  Efforts should be made to share resources with INTA and other relevant agencies from Santa Cruz and Chubut provinces, and also with the equivalent agencies in Chile. For example, a successful strategy for rebuilding livestock numbers in areas affected by thick ashfalls in Chile from the Chaiten eruption was the use of subsidies to prevent farmers from selling newborn calves. This helped promote the steady recovery of cattle numbers to pre-eruption levels.
3.3 Infrastructure
3.3.1 Electricity networks
The ashfall caused widespread disruption of electricity supplies in the study area. As we have observed for other eruptions, the predominant effect was ash contamination of distribution lines and substation insulators, which induced leakage current and insulator flashover6. Continual tripping of switches due to flashovers, combined with the presence of fine ash in switches, led to abrasion of the metallic conductors which reduced the contact between electrodes, in turn reducing their functionality.  This required ongoing replacement of the switches. Thermal generation facilities also suffered significant disruption in both Bariloche and Villa La Angostura, mainly due to ash blockage of air intakes. 
Bariloche experienced loss of electricity for >8 hours on 7 June before partial reinstatement.  Many sectors the city remained without power for >24 hours.  Intermittent cuts lasted through to the end of July.  The city is supplied from the national grid through a single transmission line and one grid exit point (GXP) substation. In this substation, coarse ashfalls of 3-4 cm caused flashovers to switching and bus-bars , cutting power to the town.  There was backup generation capacity available in the form of gas and diesel-powered generators. These were activated, but air intakes were rapidly blocked with ash, resulting in shutdown. This compounded the effects of the GXP disruption. Flashovers were also experienced across the town distribution network, causing local disruption.  Intermittent power supply in the city had significant cascading impacts on other infrastructure sectors such as the water treatment plant.
In Villa La Angostura, ashfalls of 15-17 cm caused insulator flashover on the 13.2 kV, 380 V and 220 V networks on a widespread and continuing basis. There were 20 days of electricity disruption immediately after the eruption, and the supply continued to be insecure for many months afterwards.  The town is not connected to the national grid, thus there was total reliance on the 6.1 MW thermal generation power plant.  This plant suffered problems with ash contamination of air intake filters for approximately a month afterwards on dry days, resulting in precautionary shutdowns following heavy ash falls.
At the time of our visit, issues had largely been resolved, and we observed some unique mitigation and adaptation strategies. In all locations, fire trucks were used to wash ash from contaminated equipment to reduce flashover risk.  However, ongoing ashfalls caused continual flashovers. This prompted the application of polypropylene jackets and shields to insulators in the most heavily populated areas in an attempt to reduce ash contamination. Despite some success with this method, best results were found by EPEN increasing the insulator pin length from 25 cm to 50 cm.   All 3,500 insulators in the town network were rapidly upgraded to this standard.
In Bariloche there has been significant investment in adding resilience to the power supply.  A 20 MW diesel plant has been installed to provide back-up generation capacity for the city.  Although this is still well below usual demand, (45-55 MW), this represents a significant improvement in power supply security for Bariloche, which should be applauded. The cascading impacts of power outages on other infrastructure sectors, as well as on commercial and residential customers, highlights the critical importance of a secure and reliable electricity supply.
Recommendations
A reliable electricity supply is essential for the functioning of modern society. While we were impressed with the ability of electricity providers to respond to impacts and to design effective adaptations, we suggest that there should be greater redundancy of electrical power supply across the study area.  Specifically,
The single grid exit point at Bariloche substation creates vulnerability for the city. Additional circuits within the substation would improve redundancy. Expansion of the back-up generator facility would similarly improve supply security.
Connection of Villa La Angostura to the national grid would offer greater supply security.
3.3.2    Water supplies
Each of the three different population centres studied experienced different impacts on their water supply networks. For public health-related recommendations, see Section 3.1.2.
For the Bariloche water treatment plant (WTP), the main challenge was that the plant is designed for a very low level of turbidity in the raw water source (0.2-0.4 NTU in Lago Nahuel Huapi) and thus an initial coagulation/flocculation step is not part of the treatment train. The eruption caused an increase in turbidity to 26 NTU, which was a significant challenge for the plant. Ash entered the plant via the intake, and caused abrasional damage to pump impellers as well as causing additional wear and tear on pump motors. Cleaning of ash from sand filters required a greatly increased level of maintenance, for several months. However, the only service interruption was caused by a city-wide power outage.
The Bariloche WTP provides around 80% of the city’s water supply, with the remainder coming from small springs and streams. These additional sources are utilised more during periods of high demand. Outlying settlements around Bariloche have their own self-contained water supply, treatment and distribution systems (juntas vecinales), which generally draw water from Lago Nahuel Huapi. Lake water and stream water is filtered and chlorinated; spring water is chlorinated without filtration. Staff interviewed from the Departamento Provincial de Agua (DPA) reported that in general, the eruption caused fewer problems for water supplies than in Villa la Angostura. One spring source was closed because of high levels of turbidity, but filtration systems for stream sources generally functioned well because treatment is designed for the higher levels of turbidity typical of storm flows (up to 400 NTU). Nonetheless, DPA staff reflected that while effects of the eruption for the main WTP were manageable, the event highlighted the vulnerability of the smaller treatment systems, and pointed to an increasingly urgent need to improve filtration systems.
In Villa la Angostura, the same general problems with damage to pumping equipment and clogging of intakes, pipes and sand filters were experienced. The town uses diverse sources for water supply: intakes from Lago Nahuel Huapi and Lago Correntoso, stream intakes and a well. The most severe problems were experienced with the stream-fed systems, which were inundated with ash and had to be cleared out manually. These systems still experience problems in rainy conditions. To help meet demand, a new 21-m deep well was dug and the water distributed in 10,000 litre tanks. Additional bottled drinking water was also supplied to neighbourhoods with stream-fed systems. See also the discussion in Section 3.1.2.
