Ash storms: impacts of wind-remobilised volcanic ash on rural communities and agriculture following the 1991 Hudson eruption, southern Patagonia, Chile
- First Online:
- Cite this article as:
- Wilson, T.M., Cole, J.W., Stewart, C. et al. Bull Volcanol (2011) 73: 223. doi:10.1007/s00445-010-0396-1
- 604 Views
Tephra fall from the August 1991 eruption of Volcán Hudson affected some 100,000 km2 of Patagonia and was almost immediately reworked by strong winds, creating billowing clouds of remobilised ash, or ‘ash storms’. The immediate impacts on agriculture and rural communities were severe, but were then greatly exacerbated by continuing ash storms. This paper describes the findings of a 3-week study tour of the diverse environments of southern Patagonia affected by ash storms, with an emphasis on determining the impacts of repeated ash storms on agriculture and local practices that were developed in an attempt to mitigate these impacts. Ash storms produce similar effects to initial tephra eruptions, prolonged for considerable periods. These have included the burial of farmland under dune deposits, abrasion of vegetation and contamination of feed supplies with fine ash. These impacts can then cause problems for grazing animals such as starvation, severe tooth abrasion, gastrointestinal problems, corneal abrasion and blindness, and exhaustion if sheep fleeces become laden with ash. In addition, ash storms have led to exacerbated soil erosion, human health impacts, increased cleanup requirements, sedimentation in irrigation canals, and disruption of aviation and land transport. Ash deposits were naturally stabilised most rapidly in areas with high rainfall (>1,500 mm/year) through compaction and enhanced vegetation growth. Stabilisation was slowest in windy, semi-arid regions. Destruction of vegetation and suppression of regrowth by heavy tephra fall (>100 mm) hindered the stabilisation of deposits for years, and reduced the surface friction which increased wind erosivity. Stabilisation of tephra deposits was improved by intensive tillage, use of windbreaks and where there was dense and taller vegetative cover. Long-term drought and the impracticality of mixing ash deposits with soil by tillage on large farms was a barrier to stabilising deposits and, in turn, agricultural recovery. The continuing ash storms motivated the partial evacuation of small rural towns such as Chile Chico (Chile) and Los Antiguos (Argentina) in September–December 1991, after the primary tephra fall in August 1991. Greatly increased municipal cleanup efforts had to be sustained beyond the initial tephra fall to cope with the ongoing impacts of ash storms. Throughout the 1990s, ash storms contributed to continued population migration out of the affected area, leaving hundreds of farms abandoned on the Argentine steppe. The major lesson from our study is the importance of stabilisation of ash deposits as soon as possible after the initial eruption, particularly in windy, arid climates. Suggested mitigation measures include deep cultivation of the ash into the soil and erecting windbreaks.
KeywordsVolcanic ashAsh stormAgricultureRecoveryHazard
Following Bitschene et al. (1993), we adopt the term ‘ash storm’ to describe events of intense remobilisation of fresh volcanic deposits by strong winds following an eruption. This implies that tephra or other volcanic particles are the predominant component of suspended clasts, sourced from primary tephras or lahar deposits that had buried existing soil and vegetation.
Field work was carried out in January–February 2008, in southern Patagonia, in areas that received tephra fall from the 12–15 August 1991 eruption of Volcán Hudson. Carrying out a field visit some 16 years after the eruption allowed for a comparative study of mitigation measures adopted in different locations. Because of limited time, emphasis was placed on the effects of the eruption and subsequent ash storms on agriculture; evaluating other impacts such as on tourism, human health and the economy were beyond the scope of this study.
Globally, there is very limited information on how affected communities recover following large volcanic eruptions, perhaps with the exception of communities surrounding Paricutin, Mexico (e.g. Rees 1979; Ort et al. 2008). Typically, there is substantial scientific, political and media interest when impacts are at their worst, but there is little in the way of ongoing monitoring of recovery processes despite this information being crucial for long-term volcanic risk management. One way to address this is by regular and ongoing monitoring of relevant environmental, economic and social indicators following an eruption. But in instances where no regular monitoring has occurred, a reconnaissance trip to affected areas many years after the eruption still affords the opportunity to collect the lessons learnt and establish which environmental, social or political factors were most important in shaping community or industry (in this case agricultural) recovery. This approach has been used here.
Much of the information contained within this paper is of a qualitative nature, derived from semi-structured interviews. This approach allowed detailed and in-depth data collection on complex issues. But there are several limitations of this approach which must be acknowledged, including repeatability, potential for interviewer bias, relevance of qualitative data collected, difficulties in generalising one-on-one interviews and validity (e.g. if the respondent is biased).
Pastoral livestock farming for meat and wool production at varying intensities is the dominant agricultural practice throughout the area. Farms in the Upper Ibáñez valley, southern Andes Mountains (20–50 km from Volcán Hudson), vary in size between several tens to hundreds of hectares, with low–medium stocking rates of sheep and beef cattle. In the Lower Ibáñez valley including Cerro Castillo, 50–90 km from Volcán Hudson, higher intensity sheep and beef cattle farms of a similar size occur. Warm, well-irrigated valleys, draining into Lago General Carrera/Buenos Aries, host small pockets of intensive horticultural production and pastoral farming of up to tens of hectares on the outskirts of the rural service towns in Puerto Ibáñez, Chile Chico, Los Antiguos and Perito Moreno. In the vast steppe region from Perito Moreno to the Argentine Atlantic coast, large farms (estancias) of over 20,000 ha are common.
