Landslides as climate indicators in Argentinean Central Andes (32º S) Stella M. Moreiras 1 1 Consejo
Nacional de Investigaciones Científicas y Tecnológicas (CONICET). Instituto Argentino de Nivología, Glaciología y Ciencias Ambientales (IANIGLA) – CCT Mendoza. Av. Dr Ruiz Leal s/n. Parque (CP 5500). Mendoza. ARGENTINA. Tel: +54-0615244234/4200. Fax: +54-0261-5244201.
[email protected]
Landslide occurrence in the Arid Central Andes of Argentina (~32°L) is strongly related to topography and climate conditions. In middle-low mountains and valleys, landslides are linked to summer rainfall, while in highest Andean areas, instability conditions are mainly caused by water saturation due to snow thawing in spring sessions. Besides, climate phenomena forced precipitation and slope instability in these mountain ranges. Warm phases of ENSO episodes are linked to more landslides activity in Frontal Cordillera, and a delayed signal of this phenomenon was found for Main Cordillera in the following warm session. Besides, no influence of ENSO is found for eastern Precordillera responding to Atlantic Anticyclone behavior.
Keywords: Mendoza river valley, international road to Chile, triggering mechanisms, slope instability.
Introduction Natural hazards and their socio-economic impact are strongly influenced by precipitation regime and climate phenomena in the Arid Central Andes. Summer rainstorms are a major forcing of landslide occurrence in the central Andes of Argentina; still many landslides have been also triggered by historical earthquakes (Ms>3.9). In the past century, at least 500 rainfall-induced events have affected the international road Nº 7 connecting Mendoza (Argentina) with Santiago (Chile) that runs along the valleys of Las Cuevas - Mendoza rivers (32ºS) (Moreiras, 2005; Moreiras, 2006). On summer 2013, intense summer storms triggered a large number of debris flows in the Andes of central Chile and Argentina. The main flows occurred along the Mendoza River valley (Argentina), where at least 50 debris flows caused serious disruption to the international road infrastructure and high impact to the population, mainly due to potable water supply cut offs to major cities (Moreiras and Sepúlveda, 2013) (Fig. 1). Over long-term periods, warm phase of ENSO has been linked to major landslides activity and above-normal river streamflows (Moreiras, 2005; Sepúlveda et al., 2006; Araneo and Compagnucci, 2008). However, the influence of the Pacific Anticyclone seems to be limited to high mountain areas of Main and Frontal Cordillera, while hillslope instability in eastern Precordillera is forced by humid periods of the Atlantic Anticyclone (Moreiras, 2006). Analysis of synoptic situations for summer and winter precipitation events have recently been documented (Viale and Nuñez, 2011; Viale and Garreaud, 2013); however, the extreme precipitation processes leading to debris-flow is still lacking in the study area, and need investigations.
This paper updated previous published and new unpublished data about rain-induced landslides, climate phenomena influence and variability of this landslide links along the central Andes.
1. Study area A high topography reaching up 7 km in altitude (eg. Aconcagua mount 6,958 m asl) and an arid climate characterized this portion of Arid Central Andes. However both parameters vary gradually longitudinal wise from west to east (Fig. 1). Topography of this mountain landscape decreases toward the eastern piedmont (700 m a.sl.) forcing precipitation behavior. While solid winter precipitation predominates in highest mountain areas, summer rainfall does in lower areas and valleys. Likewise, an average annual precipitation of 500 mm is measured in highest areas of the Andes diminishing until 200 mm in the Andes foothill where Mendoza city is established (Fig. 2). At the latitude of study area, Central Andes are comprised by three different geological provinces in argentine territory: Main Cordillera, Frontal Cordillera and Precordillera, from west to east. The Main Cordillera, comprising highest peak such as Aconcagua, involves Cretaceous- Jurassic marine rocks and vulcanites. Permo-Triassic volcanic Choiyoi Group outcrops mainly in the Frontal Cordillera; while Paleozoic sedimentary rocks with Permian intrusives do in Precordillera range (Fig. 2). Even though, lithology and slope are main conditioning factors for landslide distribution, this link is not analyzed in this work.