In Jacobacci, 17 wells are used as water sources. Pumphouses are enclosed, so the system proved resilient. The main challenge was in meeting water demand. The area was (and continues to be) severely affected by wind-remobilised ash. In lulls between these conditions, water demand would increase as the community cleaned up and dampened down ash in the streets, from normal usage of 1 million L/day to as high as 3 million L/day. It was a continual juggling act to meet these peaks in demand, and it was fortunate that the eruption did not occur in summer, when the demand is higher.
3.3.3    Wastewater treatment
The main problem experienced at Bariloche’s WWTP was that a large volume of ash entered the plant via sewer lines, even though storm drains and the sewerage network are in theory separated. The incoming suspended solid load almost doubled from usual level of 4500 mg/L to 8000 mg/L. It accumulated primarily in the biological reactor which is the heart of the system and cannot be taken offline. Some bypassing of partially treated sewage to the lake was necessary over the first few days, to keep the plant operational. The capacity of biological reactor has been reduced as a metre of ash has accumulated on the bottom of the 4.5 m deep tank. Management are still investigating options for clearing out this ash.
3.3.4    Transport
The most severe disruption to transport networks in the depositional area was the closure of Bariloche airport. The airport was closed at 17h00 on 4 June, just before the arrival of the ash plume. The airport did not receive an official warning, but was advised of the impending arrival of the plume by a pilot on an incoming flight. The airport was closed for one month (until 5 July) for the cleanup operations. During cleanup operations, approximately 1000 tonnes of ash were removed from the airport, of which most was disposed of by filling in hollows and depressions in the immediate area. A further initiative has been the installation of irrigation systems surrounding the runway, to encourage grass growth to trap the ash and suppress wind remobilisation.
Even though the airport re-opened for business on 5 July, it was many more months before the country’s two major airlines (LAN Chile and Aerolineas Argentinas) resumed regular services to Bariloche, as eruptive activity continued at Cordón Caullé. The decision to fly rests with individual airlines, with standard procedure to avoid flying through any ash plume. From the perspective of pilots, the problem was that they did not have a good system for identifying small, diffuse plumes. A further complication was that the ash forecasting model developed by the National Meteorological Service, and posted on their website for airlines to use, was perceived by airlines as being too ‘experimental’.  The acknowledgement of uncertainties associated with the data and modelling deterred airlines from its use.  As there were no defined safe parameters for ash plume density, there was uncertainty about whether insurance companies would continue to provide cover.  During this period (5 July – 20 October) some airlines did fly into Bariloche, and reported no problems. The airport was closed again on 20 October, until 20 December, for construction of a new runway; this project was brought forward from its scheduled date of March 2012 to take advantage of the existing disruption.
Immediately after the major ash fall on 4/5 June, and for the next two weeks, driving conditions were difficult, primarily because of low visibility in the affected areas. Severe problems related to low visibility were experienced on national route 40, between Piedra del Águila, Villa La Angostura and Bariloche, and national route 231. These two roads are the main connection for Bariloche and Villa la Angostura with the rest of Argentina and with Chile. The international border was closed for transit for several weeks. 
In Jacobacci, it was ‘completely dark’ for three days after the eruption, and municipality staff could only begin the cleanup after a week. There were some reports of engine failure due to ingestion of ash clogging air and oil filters. In areas with ash depths of over 10 cm, two-wheel drive vehicles experienced traction problems during wet conditions. Driving conditions remain treacherous at times due to airborne ash, and official advisories remain in place. Strategies to reduce ash remobilisation in built-up areas include restricting vehicle speeds and dampening ash with water. The region is in general well-equipped to cope with driving in winter conditions, and this probably conferred a degree of resilience.
The formation of thick pumice rafts on Lago Nahuel Huapi disrupted boat transport on the lake for up to 11 days after the eruption. This caused problems for the movement of people and livestock out of farming areas on the western shores of the lake which are not accessible by road (see Section 3.2.1).
3.3.5    Cleanup of urban environments
The removal of ash from business and residential districts is vital for recovery. However, cleanup operations are more complex than just removing ash; it also needs to be disposed of and stabilised to avoid future problems with remobilisation.
Villa la Angostura received up to 15-17 cm of primary ash fall. Sixteen houses suffered roof collapse, and 40 more were braced to prevent roof collapse. A fast and efficient clean-up response was undertaken by the municipality and wider community.  The initial focus was on cleaning the main roads. On the 7 June, 40 km of the main highway (Ruta 231) was closed and cleared with bulldozers. Ash removed by residents with help from volunteer brigades was placed on roadsides then collected by the municipality and taken to provisional ash dumps, located in each neighbourhood. These dump sites were located on the basis of availability, but unfortunately a child was killed whilst playing on a pile of ash that was too close to high tension power lines. Material from the dump sites is now being rapidly transferred to the main dump site which is an old quarry located in Puerto Manzano. At this site, compaction and stabilisation of the ash is planned.  A further focus of cleanup efforts in Villa la Angostura has been the clearing of natural dams higher up the streams that flow through the town. This was done in an attempt to mitigate the lahar risk as it was thought that the dams could cause the build-up of ash followed by catastrophic failure. Army teams were deployed to cut and clear debris.
Bariloche received up to 4.5 cm of ash fall, which equates to approximately 1,500,000 m3 of material across the city. In general, the city was extremely underequipped with the heavy earth moving machinery necessary for cleanup, and had to hire machinery and utilise private vehicles. The first area to be cleared was the four main streets parallel to the lake front, as this part of the CBD is very important for tourism. Cleanup of the city took two months with costs estimated to be some $USD 35 million, excluding business disruption losses. Residents were encouraged to focus on clearing their own properties and were asked to create just one pile of material per city block to facilitate removal by the municipality. Municipality efforts lasted until December 2011.  There were high rates of volunteerism in cleaning the town, particularly in 'high value' areas such as the downtown streets (commonly frequented by tourists), schools and hospitals.   Residents took the initiative to promote cleaning campaigns promoted by social networks and media. Most of the collected material (ash and other urban waste) was disposed in the old municipal quarry located in the southern part of the city, very close to the urbanised area. This dump was quickly filled so new disposal sites were selected. The most important were close to a municipal gas plant where material was accumulated in piles and covered with soil to prevent wind remobilisation; and the municipal dumping site for waste from forestry activities. 
