Keywords

3.1 Introduction

In the Semi-arid Andes, specifically in the high basin of the Río Cochiguás in the Coquimbo region, Chile, the glacial and periglacial environment is composed of complex geomorphological system, in which some landforms as rock glaciers are significant in number and size. However, this is a recent research topic in Chile, where it is still uncertain the dynamics, the inner structure and the water supply to the basins of these ice bodies. Additionally, under climatic tendencies that point towards a gradual atmospheric warming and an evident diminution in the precipitation in the area, it is necessary to study the effects that these processes would have on this type of cryogenic landforms.

This mountain landscape is characterized by steep slopes and deep valleys. The higher portion of the basin is related to a glacial and periglacial environment with the occurrence of talus, linear slopes, rock glaciers and solifluction and gelifluction slopes. Likewise, an important presence of rock and snow avalanches in the area (Paskoff 1967; Brenning 2005).

Due to the scarcity of precipitation during most of the year, the water reserves in this sector are mostly depending upon the quantity of snow accumulated in the High Cordillera and its later melting and infiltration, which feeds the basins of the main fluvial processes, as it is the case of the Elqui, Limarí and Choapa Rivers in the Coquimbo region. This climatic configuration of high mountains and their consequent fluvial streams constitute a primordial resource in the development of agricultural and cattle rising of the zone (Janke et al. 2015).

Presently, the region is being affected by the phenomenon of the “Central Chile mega-draught” that translates into a water deficit of approximately 30%, between 2010 and 2015 (Garreaud et al. 2017), and largely overcoming any other event during the rainy historical record. This problem would become more relevant in the future due to the tendencies that indicate a diminution of the rainfall and an increase in the temperature of the air in the surface of the Central Andes of Chile, which would continue also along the twenty-first century.

All this has promoted the interest of the communities living near mountain areas and the Chilean Government in investigating the availability of water resources, its dynamics through time and its spatial distribution, in which the rock glaciers have been particularly noted as potential water resources in arid zones.

This chapter aims to contribute to the knowledge of the cryosphere in the Andes by studying the morphology of rock glaciers, and the characterization of these landforms in the field. In addition, trying to relate their present characteristics with their volumetric variation over time and their possible water implications in high mountain basins.

3.2 The Study Area

The study area corresponds to the sub-basin of the Río Cochiguás, part of the basin of the Río Elqui, located between the 30°20′ and 30°29′S (Fig. 3.1a). It is limited to the north and northeast by the basin of Río Ingaguás, and has its sources in the high Cordillera at the boundary with Argentina. From an administrative point of view, the sector is located in the Paihuano County, Province of Elqui in the Coquimbo region, 46 km to the southeast of the Alcohuaz community in the Elqui Valley.

Fig. 3.1
2 maps and 2 illustrations. a, A map of South America highlights Chile and another map highlights the study area. b, the illustration highlights the Claro and Cochiguas Rivers and the study area. c, illustration highlights 2 rock glaciers labeled 1 and 2.

Study area. a Regional study area. The red square indicates the location of the detailed study area in (b). b Detailed study area showing the main rivers that originate in the basin. The black square indicates the location of the two main rock glaciers of this study. c Rock glaciers 1 and 2 studied in this chapter

Full size image

The studied rock glaciers are located at the inner part of the Quebrada Caballos that also belongs to this basin, placed between 30,44° and 30,47° south latitude, at the foot of a rock slope that indicates the international boundary between Chile and Argentina, at the heads of the Río Cochiguás.

The study area is located in the South-Central Andes, at latitude 30°28′S, where the subduction angle between the Nazca and South American plates is about 10°, a sector that is usually referred to as the “Central Chilean flat subduction segment” (Barazangi and Isacks 1976; Jordan et al. 1983; Gutscher et al. 2000) and as the “Pampean flat subduction segment” by Folguera et al. (2002). This segment is characterized by an intense coupling between the Nazca and South American plates, a highly compressed continental crust and the absence of a longitudinal valley separating the Cordillera Frontal (morphostructure of the Andes Mountains) from the Cordillera de la Costa (Pardo et al. 2006). Yáñez et al. (2001) suggest that the planar subduction in this region is related to the subduction of the Juan Fernández aseismic ridge at 33°S, which has been continuously subducting beneath South America at the same point since 10 Ma ago.