Wind remobilisation of volcanic ash
Erosivity of sediment is determined by particle size, density, morphology, soil roughness, soil moisture content and degree of binding/compaction (Sivakumar 2005). Unconsolidated, fine and often low-density ash, such as pumice deposited after an explosive volcanic eruption, can be easily remobilised by strong winds when the wind velocity at the soil surface exceeds the static threshold of the least stable soil particle (Fowler and Lopushinsky 1986). When wind force reaches the threshold value, some particles begin to vibrate and increasing wind speed causes some of them to be ejected into the atmosphere. When these particles fall back on the surface, they may cause ejection of further particles, beginning a chain reaction (Sivakumar 2005). The dynamic threshold is reached once the wind velocity is sufficient for sediment movement to be sustained (Bagnold 1941). Once ejected, particles move in suspension, by saltation or by creep, depending on their size, shape and density. Typically, the higher the wind velocity and the lower the particle density, the more likely transport will be by suspension (Sivakumar 2005). In wind tunnel testing, Fowler and Lopushinsky (1986) found that loose dry ash from the 1980 Mt St Helens eruption, with a smooth surface, was suspended at wind speeds of 11–12 km/h, whereas particles with irregular surfaces were suspended at lower wind speeds (6–9 km/h). Material <3.5 μm was displaced first. Wet ash that had been air-dried developed a firm surface, completely stabilising it to wind, even up to 69 km/h, but once this surface was disturbed, the dynamic threshold dropped to 22.5 km/h.
There have been few direct studies on extensive ash storms following a volcanic eruption (see Electronic supplementary material for details), although it is well recognised that large explosive pyroclastic eruptions generate dramatic and large-scale sedimentary and geomorphology processes (Manville and Wilson 2004; White et al. 1997). Post-eruptive erosion and re-sedimentation of pyroclastic deposits are maximised by large erupted volumes, abundant unconsolidated ash-sized material, destruction of the vegetation cover and inhibition of vegetation regrowth (Manville and Wilson 2004). Regions that experience long dry periods, regular strong seasonal winds, insufficient vegetation cover to protect the soil and inappropriate management practices which disturb the soil surface suffer the most serious wind erosion problems (Sivakumar 2005). It has been estimated from arid and semi-arid areas globally (<250–500 mm/year rainfall; Cofinas and Creighton 2001) that 24% of cultivated land and 41% of pasture land are affected by moderate to severe wind erosion (Rozanov 1990).
▪ percolating water elutriating fines from the tephra layer into the underlying soil macropores;
▪ frost heave and soil fauna activity incorporating tephra into soils, especially for thin deposits (<5–12 mm);
▪ dewatering and rainfall developing an erosion-resistant silty surface cap that may also be strengthened through chemical precipitation;
▪ regeneration and maintaining vegetation cover to reduce wind velocity at the ground surface, intercept rain drops, bind tephra particles together with root systems and humus;
▪ incorporation by cultivation or tillage;
▪ stockpiling and covering with an erosion-resistant cap; and
▪ avoiding disturbance of consolidated deposits.
Few accounts exist that describe methods that have been attempted to stabilise pyroclastic deposits over broad areas. The eruptions of Volcán Hudson in 1991 and Chaitén in 2008 and the resulting tephra deposits across Chile and Argentina point to the need for effective ways to stabilise the deposits to minimise their remobilisation by strong wind. This is the motivation of this paper.
1991 eruption of Volcán Hudson, Chile, and reported ash remobilisation by wind
Volcán Hudson (45°54′ S, 72°58′ W) erupted in two separate, partially sub-glacial phreatoplinian phases on 8–9 August and 12–15 August 1991. Located in southern Chile, Volcán Hudson is part of the Chilean Southern Volcanic Zone (33–46°S; Kratzmann et al. 2009). At least 12 Holocene explosive eruptions are sourced from Hudson, the most significant of which were those at 6,700 years BP, 3,600 years BP, and ad 1991 (Naranjo and Stern 1998). The 1991 eruption is notable for having an abrupt change from initial basalt/basaltic–andesite magma to later much larger volumes of trachyandesite and rhyodacite magmas (Naranjo et al. 1993; Kratzmann et al. 2009; Bitschene 1995). Wind dispersed ash from the first phase northwards on 8–9 August 1991 and from the second (silicic) phase ESE on 12–15 August 1991 across a narrow, elongated sector of Patagonia, covering an area (on land) of more than 100,000 km2 (Fig. 1 from Scasso et al. 1994). The eruption produced a 4.3-km3 bulk volume (2.7-km3 dense rock equivalent) of tephra fall deposits, making it one of the largest explosive eruptions of the last century (Kratzmann et al. 2009). The strongly elongated footprint of the deposit resulted from strong winds blowing to the southeast during 12–15 August. The prevailing strong westerly winds (known as the ‘roaring forties’) were supplemented by a jet stream blowing to the southeast in the upper troposphere–lower stratosphere at estimated speeds of up to 240 km/h (Scasso et al. 1994).
The 1991 tephra deposit was described by Scasso et al. (1994) as comprising up to 12 layers with marked bimodality in the grain size distribution reflecting alternating fine-ash and coarse-ash (pumice) layers in areas up to 270 km from the volcano. They also reported secondary thickening at approximately 500 km ESE of the vent and noted an overall a large proportion of fine-grained material (<0.0625 mm) and variations in bulk density with distance.
Wind remobilisation began immediately following the tephra fall on 16 August. These deposits showed parallel or cross-bedding with better sorting within each individual layer than primary fall deposits, each layer typically exhibiting unimodal distribution with a well-developed fine to very fine ash mode (Banks and Ivan 1991; Scasso et al. 1994; Bitschene et al. 1993). Strong spring winds occurred for up to a week and reworked huge quantities of ash and some terrigenous material (Scasso et al. 1994; Inbar et al. 1995). In general, ash deposits were transported by westerly winds, with much being deposited in the Atlantic Ocean (Fig. 1). Greatest erosion occurred from exposed areas, such as ridgelines, river valleys (like the Ibáñez valley) and the sparsely vegetated steppe region. The fine ash component was reportedly remobilised once winds exceeded 35 km/h, forming huge dust clouds or ‘ash storms’ (BGVN 1991a, b, c; Scasso et al. 1994; Bitschene et al. 1993).
Ash storms affected most of the deposition zone from the 12–15 August eruption and some areas on the periphery of the zone, particularly locations to the east including the large Chilean fluvioglacial valleys such as the Rio Ibáñez, and in the Argentine steppe.