During the first two days of ashfall some material was dumped in the lake both in Villa La Angostura and Bariloche.
In Jacobacci, conditions were so difficult (with darkness and low visibility) immediately after the ash fall that cleanup could only start after a week. The main streets were cleared first, using the available trucks, diggers and bulldozers in the town. Following this, residents were provided with large sacks to fill with ash cleared from their own properties. All the ash collected was placed in natural depressions to the east (downwind) of the town, and has been weighed down with building materials in a short-term attempt to stabilise it. In the longer term, there are plans to vegetate the deposits. The cleanup in this town has been prolonged by wind remobilisation of ash deposits from upwind areas, and was continuing at the time of our visit. The municipality owns two watering trucks that are constantly in use; dampening down ash is a major part of the strategy to attempt to reduce airborne ash levels, along with an emphasis on collecting, storing and stabilising the ash rapidly.
Recommendations
As the expeditious removal of ash from urban environments is essential for recovery, areas exposed to ash hazards should have plans in place beforehand covering the following aspects:
Personnel requirements
Equipment requirements, including arrangements to hire equipment if necessary
Identification of potential dump sites
Strategies for stabilisation of deposits.
3.3.6    Communications
The most reliable form of communication throughout the emergency was radio (VHF and UHF). In Bariloche, radio amateurs were instrumental in relaying information. Cellphone networks experienced problems due to overloading of networks.  There were anecdotal reports of signal attenuation caused by airborne ash and equipment failure caused by deposition of ash onto ground equipment such as antennae. The 12-hour battery life of antennae came close to being exhausted during the power outages.
3.4     Emergency management
The 2011-present eruption of PCC-VC was a complex emergency to manage and offers a range of lessons for emergency management.
At the time of our visit, the focus in Bariloche appeared to be moving on from the immediate necessities of cleanup operations and managing ashfall impacts, to using the lessons from the eruption to improve infrastructure resilience and support the recovery of local industries.  There was still an appreciable sense of emergency in Villa La Angostura and Jacobacci. There were still significant challenges in dealing with continuing remobilisation of ash causing issues for public health, agriculture and economic impacts on the tourism sector.
It was evident that there is a dedicated and capable group of emergency managers and infrastructure managers who have worked hard to respond to and recover from the eruption crisis.  However there appears to be a poorly integrated framework between local, regional and national Civil Protection elements.   The change of staff following local government elections appears to have had a negative effect on developing and maintaining emergency management capability.
3.4.1 Eruption warning and understanding of volcanic consequences
At the onset of the crisis period there had been some media reports of volcanic unrest at Condon Caullé.  Emergency managers had been aware the volcano was in a state of unrest.  Local volcanologists were active in raising public and municipal awareness of the situation at the volcano.  However, these warnings did not reach everyone due to a lack of a formal and effective warning system.  Many individuals said the eruption was a total surprise - although most of these individuals were typically not connected to official emergency management or municipal channels.
There appeared to be poor dissemination of warnings that there had been an eruption and where the volcanic ash may be dispersed.  Many infrastructure managers, officials and members of the public stated that their first warning an eruption had occurred was seeing the volcanic ash plume approaching.
In other cases, where infrastructure and municipal managers had received some warning an eruption was imminent, they did not fully appreciate the consequences of an eruption from Cordón Caullé.  There was general surprise at the thickness and extent of ashfall, and a low pre-existing understanding of what the consequences of ashfall might mean for infrastructure networks, rural and urban communities, despite previous efforts from local scientists.
Some infrastructure managers and municipal officials had a poor understanding of the Volcano Alert Level used by OVDAS-SERNAGEOMIN.  The volcano alert level system solely indicates the status of activity at the volcano, and is not a tool for ashfall warnings.  There was limited integration of this system with emergency management on the ground in Argentina.  This is complicated by being a trans-national issue.
3.4.2 Local-Regional-National Civil Protection (emergency management) framework
There appeared to be communication and coordination problems for emergency managers throughout the affected area.  We felt there was not an adequate framework for sharing information and resources.  A further problem was a lack of understanding of roles and responsibilities during a volcanic crisis.  Most problems issues were at the interface between local-regional-national Civil Protection interfaces.  It appeared there had been substantial lessons learnt.  But these need to be integrated into formal planning arrangements.
We felt that the current Civil Protection framework does not include adequate provision for coordination and utilisation of necessary expert scientific advice (specifically, on the occurrence, distribution, characteristics and impacts of a volcanic ashfall).  Though the authorities made exceptional efforts during the emergency to formalize cooperation channels with local and national scientific and academic institutions, these arrangements should be established during non-crisis periods and active throughout the duration of a volcanic crisis. There are now several important initiatives at national, provincial and municipal levels to develop monitoring, risk reduction and crisis management strategies that involves several scientific technical and academic institutions.
3.4.3 Recommendations
There were excellent examples across the study area of individuals, commercial companies and government organisations developing mitigation and adaptation strategies in response to the ashfall. These will increase resilience to future volcanic eruptions and other disasters in the future.
In general, the emergency management strategies used across the affected area were reasonably effective.  However, it appeared these were more based on individual and organisational adaptive capacity rather than pre-existing planning.  This should be a key lesson to begin comprehensive and integrated volcanic ashfall management planning.  We offer some recommendations for improving future management of volcanic crises in this region.
In the short term, there are critical steps that could improve the effectiveness of the response to future crises. The following strategies may be helpful:
Increased dissemination of volcanic hazard information. 
Improving the content, dissemination, and stakeholder uptake.  Key thematic areas for warnings should include: a) status of activity at volcano, b) when an eruption has occurred, c) likely distribution of ashfall, and d) information to prepare for and manage volcanic ashfall impacts for members of the public, farmers and infrastructure managers.  Material produced by local scientists should be used by Civil Protection to improve the awareness of Northern Patagonia's exposure to volcanic hazards.