The large-scale geomorphology in the study area is characterized by a marked increase in elevation along the west–east transect (Aguilar et al. 2011, 2013). This first-order geomorphological feature represents a topographic front that separates two elongated morphostructural units in a north–south direction, corresponding to the Coastal Cordillera to the west and the Frontal Cordillera to the east.

The Coastal Range is characterized by a series of coastal platforms that present low slope (Paskoff 1970; Ota et al. 1995; Rodríguez et al. 2015), while, to the east, this reaches its maximum elevation at 3,200 m a.s.l. This morphostructure corresponds mainly to an east-vergent homocline of Triassic-Lower Cretaceous volcano-sedimentary rocks, unconformable to the metamorphic and sedimentary basement of Devonian-Permian age (Rivano and Sepúlveda 1991).

The Cordillera Frontal, on the other hand, reaches elevations of up to 6,700 m a.s.l. It is formed by a core of Carboniferous-Permian magmatic units (Nasi et al. 1990; Pineda and Calderón 2008), which are covered to the west by a block of Triassic-Lower Cretaceous volcano-sedimentary rocks and intruded by a Late Cretaceous-Early Paleocene magmatic arc (Mpodozis and Cornejo 1988; Nasi et al. 1990; Pineda and Calderón 2008). To the east, on the other hand, the basement is intruded by a block composed mostly of Permo-Triassic volcanic and magmatic rocks overlain unconformably by Oligocene-Miocene volcano-sedimentary rocks (Maksaev et al. 1984; Nasi et al. 1990; Bissig 2001; Murillo et al. 2017).

Finally, the area east of the main topographic front corresponds to the Main Cordillera, which is defined by a core of Oligocene-Miocene volcano-sedimentary rocks (Charrier and Baeza 2002; Mpodozis et al. 2009; Jara and Charrier 2014). These rocks are covered in unconformity by Miocene subhorizontal volcanic rocks and intruded by a north–south-trending Miocene magmatic arc (Mpodozis et al. 2009; Jara and Charrier 2014).

Regarding the lithology of the sector (Fig. 3.2), a series of intrusive, metamorphic and sedimentary units outcrop in the Cochiguás River Sub-Basin, most of them with ages ranging from the Paleozoic to the Triassic (Mpodozis and Cornejo 1988). These rocks correspond to the basement of the geomorphological units detailed below and represent the main lithology of the different deposits and geoforms identified in the sector.

Fig. 3.2
A geological map highlights the sediments, rocks, and others in and around Rio Cochiguas and Rio Claro Rivers. The rivers have the sediments of alluvial and coluvial sediments. Other sediments in the region include T b, P T r c, and P T r l.

Modified from Mpodozis and Cornejo (1988)

General geological map of the area with the main geological units defined by Mpodozis and Cornejo (1988).

Full size image

The main study area (Fig. 3.2) is essentially composed with Carboniferous rocks of the Elqui Super Unit (SUE), Permian-Triassic rocks of the Ingaguás Super Unit (SUI) and Miocene rocks of the Infiernillo Unit, all of them intrusive rocks.

On the one hand, the Elqui Super Unit is a group of plutons formed by leucocratic to mesocratic granitoids, coarse-grained, with synmagmatic, cataclastic and mylonitic banding, furrowed by swarms of basic (andesitic and andesitic-basaltic) and felsic dykes. It includes from gabbros to monzogranites.

In the study area, this super unit is represented by leucocratic granodiorites and monzogranites of biotite and muscovite, of medium to coarse grain and light grey to white colour (Cochiguás Unit); and by granodiorites and cataclastic granites of very coarse grain and hypidiomorphic texture, affected by fracturing and recrystallization (El Volcán Unit).

In the lower course of the Cochiguás River, the Cochiguás Pluton, the largest of the plutons of the unit of the same name, is well exposed. In the area of the Claro River, the plutons are crossed by swarms of subvertical basic dykes, which constitute more than 50% of the total volume of the pluton in some outcrops. The El Volcán Unit, on the other hand, is located in the area near the border with Argentina, intruding the Hurtado Formation to the west and the Pastos Blancos Formation to the east. The latter marks the eastern limit of the Elqui Super Unit intrusives through an intrusive-tectonic contact.