A major ash remobilisation event was detected by the GOES satellite in early September 1991 where 55- to 65-km/h surface winds generated ash clouds extending from near Volcán Hudson to the Atlantic Ocean (BGVN 1992). The densest part of the remobilised clouds were around 250 km SE of the volcano in the steppe (BGVN 1992). Another was recorded on 27 November 1991 when NOAA-12 images recorded extensive grey plumes extending from the vicinity of Hudson more than 1,000 km across southern Argentina and the western Atlantic Ocean (Fig. 1; BGVN 1991d). The following day, ash could be tracked over the Atlantic to beyond 40°W. Additional smaller plumes were seen on 3 December satellite images (BGVN 1991d). These were initially interpreted as ash plumes from new eruptive activity, but the lack of observed eruption and no accompanying seismicity indicated wind remobilisation of the 1991 ash (BGVN 1991d).
Strong winds led to the remobilisation and accumulation of ash at Mar del Plata, on Argentina’s Atlantic coast >1,500 km NE of Hudson, on 2–3 December 1991 (Fig. 1; BGVN 1991d), and at Comodoro Rivadavia (420 km E of Hudson), depositing 2 mm of ash on 21 March 1992. Windblown ash deposition also occurred south to Río Gallegos (700 km SSE of Hudson) in early 1992 (Fig. 1; BGVN 1992). A regional survey around Los Antiguos in 1992 revealed that the region was affected by ash storms every 7 days on average, with valleys particularly affected and ash deposited in dunes and sheets (Fig. 1; Bitschene et al. 1993). Some stabilisation of ash occurred in winter due to winter precipitation and accompanying vegetation growth.
When interviewed in 2008, people across the entire study area (southern Patagonia) reported that wind-blown ash was “extremely bad” for around 4–6 months after the eruption, during the period of seasonal high winds in October to February (Fig. 3). During intense wind events, ash storms reduced visibility to <1 m in Puerto Ibáñez and Chile Chico and forced people to constantly seek shelter indoors. By 1996, ash storms had decreased in severity and frequency, but areas with tephra fall deposits >50 mm were occasionally still affected, such as Chile Chico, Los Antiguos and Perito Moreno. In areas with smooth topography and sparse vegetation, such as at Puerto Ibáñez and in the Argentine coastal steppe, ash storms continue as a hazard to the present day, but with reduced frequency and impacts. Along the Argentine coast, effects on farming, transport routes and the few isolated settlements were extreme where ash was blown from the interior across the continent to the coast. The semi-arid conditions and sparse vegetation growth meant that the fine ash deposits did not readily stabilise.
Soil grain size changes—1991 to 2008
Wind-remobilised ash particles were reported to be typically <5 mm immediately post-eruption (BGVN 1991c). To test this, grain size distributions of the topsoil/ash deposit were taken from undisturbed and uncultivated sites (avoiding eroded areas) in 2008 in order to compare with the grain size of the 1991 primary tephra. Samples were taken at Upper Ibáñez valley, Cerro Castillo, Puerto Ibáñez, Chile Chico and a site in the Argentine steppe (Estancia sample), 270 km from Volcán Hudson, for comparison with data from Banks and Ivan (1991) and Scasso et al. (1994) (Fig. 1). The original tephra thickness at these locations was 50–1,500 mm, respectively (Fig. 4a; Naranjo et al. 1993; Scasso et al. 1994). Efforts were made to representatively sample all of the primary tephra deposit, where visible. The 2008 deposit samples also show decreasing grain size with increasing distance from the volcano (Fig. 4b). The 2008 samples are consistently coarser than the 1991 samples from the same localities and lack the fine tail of the original ash distribution. This may indicate that the fine ash component has been eroded by wind. This correlates to reports immediately post-eruption that remobilised ash particles in 1991 were typically <5 mm (BGVN 1991c). Note that Banks and Ivan (1991) did not sieve below 63 μm; however, they and Scasso et al. (1994) sampled tephra fall at Chile Chico, providing a check between both datasets. A good correlation is shown between the two datasets up to 4ϕ (63 μm), and the data have been extrapolated to approximate the more detailed analysis of Scasso et al. (1994). On this basis, extrapolations are plotted for the rest of the Banks and Ivan data for Ibáñez valley and Cerro Castillo in Fig. 4a.
The 2008 Estancia sample is much coarser-grained than the 1991 deposit, indicating that there has been extensive erosion or other reworking of the deposit, in particular the fine fraction >5ϕ (<32 μm). The very fine ash component (i.e. <8 μm or >7 phi) is completely missing in the 2008 Chile Chico and Estancia samples. This is supported by Inbar et al. (1995) who reported that the mean size of airborne particles in Santa Cruz province, Argentina, collected from static and dynamic samplers during September 1992, was between 20 and 30 μm, and most of the samples were <60 μm.
Ash dune formation
Wandering dunes were reported by Bitschene et al. (1993) at Los Antiguos and continuing eastward in 1991 and 1992.In the region between Los Antiguos and Perito Moreno, dunes were reported to be between 0.5 and 1.5 m high in valley areas, 5–10 years following the eruption (BGVN 1992). In Perito Moreno, the original 50-mm fall deposit was overlain by up to 450 mm of remobilised deposits in 1991 and 1992 (Scasso et al. 1994). Bitschene et al. (1993) estimated that 70% of ash deposits east of the Perito Moreno–Bajo Caracoles line were eroded by September 1992 and deposited in the South Atlantic Ocean.
Mapping, sampling and qualitative interview data from farmers and emergency management officials indicated that windblown ash dune formation occurred primarily >70 km from the volcano. In the Upper Ibáñez valley, few wind-blown dunes formed, presumably due to the coarse grain size of the fall deposits and suspension (rather than saltation) of the fine component of ash deposits by strong winds between 1991 and 1996. Small dunes, 10–20 cm in height, buried pastures and affected farming operations in the Lower Ibáñez valley, close to Cerro Castillo approximately 70 km from Volcán Hudson throughout the 1990s and were still observable in 2008 on exposed fields. In Puerto Ibáñez, dune formation was significant, with the 50- to 80-mm fall deposit overlain by up to 2 m of ash dunes by 2008. In February 2008, ash was still retained in the lee of vegetation in the tephra plume region and in some cases persisted as a reworked layer above the soil exhibiting low-angle cross-bed textures. Dunes were between 0.2 and 0.5 m high in sparsely vegetated steppe valleys.