Better integration of information from Chilean volcano observatories into Argentine Civil Protection (emergency management) framework. 
o    Coordination of this information by a central agency for Argentina may be advisable, but local and regional emergency managers should be aware of where to access warning information. 
o    Increased education about the OVDAS-SERNAGEOMIN volcanic alert level system.
o    Establish formal communication between OVDAS-SERNAGEOMIN and Civil Protection authorities in Northern Patagonia.
Further development of ash plume modelling capability
o    Better integration/coordination of end-user needs for ash plume models.  If well designed, this would have widespread application.
o    Development of near-real time ash plume modelling capability for warning and preliminary impact assessment.  Fall3D models were available via the internet during the eruption, which forecast the next 2 to 3 days, but were not widely disseminated.
In the longer term, maintaining a solely focused, permanent and professionally-trained emergency management personnel is essential for maintaining and further developing emergency management capacity.  Emergency management professionals must also be given the time and resources to adequately plan and implement recovery actions and strategies.
Embedding the lessons learned from this disaster is essential.  In our experience there is a short window of time during which there will be political, social and economic appetite to commit financial and time resources to improving disaster resilience in affected areas.  This should be maximised. In particular, there is an opportunity to upgrade substandard infrastructure systems, improve agricultural productivity and the local-regional-national Civil Protection emergency management framework.
Improved linkage and collaboration between science agencies and emergency management is required.  In particular there needs to be more effective provision and dissemination of the following information:
Assign a high priority to accurate mapping of the distribution, thicknesses and other characteristics (grainsize, chemical composition) of ashfall deposits.  Such isopach maps are essential for informing impact assessment and recovery planning.
Analysis of ash characteristics using standardised methods for:
o    Respiratory hazards
o    Leachate composition, to assess hazards from leachable elements
o    Analysis needs to be prompt and disseminated widely to all stakeholders
Establish permanent, integrated stakeholder and science groups for volcanic hazard information provision and sharing.  This should be a two way process where expert information is provided and stakeholders identify their needs and information gaps in a collaborative and responsive system.  It may be useful to have separate groups for health, infrastructure and primary industries.
o    For example, infrastructure companies should consider forming an Engineering Lifelines group, which shares information about risks, and works as a group to reduce cascading failure due to interdependency. 
o    There would be considerable value in sharing information on volcanic ash impacts.  For example, many infrastructure companies affected by the 2008 Chaiten eruption could have given information and advice to those affected during the 2011-present Cordón-Caullé eruption.
Finally, we suggest that many of these recommendations would be facilitated by the establishment of a volcano observatory organisation in Argentina. Such an observatory would monitor volcanic hazards which may affect Argentine territory. This would require close cooperation and information sharing with the Chilean OVDAS.  The observatory could also serve as the catalyst and advocate for volcano hazards awareness and emergency preparedness in Argentina.  This model is successfully used in many other countries at risk from volcanic hazards.

6.0    ACKNOWLEDGEMENTS
We would like to thank everyone who took the time to answer our questions and share their experiences, photographs and data. In Bariloche, we are grateful to Claudio Knaup (former Civil Defense emergency expert) and Gabriel Cazalá (from the Municipality), Bariloche International Airport, Departamento Provincial de Aguas, INTA, Cooperativa de Electricidad Ltda., Guillermo Mujica, Carlos Fullana and Horacio Fernández. Also to Analena Santagni, Lic. Silvia Uber and Dra. Andrea Tombari (University of Rio Negro). In Villa la Angostura, Prof. Roberto Cacault, Marcos Arretche, Fernando Anselmi, Alejandro Murcia, Janet Galera, Alejandra Piedecasas, Andrés Sandoval,  Hernán Garabali, Edgardo Carignano and Javier Abraham of EPEN provided us with valuable information including a field trip. From Jacobacci we would like to especially thank Dr. Mario Del Carpio, Ailén Rodriguez (Environmental Coordinator), Helena Herrero (Hsopital Director), Guido Santana and Alberto Mondillo (Epidemiology), Juan Escobar, Jose Mellado and Idelma Sarlor (Coop de Agua).  From Pilcaniyeu Hospital, Dr. Marcelo Graemiger and Maria Laura Sassone. From Comallo; Marta Ester Llanos, Miguel Angel Cunilof, Andrés Nahuel Colaso y Bibiana Favre (Primary School Supervisor). From Zona IV (Neuquén) we thank Dra. Fernanda Hadad, Dr. Daniel Ricardi, Dr. Ricardo Powel and Dr. Alejandro Ojeda (From the Ministry of Health, Subsecretaria de Salud de Neuquén).  The Inibioma team wants to thank Dr. E Gómez from IADO-CONICET for providing grain-size analyses.  Thank you to Dr Roberto Volante.  Thank you to the many farmers for allowing interviews. And finally, particular thanks to David Dewar for outstanding translation support.
The New Zealand team was funded by the New Zealand Ministry of Science and Innovation through the Natural Hazard Research Platform subcontract: C05X0804. Additional support was provided by the New Zealand Earthquake Commission and Auckland Council through the DEVORA project.  The INIBIOMA team was funded by CONICET (Special fund for the emergency and research funding PIP 2011 0311 GI) and by the Scientific Cooperation Agreement signed between Universidad Nacional del Comahue and the province of Neuquén.


APPENDIX ONE: HEALTH HAZARDS OF VOLCANIC ASH FROM THE PUYEHUE-CORDÓN CAULLÉ ERUPTION
Dr Peter J Baxter
University of Cambridge, UK
April, 2012
pjb21@medschl.cam.ac.uk
1. Background
1.1.  The current status of the ash problem: ash deposits from the PCC-VC eruption that fell during the start of activity in 3-5 June 2011 are gradually being incorporated in the soil in areas where there is regular rainfall, but in the impacted dry areas in the Patagonia steppes the deposits are persistent and very high levels of exposure to fine particulate matter (PM) in the air occur to the whole population on a regular basis during days with strong winds.  The concern is that these conditions, which are badly affecting settlements along the Linea Sur, and other ash-impacted parts of the Steppe, may possibly take years to   resolve.  PCC-VC continues to have small ash eruptions and large deposits are still visible blanketing the mountain range above Bariloche and Villa La Angostura. 