2. Precipitation threshold value Daily precipitation records suggest a relative low rainfall threshold for landslide occurrence in middle elevations ranging 1500 to 2700 m asl. during South Hemisphere summer (Dec-Feb). A daily precipitation range of 6.5 to 12.9 mm has been determined for landsliding in those locations not farer than 10 km of meteorological stations. This low threshold could be partially explained by the reduced amount of annual precipitation (200 mm) and the abundant generation of debris in these mountain areas. Errors in determination could have resulted from impossibility of determine intensity of rainfall. Mountain meteorological stations measure 24 hours precipitation. Likewise, meteorological records are scarce in the region limiting a precise determination of the threshold values. Meteorological stations are located along the Mendoza river valley, but no data exists for remote highest areas where events have been reported as well. Antecedent precipitation plays an important role. Mean values of accumulated rainfall reach to 28 mm whether a 5-day precipitation window previous to the landslide events is taken in account. In fact, 50 debris flows induced by rainfall during 2013 rainfall were associated to 29 days of weather anomalies in the Central Andes (Moreiras and Sepúlveda, 2013) (Fig. 1). Landslides are also associated to snowfall precipitation taking place during South Hemisphere winter in higher topography (See Figure 2). Greater slope instability is recorded in the following spring period associated to snow thawing (Moreiras et al., 2012). Herein, landslide triggering factors (rain /snow thawing) and temporal distribution (summer/spring) varies in the different ranges of the Argentinean Central Andes. Whereas debris flows and rockfall induced by rainfall are clustered in summer periods (December to February) along the valleys and lowest areas; debris flows and debris avalanches are consequence of snowmelt in highest areas of Main and Frontal cordilleras during spring (September-November).
3. Climatic anomalies forcing Temporal variations in landslide occurrence are related to climatic anomalies linked to the Pacific Anticyclone (ENSO). Humidity coming from western Pacific Ocean falls as snow in highest Andes areas and as rain in lowlands in winter season. During warm-phase of ENSO (El Niño) this precipitation is above mean annual records, while a precipitation below mean annual precipitation is recorded during cold phase (La Niña). As a consequence, landslides induced by rainfall are more common during El Niño years in the lower areas of the Frontal Cordillera, decreasing in number during La Niña events (Moreiras, 2006). Likewise, severity of debris flows generated by snow melting in the Main Cordillera is greater in the following year of above mean annual precipitation linked to the warm phase of ENSO phenomena (Moreiras et al., 2012). The Precordillera range, located east of the Frontal Cordillera, does not seem to be sensible to Pacific Anticyclone. Not significant differences were recorded between cold-warm ENSO events in this range (Moreiras, 2005; Moreiras, 2006). In contrast, slope instability in the Precordillera rises during wet periods induced by the incursion of wet air masses from the Atlantic Anticyclone during summer. Uspallata valley looks like a natural limit of influence of Pacific and Atlantic Oceans. Rain induced landslides are associated to behavior of both Atlantic and Pacific anticyclones in surrounding areas of this valley. Instability on eastern hillslope of the Cordon del Plata range, just toward the South of Uspallata, does not reflect any increase during warm phase anomaly of the El Niño. However, impacts and damages of historical events occurred during this anomaly resulted significantly more severe (Páez et al., 2013).