Ash dunes were reported at Tres Cerros in 1991 and continue to affect farming operations in 2008, with ash blowing in from the west. In Puerto San Julian, there was initially insufficient fine ash to form dunes greater than several centimetres high, but subsequent ash storms over the succeeding 17 years have formed dunes 10–20 cm in height (Antonio Tomasso, 2008, personal communication).
Composition of reworked material
Lithology data for samples of ash dunes from southwest of Puerto Ibáñez based on point counts of 500 particles
Grain size (μm)
Pumice: white vesicular (%)
Pumice: banded (%)
Pumice: grey (%)
Crystal: plagioclase (%)
Crystal: clinopyroxene (%)
Crystal: K-feldspar (pink cream, %)
Crystal: orthopyroxene (%)
Crystal: amphibole or hornblende (%)
Lithic: hydrothermal (%)
Lithic: grey steel (%)
Lithic: vesicular (%)
Impacts on agriculture
In areas that received >100 mm of tephra deposits such as at Puerto Ibáñez, Chile Chico and Los Antiguos, vegetation was not totally buried, but the subsequent ash storms and resultant deposition made vegetation survival and growth even more tenuous or impossible for many years. The Argentine coastal steppe region received less than a few centimetres of tephra fall, but was severely affected by ash storms, causing significant disruption to farming activities and damage in the years subsequent to the eruption. The wind storms created frequent tephra fall-like conditions, and fast-moving saltating ash particles abraded exposed vegetation and affected livestock. The extended impacts have contributed to a severe depression of the agricultural economy in the region (Shaquib Hamer, Don Julio Cerda Cordero, Don Hector Sandin, 2008, personal communications) which, in the early 1990s, was also dealing with the effects of hyperinflation, rising unemployment and reduced import protectionism (World Bank 1993; INDEC 2009). Some interviewees made mention of the tough economic conditions in the early 1990s, though no mention was made of the 2001 economic collapse in Argentina.
The 1991 eruption of Volcán Hudson caused severe and widespread impacts on livestock (Bitschene et al. 1993; Rubin et al. 1994; Bitschene 1995; Inbar et al. 1995). There is no accurate estimate for total livestock mortality due to tephra fall and subsequent ash storms, but it may be as high as one million sheep and several thousands of cattle.
Detailed descriptions of the impacts on livestock were collected from farmers across the study region and are summarised here. Nearly all farmers reported that impacts were extreme for approximately 3 months following the eruption. Most interviewees recalled that ash storms had a significant ongoing impact to livestock through repeated coverage and destruction of pasture for months to years after the tephra fall. They also suffered gastrointestinal problems from ingesting ash-covered feed. In addition, there was severe loading of fleeces with ash, causing exhaustion which commonly led to mortality. Corneal abrasion leading to blindness occurred in extreme cases. In the weeks and months following the eruption, farmers in Chile Chico were forced to sell their livestock due to uncertainty about pasture recovery and ongoing ash storms impacting livestock and destroying feed availability (Wilson et al. 2009). In the steppe region, it was impossible to evacuate livestock or supply supplementary feed, in part because the poor visibility during ash storms made it extremely difficult to find livestock (Wilson et al. 2009).
The large farms (ranches) in the steppe were economically marginal before the eruption, with herds in poor condition and land typically overstocked (Rubin et al. 1994; Shaquib Hamer, 2008, personal communication). These ranches suffered extreme livestock losses following the tephra falls and ash storms. Fine windblown ash covered the sparse available feed of native tussock grasses and shrubs and also contaminated surface water supplies (Peri and Bloomberg 2002). A 120,000-ha ranch close to Perito Moreno lost 50% of its 30,000 sheep mainly due to ash ingestion and starvation when animals were faced with ash-covered feed and contaminated water supplies. However, losses were greatly reduced close to the homestead where livestock could easily be herded, given supplementary feed and taken to water. In outer areas, they were very difficult to rescue. Dramatic reductions in stock numbers were required on the farm following the tephra fall and ensuing windblown ash storms. Sheep were sold, rented out and some gifted (with the lambs sent back). A 20,000-ha ranch at Tres Cerros reported that in the first month following the tephra fall, half of its 12,000 sheep were lost. Following the intense period of ash storms in late 1991, 11,000 (approximately 90%) of the animals had died. Those that remained suffered severe abrasion to their front teeth from grazing ash-covered feed.
Ash storms severely affected crops and pastures from suspended or saltating ash particles. Plant tissue was abraded or damaged by the fast-moving, sharp particles, and the constantly shifting ash deposits reduced photosynthetic activity and led to dehydration and burial (Ambrust 1984; Black and Mack 1984; Fryrear 1990). Impacts to vegetation depended on the strength and timing of ash storms with respect to the stage of plant development. Mature plants were most resilient, and stems were more resilient than leaves. New seedlings growing through ash deposits were most susceptible, slowing vegetation recovery. This further reduced feed availability especially for close-grazing animals. Cows could reach higher shrubs and trees which were less vulnerable to damage from saltating particles. The most severely affected pastures received <5 mm of the initial tephra fall and were in exposed areas of the Lower Ibáñez valley and the vast Argentine steppe (between Los Antiguos and the Atlantic coast where winds were over 120 km/h and ash storms cut grass at soil level and “abraded poplar trees like sand paper”; Don Hector Sandin, 2008, personal communication). In Tres Cerros, it was only around the year 2000 that vegetation began to reestablish (Don Hugo Ciselli, 2008, personal communication).