1.2. This note summarises the continuing concerns over the risk to health from the continuing exposure of populations to ash and which arose out of our meetings with local health and environmental professionals.  Meetings were held in Bariloche, Villa La Angostura, San Martin de Los Andes, Junin de Los Andes, Pilcaniyeu, Comallo and Jacobacci. 
1.3. Heavy ash falls like those experienced in Argentina in June 2011 inevitably cause anxiety about possible short and long term health effects of inhaling fine ash and may be an important underlying factor for people leaving the affected areas to go to live and work elsewhere.   Economic factors like the fall in tourism, the closure of local airports, and the losses to farm livestock are some of the most important threats to livelihoods after the eruption, but health concerns may also be an important consideration why businesses close and people move away.   It is therefore an essential aspect of community resilience and sustainability for the authorities to provide timely and reliable advice on what is known about the likely health hazards after an eruption and to have plans for limiting exposure of the population to ash, especially in outdoor workers and children, and establish guidance for health professionals.   
1.4.   Although local laboratories had done valuable work in characterising the ash and its water-soluble leachates, our impression was that information had not been officially co-ordinated or disseminated by public health officials.  In addition, local physicians we met were confused about whether ash should be taken seriously as a respiratory health hazard and on the need for further research, especially in children.  There was general agreement amongst the health professionals we met in Bariloche, San Martin and Junin that acute respiratory problems had not significantly increased or worsened amongst their patients who attended clinics or the hospitals following the June ash falls.   Some epidemiological studies on patient attendances had been done, but were not possible to interpret because of the fall in tourism and the numbers of local residents compared to previous years’ figures.
The aspects that need further consideration in any assessment on the risks of the ash to respiratory health, and for the planning of mitigation measures, will be outlined next.
2.  Ash characterisation.
2.1.  Grain size distribution.  Claire Horwell (Durham University, UK) received five ash samples from this eruption: two were coarse material and three had 4.5 – 9.0 weight% in the less than 4 µm diameter (respirable) range, and about double that amount in the less than 10 µm diameter (thoracic) range.  We would expect the ash resuspended from deposits of this fine ash to contribute substantially to PM10 and PM2.5 measurements in the ambient air, with large numbers of particles being capable of being inhaled into the air sacs (alveoli) deep inside the lungs.
2.2.  Crystalline silica.  Claire Horwell found no detectable crystalline silica in her samples.  I collected two samples of very fine ash piled up against scrub bushes by the wind outside Jacobacci and Comallo, and had them analysed by Jennifer Le Blond using the XRD-PLS instrument at the Natural History Museum, London. No   detectable crystalline silica was found.  This result is very reassuring, in that it goes some way to excluding a major potential cause of fibrotic lung disease (silicosis and mixed dust fibrosis), and a causal factor for chronic obstructive pulmonary disease (COPD) and lung cancer.  However, the main ash fall activity in June continued over two-three days and the area covered by ash remains very extensive: more analyses of ash samples taken from a wider area, together with the samples collected at different stages of the eruption by volcanologist Gustavo Villarosa, warrant additional analyses to provide final confirmation of the absence of crystalline silica.
3.  Exposure measurements.
3.1.  Since the June 2011 eruption, exposure to ash has been mainly from persistent deposits on the ground that become resuspended by winds or human activity.  Rain makes a crucial difference to resuspension and exposure to ash.  The cities of Bariloche (pop. 120,000), Villa La Angostura (pop. 11,000), San Martin (pop. 24,000) and Junin (pop. 10,000) are favoured by regular rain or snow fall in the winter, and overall exposure to ash is likely to be low in comparison with settlements in the arid areas of the Patagonia Steppe (e.g., Comallo – pop. 2000 and Jacobacci  - pop. 5000).
3.2.  A notable observation is that the ash deposits are so persistent on the mountains after winter snow; in dry, windy weather they are likely to add to the particulate air pollution in the nearby cities.  Less surprising, but potentially much more hazardous to health, are the persistent and extensive ash deposits along the Linea Sur settlements we visited which are blown into ash storms by the strong winds that last the whole afternoon on four days a week on average throughout the year.
3.3.  Daily measurements of ambient air PM10 (and PM2.5) are needed to understand  the background levels of fine ash people are exposed to in different cities and settlements, both hourly and as 24 hour average means.   Apart from a monitor at the Bariloche airport, used to check ash conditions for aircraft, and a few results from an air monitor used by SEGEMAR, the amount of available exposure data is very limited.
3.4. Individual exposures to ash in dry conditions are going to be higher than background readings, especially in outdoor workers, like farmers.  At times, some individuals doing dusty tasks or walking in ash deposits, for example, could have very high transient peak exposures, but their average exposure over 24 hours (including when levels are low inside homes, especially at night) should be much lower.
3.5.  Exposure measurements inside and outside buildings will be useful in assessing the effectiveness of ash clean-up measures and the sealing of houses, schools and other buildings against ash entry, especially under the dry, windy conditions in the steppes.
3.6. At the time of our visit none of these exposure measurements were being undertaken for lack of available equipment. 
4.  Clinical effects of exposure to fine particles (PM2.5) in ash-impacted areas.
4.1.  Acute irritant effects of particles on the upper respiratory tract, eyes and possibly exposed skin would be expected when ash levels are highly elevated in the air.  Individuals with pre-existing chronic lung problems, including those with asthma,   would be expected to be affected most.  In urban areas like Bariloche, Villa La Angostura, San Martin and Junin the number of people badly affected will be mainly limited by the wider availability of weather resistant housing that reduces the ingress of ash and the regular rainfall suppressing resuspension, as well as ready access to medical care for many people.