4. Future scenarios Climate change might affect the timing, phase and variability of precipitation, along with the resulting damaging events. General circulation model simulations predict a significant decrease in precipitation over the highest Andean region for the next decades (Cubasch et al., 2001), and an overall negative trend (Quintana, 2004) has been observed in annual and winter rainfall over Central Chile. However rainfall anomalies took place in last decades, like in summer of 2013 triggering exceptional volume of material and summer of 2014. Besides, a greater landslide activity is predicted as El Niño climate phenomenon may become more frequent and severe in the future (Diaz and Kiladis, 1992; Yamaguchi and Noda, 2006) suggesting a higher landslide hazard in the highest Central Andes involving Main and Frontal cordilleras. As well a positive trend for summer rainfall linked to Atlantic Anticyclone predicts more debris flows and rockfall in lower areas and valleys where main infrastructures and cities are established. Elevation of the 0°C isotherm as a response to global warming is forcing slope instability. Areas below 0ºC isotherm will be water –saturated and prone to mobilize. Rockfall and debris-flow records have increased during the last three decades due to both precipitation and temperature increases in the middle altitudes (Fig. 3). These changes could significantly impact facilities, agriculture, water resources and natural ecosystems. In particular, the mountain areas of Central Chile and Argentina were affected by damaging debris-flows associated with extreme summertime and wintertime precipitation events (e.g. Sepúlveda and Padilla, 2008; Moreiras and Sepúlveda, 2013). Analysis of synoptic situations
for summer and winter precipitation events have recently been documented (Viale and Nuñez, 2011; Viale and Garreaud); however, the extreme precipitation processes leading to debris-flow is still lacking in the study area. However, the relationships between landslide occurrence in Main Cordillera, Frontal Cordillera, and Precordillera with climatic anomalies induced by the Pacific and Atlantic oceans respectively, suggest the possibility of using the landslide chronologies in the Central Andes as a geo-indicator of global climate change.
References Araneo D. and Compagnucci R. 2008. Atmospheric Circulation features associated to Argentinean Andean rivers discharge variability. Geophysical Research Letters, ISSN: 0094-8276. Vol. 35, L01805, doi:10.1029/2007GL032427, pp. 1-6. Moreiras S.M. 2005. Climatic effect of ENSO associated with landslide occurrence in the Central Andes, Mendoza province, Argentina. Landslides Vol 2 (1): 53-59. DOI 10.007/s10346-005-0046-4. Moreiras S. M. 2006. Frequency of debris flows and rockfall along the Mendoza river valley (Central Andes), Argentina. Special Issue Holocene Environmental Catastrophes in South America. Quaternary International 158: 110-121. Moreiras S.M., Lauro C., and Mastrantonio L. 2012. Stability analysis and morphometric characterization of palaeo-lakes of the Benjamin Matienzo Basin- Las Cuevas River, Argentina. Natural Hazards 62 (2): 593-611. DOI: 10.1007/s11069-012-0095-7 Moreiras S.M., Lisboa S., and Mastrantonio L. 2012. The role of snow melting upon landslides in the central Argentinean Andes. Earth Surface and Processes Landforms. Special issue on Historical Range of Variability. Guest editors: Ellen Wohl and Sara Rathburn DOI 10.1002/esp.3239. Moreiras, S.M., and Sepúlveda, S.A. 2013. The high social and economic impact 2013 summer debris flow events in central Chile and Argentina. Bollettino di Geofisica Teorica ed Applicata, Vol. 54, Supp. 2, 181-184. Páez M.S., Moreiras SM, Brenning A., and Giambiagi L.B. 2013. Flujos de detritos-aluviones históricos en la cuenca del Río Blanco (32°55’-33°10’ Y 69°10’-69°25’), Mendoza. Revista de la Asociación Geológica Argentina 70 (4): 488-498. Sepúlveda. S.A.. Rebolledo. S.. Vargas. G.. 2006. Recent catastrophic debris flows in Chile: geological hazard. climatic relationships and human response. Quaternary International. 158. 83-95. Viale M. and Garreaud R. 2013. Summer precipitation events over the western slopes of the subtropical Andes. Mon. Wea. Rev. accepted. Viale M. and Nuñez M.N. 2011: Climatology of Winter Orographic Precipitation over the Subtropical Central Andes and associated synoptic and regional characteristics. Journal of Hydrometeorology. Vol. 12. Nº4. 481507.
Figure 1: Recent damaging events recorded along the Mendoza River valley induced by rainfall.
Figure 2: Above: Location of study area indicating localities and meteorological stations along the Mendoza River valley on a SRTM image. Below: Meteorological records: mean monthly precipitation in blue bars and mean monthly precipitation temperatures in red line.
Figure 3: Records of landslides induced by rainfall along the Mendoza River valley from 1950 to 2006. Red line indicates the percentage of total events.