In the Lower Ibáñez valley, Puerto Ibáñez and ranches around Perito Moreno and Tres Cerros, vegetation was covered by ∼20 mm to 3 m of ash sheets or dunes. Fields in the lower Ibáñez were most heavily impacted and ash covered all feed. In the Ibáñez valley and Perito Moreno, farmers reported that most of the ash had blown away or stabilised after 3 to 5 years (Shaquib Hamer, 2008, personal communication).
Seedlings and young recovering grasses were vulnerable to repeated burial by small dunes 20–100 mm high. This was a significant problem between 1991 and 1994 in the Lower Ibáñez valley, Chile Chico, Perito Moreno and the Argentine steppe. The burial of seedlings brings their meristems in contact with warm soil surfaces and can accentuate nutrient and water deficiencies, which hampers seedling recovery (Michels et al. 1993, 1995). Buried plants are known to suffer delayed development compared to unburied plants and reduced number of tillers and panicle length (Michels et al. 1993). Totally buried plants were starved of sunlight if ash cover was not removed by further dune movement or human mitigation strategies.
In Perito Moreno, remobilised ash contaminated alfalfa, which reduced harvest of this important supplementary feed. Mowers and balers could not operate on the ash-contaminated crop due to abrasion and clogging of moving parts by fine ash. Cutter bars and mower blades were mostly ruined. Farmers would typically only cut and bale on a calm day; otherwise, the feed would simply be covered in ash.
There was no direct damage to crops during the tephra fall because it was winter (August) and seasonal crops were not growing. However, the ash storms generated persistent horticultural problems in Puerto Ibáñez, Chile Chico and Los Antiguos during the following growing seasons. Losses and failures occurred in crops such as potatoes, tomatoes, garlic and alfalfa due to ash cover and ash storm abrasion, even within fields protected by poplar and willow tree windbreaks. Farmers without windbreak protection could not grow horticultural crops for around 6 years after the tephra fall, and even in 2008, ash storms still made cultivation difficult.
In the immediate area, farming operations have been severely disrupted. Fields buried in ash dunes were previously highly fertile with high-yielding tomato, potato and corn crops, and pastures for intensive pastoral production (Ulyses Pededa, 2008, personal communication). Seeds cannot be sown in the coarse lahar deposits and windblown sediment constantly abrades or buries new growth. The stocking rate has been reduced from around 100 cows (per 40 ha) in 1991 to six in 2008, and all non-greenhouse horticultural production has ceased. The cattle suffer continual eye and respiratory problems. Only 30 ha of wind-protected land was farmed in 2008 compared to the original 150 ha before the eruption.
Of wider importance to the urban Puerto Ibáñez community is that the affected area is upwind of the township and sources ash storms that affect the town. Houses on the southwestern side of Puerto Ibáñez have increasing interior contamination issues since 2006 comparable to the intense windstorm period of 1991–1995. Abrasion damage occurs on greenhouses and sub-aerial horticultural crops. Farmland to the east is slowly being buried and lost to the advancing dunes.
The cherry harvest in Los Antiguos (the dominant crop of the area) was exceptionally poor for several years following the 1991 tephra fall, impacting the Los Antiguos economy and the many people who relied on the cherry harvest (Inbar et al. 1995). The airborne ash harmed pollinating insects, particularly bees, reducing fertility amongst the trees (Johansen et al. 1981; Brown and bin Hussain 1981; Marske et al. 2007), thus limiting pollination and fruit set (Peri and Bloomberg 2002). Bird populations were also reduced by the strong spring winds and ashy conditions.
Remobilised ash had an abrasive effect on apple, peach, pear and apricot crops. When attempts were made to clean ash-covered fruit during harvest, it damaged the fruit. Tomatoes covered in fine ash coatings following an ash storm in early 1992 were considered contaminated and refused by buyers. Depressed yields continued for several years, but most interviewees agreed they had recovered by 1993–1994 as ash storm intensity and frequency subsided, and by 1996–1997, the harvests were back to normal. This was an extremely difficult time for orchardists, needing to survive for essentially two seasons without income, relying on savings and government aid. “Every now and then life was pretty bad in those 2 years” (Ananias Jonnutz – Productov, 2008, personal communication).
Agricultural mitigation measures
In the area to the southwest of Puerto Ibáñez township, many additional semi-permanent windbreaks have been hastily erected since 2006 to mitigate the effects of windblown ash (Figs. 6 and 7) to protect homesteads and greenhouses. They are nested closely beside an established willow/poplar windbreak to provide increased protection from damaging saltating particles. Each windbreak uses two rows of willows with large logs and aluminium sheets to block the wind so smaller willows can grow behind them. The first exposed row is often strongly abraded. This acts mainly as a barrier to allow the second row of willows to grow successfully. There is enough space between each row so that a wheelbarrow can move between them for clearing fine particles. So far, approximately 400 m of this windbreak design has been installed. Apart from the physical barrier orientated parallel to the permanent windbreak, the system also uses groynes orientated perpendicularly to increase surface roughness by providing a baffling effect, which reduces wind speed and causes particle deposition.
This farmer-designed and built system provides some protection, but its low height and density has been inadequate to protect nearby houses from contamination and greenhouses and exposed horticultural crops from damage. Windbreaks are unable to be installed fast enough (i.e. over 18- to 24-month timescales) to provide adequate protection for the recent problems. Although some assistance was given to small farmers for windbreaks in this area to control remobilised ash, it has been insufficient to effectively control the ash storms. Local farmers recommended that windbreaks should have been implemented more rapidly and with greater government assistance given the protection they afford residential houses in Puerto Ibáñez.
Revegetation of ash deposits
Farmers strove to revegetate ash-covered soils throughout the area impacted by ash storms to reestablish feed for livestock and stabilise the ash and reduce ash storms. Vegetation recovery stabilised ash deposits by reducing wind velocity at the ground surface, intercepting rain drops and binding ash particles together with root systems and decomposing organic matter (Folsom 1986; He et al. 1993; Sivakumar 2005; Peri and Bloomberg 2002). Surface crop residues also reduce water loss, increase soil surface roughness (reduce wind velocity) and shield soil from saltating particles (Michels et al. 1995; Ambrust 1984; Sivakumar 2005). Decomposition of crop residues also adds valuable organic matter to volcanic deposits (Sivakumar 2005).