4.2.   By contrast, in the Patagonian Steppe, the lower than normal rainfall in the last five years, the prevailing dry conditions and high winds mean that extremely high levels of PM are attained and the housing is generally of poorer quality and not built to resist the entry of ash; ready access to medical care is also more limited in rural communities.  Eye problems, for example, were very common and in some people required admission to hospital.  Children spent four months living in severe ash conditions after the June ash falls until schools were allowed to re-open, when they moved to the towns and stayed in dormitories where their exposure to ash would probably have been much less than in their home areas because the school buildings can be more adequately sealed to protect against ash entry than their homes. 
4.3.   In the sheep farming setting of the steppe, where ash storms lasting several hours occur as often as four days a week, finding out about health conditions like asthma in children, or worsening of health in adults with pre-existing health problems, is more difficult than in large cities and requires formal epidemiological surveys, as reliance cannot be placed on the numbers of attendances at clinics and hospitals.  The impact on lives can be substantial, however, and families may decide to leave the area if one of the family members is badly affected by the ash. The prevalence of asthmatic symptoms   in children aged 13-14 yr in Argentina (Neuquen) is typically about 10% in one published ISAAC survey.  Patients with advanced heart and lung conditions should consider talking to their doctors about moving, if the option is available, whilst the severe exposure conditions last. 
4.4.  Additional adverse factors that need to be considered in some of the rural communities include the indoor burning of fossil fuels and tobacco smoking (both of which may expose households to further particulate pollution). Poverty and malnutrition is associated with tuberculosis and also childhood pneumonia (the commonest cause of illness and death in children under five years of age in low to medium economies), both of which may increase in incidence and severity due to the high ash exposure.
4.5.  The evidence for long term health effects developing as a result of the current high levels of exposure to fine volcanic ash is less secure than the acute effects just described, as the necessary epidemiological surveys of large populations after volcanic eruptions have not been undertaken anywhere in the world.  It is plausible, however, that lengthy and elevated exposures to volcanic PM2.5 could contribute to, for example, impairment of lung growth in children and the development of chronic obstructive pulmonary disease (COPD) in adults (WHO, 2006).  COPD is the fourth most common cause of death worldwide (due to become the third by 2030) and the lifetime risk of contracting it is one in four; the key casual factor is active smoking.  There is a very high prevalence of smoking in Argentina.  COPD can also be caused, or contributed to, by environmental insults (including mineral dusts).
5.  Toxicological aspects.
5.1. The review on the respiratory health effects of volcanic ash by Horwell and Baxter (2006) summarises the scientific literature up to and including 2005.   Since that time, research on combustion-derived PM has been applied to developing policies for reducing air pollution in cities and evaluating its public health impact (COMEAP, 2009; 2010).  How relevant is this research to volcanic ash?
5.2.  The annual average concentration of PM2.5 pollution in urban air of the UK and USA is in the order of 7-30 µg/m3.  Within this range and without specifying which components of PM2.5 may be the most important, experts have come to some conclusions about the relationship, or risk coefficients, between exposure to PM2.5 and mortality risk using the results from epidemiological studies in large populations.  (A risk coefficient summarises the percentage change in health outcome per unit change in air pollution concentration).   A short term, incremental rise of 10 µg/m3 of PM10 over background levels is sufficient to increase deaths from cardio-respiratory diseases by about one per cent (0.6%) over hours and days afterwards.  Long term exposure to urban air with a raised background concentration of 10 µg/m3 of PM2.5 may increase all-cause mortality (mainly from cardio-respiratory diseases and lung cancer) by 6%.  The mechanisms are still far from being understood, but toxicologists favour an explanation based on the inflammatory properties of the ultra-fine carbonaceous particles, together with their associated organics and transition metals formed from the chemical conversion of combustion gases emitted by traffic (WHO, 2006). 
5.3.  By contrast, there are no epidemiological studies that have followed up mortality in populations after volcanic eruptions and it is impossible to provide similar risk estimates for PM2.5 in volcanic ash.   There is a general consensus, however, that mineral dusts are less toxic than combustion-derived particles in urban air.   For example, we do not observe the large increases in hospital admissions (or mortality) that used to occur in the infamous smog episodes due to coal burning in the past, even though the concentrations of fine particles of ash in the air after eruptions may be even more elevated than the combustion-derived particles in the deadly smogs. 
5.6.  Nevertheless, we should regard the PM2.5 fraction of airborne ash as the most important for health risk assessment, as particles below 2.5 µm in aero-dynamic diameter deposit effectively in the alveoli, especially in high risk sub-groups in the population, such as the sick and elderly, or children.  Ultrafine ash particles may also be present.  Coarser particles should not be ignored, however, as they may act on the larger airways and trigger asthma and bronchitis symptoms, for example; hence, the PM10 (the thoracic fraction) is also a useful measure and it incorporates PM2.5.  The risk coefficients discussed above for PM2.5 and PM10 in urban air are useful as they define a relatively small level of risk in a large exposed population and exposure to ash at such low concentrations is unlikely to carry as high a risk.  Extrapolation of these or other risk coefficients to the much higher concentrations usually seen in eruptions is not, however, straightforward in the current state of knowledge.  
6.  Public health research
6.1.  In view of these uncertainties, further research should be undertaken on the health effects of the P-CC ash if the population is to be adequately informed about the risk of past and present exposure.  This is important because of the continuing high exposures in the Patagonia steppe which may continue for months or years to come, as the populations in the rural settlements have already been, and continue to be, very highly exposed on a daily basis to PM2.5.  The following studies are suggested as priorities:
Routine data collection of mortality statistics and hospital/clinic attendances should be reviewed and monitored in the ash fall areas.  TB and childhood pneumonia need to be included in the monitoring of poor rural areas.
A study of asthma and lung function in school children should be considered as a priority.  School children in Linea Sur settlements are a high exposure group and warrant long term follow up; low to medium exposure groups for   comparison could be chosen from the towns in the tourist areas where they receive much more rain.