Where ash was <100 mm, farmers in Puerto Ibáñez, Chile Chico and Los Antiguos ploughed the ash into the buried soil with rotary hoes. In the short term, this cultivation destabilised the ash deposits by breaking up the resistant rain-compacted crust, but incorporation of the ash into the existing soil consolidated and stabilised ash deposits and restored soil fertility. This cultivation had to be carried out with considerable care as it potentially pulverised buried soil, making it also more susceptible to erosion in dry or windy conditions. It also removed any remaining vegetation cover binding the soil and providing shelter from the wind. A chisel plough was most effective because it permitted the cultivation of vegetated surfaces and maintained a rough, well-textured surface (Sivakumar 2005). Where ash deposits were overthickened to >300 mm, ploughs were unable to penetrate the ash and farmers waited for the wind to blow some of the ash away before ploughing. In some instances, hay mulch was used to increase organic matter in the ash to promote plant growth and assist binding the ash together to more rapidly form topsoil. White clover was the most successful pasture variety to reestablish in this type of material. In some areas, forestry was also established.
Rural town cleanup and ash disposal
House and business roofs were swept clear of tephra fall to mitigate against potential roof collapse and as part of individual cleanup operations. In many instances, the ash needed to be hosed off roofs as it was too heavy to sweep off with brooms. Most residential and business roofs consisted of sheets of corrugated iron on a wooden frame, often unpainted. Older houses (>30 years) commonly used wooden tiles, nailed to a wooden frame. Refer to Wilson et al. (2009) for further details on the impact of tephra fall to roofs.
Fine ash penetrated houses and buildings during each ash storm event. This is despite houses being generally well-constructed, clad and heated (usually by a woodstove) due to the common frosts and snowfalls in winter. Fine ash would enter small gaps in the walls, doors and windows, or penetrate through the roof, contaminating exposed surfaces (food, clothing and furniture). In the hours and days following each ash storm, there was so much ash inside houses that electric lights were rendered useless. Doors and windows were taped up, but ash still penetrated into houses. One interviewee in Chile Chico reported wearing out three vacuum cleaners cleaning ash from her house over an 8-month period. Other interviewees boiled water regularly to remove suspended ash particles inside houses.
Household contamination in the Upper Ibáñez river valley largely ceased by 1995, apparently due to the stabilisation and erosion of fine ash in the area. However, in Puerto Ibáñez, household contamination was still an issue in 2008, albeit with reduced frequency and intensity. Residents in Chile Chico reported that they needed to constantly clean their houses, roofs and roof interiors of ash for 8–12 months following the eruption. Even today, contamination remains a public health concern in Tres Cerros and Puerto San Julian, with up to 70 mm of fine ash still trapped in attics and roof cavities (Antonio Tomasso, 2008, personal communication).
Municipal officials in Puerto Ibáñez (population ∼3,000), Chile Chico (∼4,000), Los Antiguos (∼2,500) and Perito Moreno (∼3,500) acknowledged that a fast and efficient cleanup of ash was essential to stop remobilisation in the urban environment. Cleanup operations were complicated by the persistent ash storms remobilising ash, shortages in earth moving equipment (initially) and evacuations out of towns which reduced the available workforce. Ash in Puerto Ibáñez, Chile Chico, Los Antiguos and Perito Moreno was typically piled up on roads and footpaths by individual households and businesses following cleaning and later loaded into trucks by hand or mechanical diggers and buried at ash dumps away from town to prevent further remobilisation. The piles of ash on roads and footpaths were easily remobilised by wind, causing additional household contamination, reduced visibility, surface water contamination and public health impacts. Dump trucks, graders and diggers were sent by the Chilean and Argentine regional governments to assist with cleanup operations. These operations took 30–60 days, with Chile Chico municipality reporting that around five million cubic metre of ash was removed from the town. Chile Chico and Los Antigos covered dumped ash with soil to prevent its remobilisation. However, at Perito Moreno, without capping, the ash dump suffered remobilisation problems. Ash dumps at Chile Chico and Los Antiguos were cited in the town garbage dumps, both situated in valleys to the south of both towns.
In Puerto San Julian, despite thin tephra fall (5 mm; Scasso et al. 1994), continual deposition of windblown ash from inland sources led municipal authorities to decide against promoting a community cleanup operation, which may have contributed to the perceived current health issue.
Other infrastructure impacts
On 2–3 December 1991, flights from Camet airport at Mar del Plata were cancelled because of reduced visibility and mounds of ash blocking runways following an ash storm (BGVN 1991c; Guffanti et al. 2009). Airlines took evasive action to direct flight paths away from the zone affected by ash storms for several years following the eruption (BGVN 1991c).
Ash-covered roads following the eruption and subsequent ash storms led to frequent road closures due to extremely poor visibility and damage to vehicles. Visibility along the main Argentine coastal highway between Comodoro Rivadavia and Puerto San Julian was affected for up to 6 months following the eruption, which disrupted food deliveries in Tres Cerros and Puerto San Julian. Even with lights on motorists could not see, resulting in a number of accidents. Near Perito Moreno, roads were initially not visible beneath the primary tephra fall deposit, which required clearing with bulldozers. However, they were again frequently disrupted for hours to days due to deposition of remobilised ash following ash storms for 3–4 months following the eruption. The closure of roads stopped food, water and milk from arriving.
The fine volcanic ash was commonly very damaging when ingested by vehicle engines in all areas for several years after the tephra fall during ash storms. It blocked air filters, abraded seals and was highly abrasive to moving parts of the engine, such as pistons. Other parts of the vehicle (brakes, wheel bearings, etc.) were also potentially affected. As a consequence, machinery usage and vehicle travel was reduced or avoided during ash storms. Engines required frequent reconditioning. In Tres Cerros, a turbocharger was sometimes installed, which was successful in flushing air out and reducing ash ingestion to the engine (Don Hugo Ciselli, 2008, personal communication). This appeared to be a local initiative and was not reported to be undertaken elsewhere. Oil and air filters were rapidly blocked with ash. In Tres Cerros, oil and air filters, and the oil itself were changed every 15 days following the initial tephra fall and each subsequent ash storm event for up to 4 months. Ash mixed with engine oil to form a paste, reducing effective lubrication and not allowing oil to be easily drained from the engine. In extreme cases, this led to engines losing power or seizing up. Windows and paintwork were also damaged by ash blasting (Don Hugo Ciselli, 2008, personal communication). In Los Antiguos and Puerto San Julian, wet sacks were put over air filters to trap ash. This apparently worked quite successfully if regularly changed or cleaned. Mechanic Don Hugo Ciselli (2008, personal communication) reported that much of his time in the months following the eruption and initial 4-month period of intense ash storms was spent cleaning motors of ingested fine ash with compressed air.
Electricity was turned off for a month in Rio Gallegos (Fig. 1, inset) following the eruption due to the severe remobilisation of ash associated with the strong winds creating fears of ash building up on insulators and creating flashover hazards. Tres Cerros and Puerto San Julian continued to operate their own diesel generators, which worked successfully throughout the ashy conditions. Satellite communication in Los Antiguos reportedly failed during the ash storms, but HF radio and telephone systems remained in operation throughout (Antonio Tomasso, 2008, personal communication).
Domestic water in Tres Cerros was sourced from 100- to 120-m deep wells and was hence safe, although disrupted when electric- and windmill-powered pumps were badly impacted by ash storms in a similar way to vehicles. Windmill bearings were particularly vulnerable to seizing.
Effects on human health
A full investigation of human health effects was beyond the scope of this study, but some observations were made and these are recorded below.
Following concern about fluoride toxicity from the tephra fall, studies by Banks and Ivan (1991) and Rubin et al. (1994) evaluated ash leachates, natural waters, and grass and concluded that fluorine was not present at toxic levels. These studies did allay some health concerns. Fears remained high, however, for health impacts from fine suspended ash during ash storms. Common complaints included irritated respiratory tracts (especially in children and asthmatics), irritation of eyes and skin (with effects similar to sunburn), along with bronchial problems and allergies developing after the eruption..
Most people stayed indoors during ash storms. Work attendances dropped due to additional time house cleaning, illness or inability to travel during ash storms. Civil Defence authorities in most communities distributed masks (which were not always available) after the eruption and recommended the use of goggles in the ashy conditions. In areas such as Puerto Ibáñez and Tres Cerros, masks were still required in 2008 due to ash storms. In some cases, the masks provided were inappropriate for dealing with the fine ash particles. These masks, designed to protect against chemical weapons, were donated to Los Antiguos from Israel, but they took weeks to arrive and were rapidly blocked by fine ash. Many of the masks sold were not capable of filtering fine ash, so multiple masks had to be worn at once. In some cases when authorities advised masks would be required, entrepreneurs brought all available masks and sold them on at inflated prices (Luis Fernando Sandoval Figueroa, 2008, personal communication). Bitschene (1995) noted that Chile Chico hospital did not record any elevated rates of sickness or mortality in the months after the eruption. The lack of any formal health assessment following the eruption was of concern to a number of locals interviewed.
Evacuation and relocation
In all areas, fears of health impacts and household contamination by ash were cited by many people interviewed as one of the reasons for families temporarily evacuating or permanently relocating from areas exposed to ash storms.
In the Upper Ibáñez valley area, farmers were typically forced to evacuate due to the severe conditions created by the tephra fall, and attempts to return during the first 3–4 years were thwarted by intense remobilisation of ash deposits in ash storms.
The role of ash storms in the displacement of farmers in the Lower Ibáñez valley was more complex. In areas receiving 300–500 mm of tephra fall, farmers usually evacuated. They were forced to remain in temporary housing for up to 3–4 years, periodically returning to check on farms and homesteads, whilst ash storms suppressed vegetation recovery and contaminated homesteads. Ash storms eventually eroded enough of the ash deposits to allow vegetation recovery by 1994-1995 and subsequent repatriation.
No large-scale human evacuation initially occurred from the Puerto Ibáñez area, although some families evacuated to Coihaique (Fig. 1, inset) immediately following the eruption. Over 600 (mainly women and children) were evacuated from Chile Chico, Los Antiguos and Perito Moreno several weeks after the tephra fall due to ash storms creating strong public health concerns (Bitschene 1995). Men stayed behind to clean up and manage farms, business or municipal services. There were also fears of further heavy tephra fall affecting the area. Elderly people and young children were evacuated from Chile Chico to Coihaique at the end of August for around 4 months to avoid the ash storms. In Perito Moreno, most farmers remained on their farms during and following the ash, but children were evacuated as soon as possible due to health hazard fears. Visibility was very poor from high concentrations of suspended ash carried by strong winds, so evacuation traffic from the towns would wait for periods of good visibility before leaving.
Widespread evacuation of remote, extensive farming steppe areas surrounding Chile Chico, Los Antiguos, Tres Cerros and Puerto San Julian occurred due to the tephra fall and intense initial ash storms. Farmers initially left their land to live in community shelters or with relatives because of the difficult conditions created by the repeated ash storms. This created additional welfare demands in the rural towns affected by tephra fall. For example, in Puerto San Julian, the municipality provided bedding and food to up to 300 evacuated farmers (Alberto James Alder, 2008, personal communication). The evacuations ultimately became a permanent relocation or abandonment of farms in many cases because ash storms continued to impact vegetation and livestock. Most went back to check on their farms, but ultimately concluded sheep farming was no longer viable.