A study of respiratory symptoms and lung function in a group of adults, e.g., farmers, in the Linea Sur and a comparison group in rural areas where there is more rainfall.  A study cohort could be established for long-term follow-up.
Various objective variables for study could include inflammatory markers for systemic inflammation, as a sign of high exposure to fine ash and evidence of potential risk of cardio-vascular disease impact. Psychological stress (anxiety and depression) was reported to us as common in some rural areas: this also warrants further scientific investigation.
Eye symptoms have been common and in some cases severe enough to require hospital admission: further study is needed as ocular effects of this severity due to volcanic ash are very unusual.
6.2.  These studies involving samples of the population need to be expertly conducted with sufficient power and quality, and to continue for long enough, to be able to reassure people that adult and child health are not being irreversibly affected by continuing ash exposure, in particular in the steppes. 
7.  Measuring exposure to PM.
7.1.  Measuring PM in the ambient air is an essential part of risk assessment and would be a key parameter in the epidemiological studies listed above.   PM10 and PM2.5 are invisible and cannot be estimated without proper instrumentation.
Particle air monitoring stations (as installed in major EU and USA cities) should be established in the study areas to record hourly and daily means of background PM10 and PM2.5.  These require a technician’s support and are costly.  The results can be automatically transmitted to a central office. 
Hand-held monitoring instruments like the DustTrak should be used to check PM levels in the outside air in towns and small settlements, and indoors in schools and houses to measure the effectiveness of sealing buildings against the ingress of ash.  They can be set to record for long periods and downloaded.
Exposures to PM in outdoor workers, or direct measures of exposure in other individuals, can be measured using instruments such as the Sidepak, which is a small device that works like the DustTrak, but fits on to a belt.
An exposure survey using the DustTrak and Sidepak instruments should be undertaken in the worst affected towns, such as Pilcaniyeu, PichiLeufú, Corralito, Ing. Clemente Onelli, Comallo and Jacobacci, to establish the current mean background levels of PM10 and PM2.5 in the ambient air under a range of representative weather conditions, and to define the much higher mean personal exposures of children and representative groups of adults.  See the methodology used on Montserrat by Alison Searl and others (Occup. Environ. Med. 2002; 59; 523-531).
7.2.  Some limited air sampling by collecting PM10 on a filter with a pump has been undertaken by SEGEMAR.  Significantly, we saw a mean result for one whole week (day and night) of air sampling at a location outside the hospital in Jacobacci in August, when conditions were still bad; the result was 600 µg/m3 – a very high figure. 
8.  Reducing exposure to ash.
8.1.  The most important determining factor in exposure to volcanic ash is rainfall.  It suppresses resuspension of ash, clears suspended ash from the air, and washes ash away down slopes.  It encourages the growth of grass through deposits of ash in flat fields or lawns around houses, and the incorporation of ash in soils in cultivated fields. 
8.2.  The low rainfall in the steppe has the opposite effect, with fine ash blown into piles against shrubs and waiting to be resuspended by strong winds.  Ash storms, when visibility is reduced to less than one kilometre, occur on four days a week on average, last for several hours in the afternoons, and are worse in spring and summer. 
8.3.  We were in Jacobacci during one of these ash storms and PM10 was measured at  919 µg/m3 (range: 277 - 6530) in the street over 15 min., and 625 µg/m3 (range: 374 - 1300) inside a restaurant.  At about the same time inside the hospital, where windows were sealed up against the ash, the airborne concentrations on spot checks were around 130 µg/m3 and 30 µg/m3 in the (empty) operating theatre.    This showed the effectiveness of measures taken to prevent the infiltration of ash in an ash storm.  The conditions got worse later when we went out of town; the winds accelerated and started to pick up soil as well as ash, and the visibility dropped to a few metres.
8.4.  Essential measures to reduce the average exposure to ash are to seal buildings against wind-borne ash and keep streets, and the surrounds of schools, homes and other buildings free of ash.  This requires constant effort in the Linea Sur settlements, but it should be available to everyone to be able to massively reduce their exposure by going inside during ash storms.  Dwellings need to be upgraded where necessary to keep the wind-blown ash out.  In Comallo, ash was lying in main streets and on playing fields; by contrast, in Jacobacci the streets were regularly cleared of ash.
8.5.  Effective, light-weight masks (N95, or FP2/FP3 standard) are available but were not widely used.  For very dusty work or other high exposure periods they should always be worn.  They should also be available for people with chronic respiratory problems.  Suitable masks for children are not generally supplied by manufacturers.
9. Animal studies.
9.1.  Sheep and cattle are highly exposed to ash in the fields.  Blindness was commonly reported and many animals died from lack of food and water in the early eruption crisis.  Rumens became blocked by boluses formed from a mixture of ingested ash and wool from the sheep’s own coat. Respiratory problems in livestock were not reported, but attempts should be made to study the lung tissue from dead sheep for histological evidence of early chronic inflammatory changes due to ash.  Negative findings in sheep with a history of lengthy and high exposure to ash would be reassuring evidence for humans living in the ash storm areas. 
10.  Impacts of Andean eruptions.
10.1.  A major ash eruption impacting on Argentina occurred about every 10 years in the 20th century.  The eruption of Hudson in 1991 was the third largest in the world in the 20th century and had a similar impact to the present P-CC one in the arid areas of Patagonia.  The lessons of the Hudson eruption need to be studied.  It would also be important to know if health professionals were finding evidence of fibrosis in chest X-rays taken in patients for other reasons and which might be attributable to exposures to ash following this eruption. It provides another important opportunity for studying respiratory health impacts in settlements arising from exposures to ash that have continued for many years after the initial eruption.  
10.2.  Chaiten erupted in 2008 and the initial ash fall across Argentina did not contain cristobalite.  The volcano remained active and the ash from subsequent smaller eruptions was found to contain up to 20% cristobalite.  Public health officials need to be prepared for a major future eruption of Chaiten in order to ensure that the ash is analysed for cristobalite without delay.  This concentration of cristobalite in ash deposits would present a serious potential hazard of silicosis in the impacted settlements of the Patagonian steppes and population-wide measures to reduce exposure would have to be urgently introduced.
General references
Electronic versions are also available on-line
Horwell CJ and Baxter PJ.  The respiratory health hazards of volcanic ash: a review for volcanic risk mitigation.  Bull. Volcanol. 2006;69: 1-24.
WHO Europe. Air Quality Guidelines.  Global Update 2005.  Copenhagen: WHO, 2006.
Committee on the Medical Effects of Air Pollution (COMEAP).  Long term exposure to air pollution: effects on mortality.  A report by COMEAP.  London: Health Protection Agency, 2009
Committee on the Medical Effects of Air Pollution (COMEAP). The mortality effects of long term exposure to particulate air pollution.  A report by COMEAP.  London: Health Protection Agency, 2010.



APPENDIX TWO: DATA FROM VILLA LA ANGOSTURA
Table 1 Tap water (agua de red) collected on 11 July 2011 from different neighbourhoods of Villa la Angostura. Data kindly provided by laboratory of Bromatologia Municipal, courtesy of Alejandro Murcia.

Neighbourhood     Brief description of water supply system    Summary of water quality (microbiological)     Residual chlorineφ      
Las Bandurrias    Intake in Lago Nahuel Huapi, storage in three tanks of 30 m3 each, disinfection with sodium hypochlorite. The system is considered well maintained.     Fit for human consumption    None      
Villa Correntoso    As above    Not fit for human consumption (microbial indicators exceed allowable levels)    None      
Club Cumelen
Tank #1    Private system, has its own intake in Lago Nahuel Huapi. No filtration, just chlorination.     Fit for human consumption    <0.1 mg/L      
Club Cumelen
Tank #2    As above    Fit for human consumption    <0.1 mg/L      
Puerto Manzano    There is a stream intake on a tributary of the Rio Bonito, and a pump from the main Rio Bonito. The system has coarse screening of organic debris and a sand trap, but no sand filters. There are major problems with organic debris from the stream intake and it would be preferable to use only the pumped water. There is a chlorine dosing pump but the quantity of suspended solids and organic matter in the system prevents an adequate chlorine residual in the distribution system.      Not fit for human consumption (turbidity identified visually in water)    None      
Las Piedritas
‘House 4’    There is a stream intake in the Arroyo Colorado, followed by a sand trap and then slow sand filters which are in a state of disrepair and poorly maintained.     Not fit for human consumption (turbidity identified visually in water)    <0.1 mg/L       
Las Piedritas
‘School 186’    As above     Not fit for human consumption (turbidity identified visually in water), and also a break identified in the chlorine dosing system)     <0.1 mg/L       
‘In front of the fire station’ (El Once neighbourhood)     Lomas del Correntoso treatment system has intakes on Lago Nahuel Huapi and Lago Correntoso. An additional intake on the Arroyo Piedritas is currently not used because of problems with suspended solids in flood flows. For the Lago Correntoso intake, there is a new filtration system with four rapid sand filters. However, this water is then mixed with unfiltered water from the other intakes. The stored water is then chlorinated.     Fit for human consumption    0.5 mg/L    
φ to prevent re-infection in the distribution system, residual chlorine should be in the range 0.2-0.5 mg/L.
APPENDIX THREE: Previous investigations on 2008 Chaiten eruption and 1991 Hudson eruption
Lessons from Hudson – coping with wind remobilisation of ash
Our field team conducted a study (in January 2008) of long-term recovery of rural communities and farming operations in the area affected by ashfall from the 1991 eruption of Hudson volcano, located approximately 600 km south of Puyehue-Cordón Caullé. The Hudson eruption was the third-largest of the 20th century, and deposited ashfall over approximately 150,000 km2 of land in Santa Cruz province. For several years after the eruption, recovery of farming areas was severely inhibited by problems caused by wind-remobilised ash (‘ash storms’, see Wilson et al. 2011a). Seedlings were sheared off by windblown ash or buried by ash deposits, inhibiting any pasture growth. Livestock health continued to suffer, with gastrointestinal blockages, tooth abrasion and eye irritation continuing. Many of the marginal farms were abandoned and there were extensive migrations out of the area throughout the 1990s.
Lessons from this event are applicable to the 2011 Cordón Caullé eruption, especially for the steppe region where remobilised ash is already causing problems for agricultural recovery.
Farmers who acted quickly to stabilise ash fall deposits by cultivating them into the soil and replanting had a clear advantage over those who did not. Cultivation using a chisel plough (to maintain a textured surface) and hay mulch (to increase soil organic content) were found to be most effective.
The use of shelter belts of willow or poplar was valuable in providing protection for crops from windblown ash. These are not an immediate solution as they require time to become established.
Diversification of production into greenhouse cultivation was advantageous as it provided farmers with an additional income stream and are relatively resilient to conditions with high levels of windblown ash.
The development of district irrigation schemes has allowed greater resilience to climatic variability, and greatly increased production volumes.
References
Wilson, T.M., Cole, J.W., Stewart, C., Cronin, S.J., Johnston, D.M., 2011a. Ash Storm: Impacts of wind-remobilised volcanic ash on rural communities and agriculture following the 1991 Hudson eruption, southern Patagonia, Chile. Bulletin of Volcanology 73 (3): 223-239
Wilson, T.M., Cronin S.J., Stewart, C., Cole, J.W., Johnston, D.M. 2011b. Agricultural recovery following the 1991 eruption of Vulcan Hudson. Natural Hazards 57 (2): 185-212
Wilson, T.M., Cole, J.W., Johnston, D.M. Cronin S.J., Stewart, C.,  Dantas, A., 2012. Short- and long-term evacuation of people and livestock during a volcanic crisis: lessons from the 1991 eruption of Volcán Hudson, Chile. Journal of Applied Volcanology 1:2 doi:10.1186/2191-5040-1-2