As many as 80% of extensive pastoral farmers in the steppe were estimated to have abandoned their farms due to poor productivity and health fears from windblown ash by former Perito Moreno agricultural officer Shaquib Hamer (2008, personal communication) and other local residents and public officials. The former Mayor of Puerto San Julian, Alberto James Alder, reported that approximately 7,000 people were forced from their farms in the local area. In 2008, only few farmers remain in the steppe area affected by ash deposition. Many of these farms were struggling prior to the eruption with dry conditions and falling wool and meat prices. The eruption and ongoing ash storms appeared to greatly accelerate their decline and force many from the land.
Summary and conclusions
Tephra fall from the August 1991 eruption of Volcán Hudson affected some 100,000 km2 of Patagonia and was almost immediately reworked by strong winds, creating billowing clouds of remobilised ash, or ‘ash storms’. The immediate impacts on agriculture and rural communities were severe, but were then greatly exacerbated by continuing ash storms. Ash storms produce similar effects to initial tephra eruptions, prolonged for considerable periods.
Fine ash deposits were the most susceptible to erosion, especially in the distal portions of the tephra fall plume. In areas with higher rainfall in the Andean region, such as the Upper Ibáñez valley, ash deposits stabilised most rapidly by rainfall compaction and enhanced vegetation growth. Semi-arid regions around Perito Moreno and in the Argentine steppe took the longest period to recover. Destruction of vegetation and suppression of vegetation growth occurred when tephra fall was >100 mm. This also reduced surface friction, allowing increased wind erosivity. Stabilisation of ash deposits was aided by tillage into buried soils. Thin ash deposits were most rapidly stabilised, but thicker deposits were highly erodible for years. Fluvial erosion of tephra deposits occurred contemporaneously with wind erosion. Lahar deposition at Puerto Ibáñez stockpiled sediment, which was later a significant source for wind erosion.
Ash storms following the 1991 eruption of Volcán Hudson created significant and ongoing impacts on rural communities in southern Patagonia, prolonging their recovery from the disaster. This highlights the fact that impacts from significant tephra falls do not end with the eruption; several years of destruction may follow. Ash storms repeatedly buried farmland under ash dunes, abraded vegetation, contaminated feed supplies and generated animal health impacts (tooth abrasion, gastrointestinal problems, blindness, impeded movement and exhaustion). It also eroded topsoils and filled irrigation canals and stock water supplies with ash. These impacts forced the adjustment and abandonment of some farming practices. The widespread and heavy tephra fall, combined with ongoing ash storms, forced a major reduction in livestock and abandonment of farms across the region. This was a major problem in the downwind distal parts of the tephra plume despite having received relatively thin tephra falls. Wind erosion provided some benefits in places by reducing the thickness of ash deposits to allow faster and/or more effective recovery of pasture and crops.
Windbreaks were effective at protecting horticultural and pastoral production, and greenhouses provided useful additions to farm incomes. Mechanical cultivation served to stabilise ash deposits in intensive agricultural areas. However, large-scale sediment stabilisation programmes were not implemented due to the large areas that would have required treatment, economic constraints and sparse population.
Some farms failed through continual low production or destruction of their livestock, soil or pasture. However, others have remained in the impacted area, with some thriving. Farmers who cultivated soils and/or installed greenhouses appear to have been more successful.
Rural towns and lifelines were badly impacted, with major town cleanup activities required to stabilise ash deposits. Perceived public health problems, mainly related to possible respiratory impacts from airborne ash fall, and uncertainty about them were an important driver of short-term evacuation and permanent relocation decisions. Although, the ongoing impact on agricultural productivity was probably the main factor.
Remobilisation of ash deposits by strong winds in semi-arid southern Patagonia has had severe and ongoing impacts which should be anticipated by civil defence and emergency management authorities in areas of similar environmental conditions that are exposed to high volcanic hazard. Disaster risk reduction should not simply be focused on direct tephra fall impacts but include the effects of ash storms.
In areas subject to strong winds, early stabilisation of ash deposits (by deep cultivation, erection of windbreaks, etc.) may be required in order to avoid ongoing problems with ash storms. Assessment of potential problem areas is essential post-eruption, similar to lahar zone identification, to avoid instances like Puerto Ibáñez where communities suffer extreme impacts and adopt mitigation strategies too late to be effective. Relief and recovery policies must be sufficiently flexible to allow farmers to adopt mitigation measures, such as greenhouses, windbreaks and increased cultivation in areas where continued agricultural production is still possible. However, this must be conducted within the bounds of sustainable agriculture as some areas may simply never recover due to external stresses, such as the Argentine steppe area.
In a wider sense, this study illustrates the valuable lessons about the recovery process that can be learned from long-term studies. The limitations of the study highlight the need for longitudinal post-disaster studies during the recovery period as there is much that can be learnt by timely observations before time-dependent data perish. Future work should also be undertaken in the study area to investigate the impacts on public health, sociological perspectives, local economies and businesses (including tourism) to provide a multidisciplinary understanding of how the crisis has evolved to the present day.
The authors would like to acknowledge funding assistance from the New Zealand Ministry of Agriculture and Forestry Grant POR/SUS 7802/40 (Wilson and Cole), the New Zealand Earthquake Commission, Foundation of Research Science and Technology Grants C05X0804 (Johnston and Cole) and MAUX0401 (Cronin), and the University of Canterbury Mason Trust. David Dewar provided outstanding translation and field support. Sincere thanks to interviewees, in particular Don Julio Cerda Cordero, Veterinarian, SAG; Ulyses Pededa, farmer and artist, Puerto Ibáñez; Ananias Jonnutz – Productov, Orchardist, Los Antiguos; Don Hector Sandin, ranch owner, Los Antiguos/Perito Moreno; Luis Fernando Sandoval Figueroa, Civil Defence Officer, Los Antiguos; Shaquib Hamer, former Agricultural Officer for El Instituto Nacional de Tecnología Agropecuaria, Argentina (retired), Perito Moreno; Don Hugo Ciselli, mechanic and farmer at Tres Cerros; Antonio Tomasso, cable television installation specialist and farmer, Puerto San Julian; Alberto James Alder, former Mayor, Puerto San Julian.