The Ingaguás Super Unit, on the other hand, is an association of generally hololeucocratic plutons, which are mostly disposed to the east of the outcrops of the Elqui Super Unit. The granitoids of this super unit intrude both the schists of the El Cepo Metamorphic Complex and the Pastos Blancos Formation and the granitoids of the Elqui Super Unit.

In the central sector of the study area (Fig. 3.2), this super unit is represented by medium-grained “pink” leucocratic monzogranites and syenogranites of the El León Unit.

The last group of intrusive rocks is composed of the Infiernillo Unit, of Miocene age, which correspond to medium-fine-grained quartz diorites, granodiorites, tonalites and subordinate andesitic porphyries. The bedding rocks are the El Cepo Metamorphic Complex, the Pastos Blancos Formation, Paleozoic granitoids and the Doña Ana Formation. Rocks of this unit outcrop in the eastern sector of the study area, in a N–S strip near the border with Argentina.

In addition to the intrusive rocks mentioned above, there is an important area to the south-southeast of the study area made up of Upper Paleozoic-Lower Triassic volcanic rocks of the Pastos Blancos Formation, which correspond mainly to rhyolitic lavas, ignimbritic flows, tuffs, intercalations of volcanic breccias and scarce levels of sandstones and conglomerates.

About the climate, in general terms, the study area is located in a transition zone between arid and semi-humid climates, immediately south of what is known as the “Arid Diagonal”, which crosses South America from the Gulf of Guayaquil, in Ecuador, and ends on the Atlantic coast of Patagonia, between 25° and 27°S (Eriksen et al. 1983; Zech et al. 2006; Ginot et al. 2006). To the north of this diagonal there is a greater influence of the Amazon System (SMS, South American Monsoon System), while to the south it is conditioned by the Pacific Circulation System and the westerlies, where glaciers generated by humid fronts from the Pacific and winter precipitation develop (Ammann et al. 2001).

The climate of the area also depends on the presence of the Pacific anticyclone. This is a high-pressure system located in the southeastern Pacific Ocean where the atmospheric pressure is higher than the surrounding area, which produces a downward movement of the air masses that inhibits the development of cloudiness and precipitation, and favours clear skies with high solar radiation (Escobar and Aceituno 1998).

In general, the current climate of the study area according to the classification of W. Köppen (in Errázuriz et al. 1998) corresponds to tundra due to the effect of altitude with little or no precipitation. This climate is a subtype of the high-altitude marginal desert climate and develops continuously in the highest part of the Andes Mountains, from 28°S.

Regarding seasonal variations, the area is characterized by cold winters and dry summers (Fiebig-Wittmaack et al. 2012), by a strong diurnal temperature variation between day and night, and a strong altitudinal effect on temperature (Azócar 2013). Most of the area's humidity is due to solid precipitation (snow) between May and August (Gascoin et al. 2010). However, during the summer, small snowfalls caused by humid air masses from the eastern side of the Andes are observed, mainly between January and March (Garreaud and Rutllant 1997) and also by convection of humid air from the west in the afternoons.

The 0 °C Mean Annual Air Temperature (MAAT) isotherm is located at approximately 4,150 m a.s.l. However, it tends to be at 3,700 m during the coldest month (July) and over 4,800 m in the hottest month (January) (Azócar 2013).

On the other hand, in the highest zones of the study area, and in particular where rock glaciers 1 and 2 are located, the MAAT varies between 0 °C (at 4,150 m) and −2° (at 4,400 m), according to the calculation developed by Azócar (2013) (Fig. 3.3).

Fig. 3.3
A study area map highlights the mean air temperature. Mean air temp decreases from 4 to negative 4 moving from the north to the southeast. A northward compass is on the top right and a scale measuring 2 kilometers in on the bottom right.

Annual mean air temperature curves in the study area. Schematic developed from the MAAT modelling of Azócar (2013). UTM 19S coordinates

Full size image

3.3 Periglacial Environment and Rock Glaciers

To understand what these cryo-landforms are, it is necessary to understand the context under which they have developed. According to the Terminological Guidebook of the South American Cryology (Trombotto et al. 2014), the periglacial environment is a cold and cryogenic environment with the occurrence of permafrost in depth, or permanently frozen soils, and the possible presence of underground ice trapped and preserved under natural conditions after a long time. Moreover, the freezing processes dominate with cycles of freezing and melting that affect the rocks and the upper portion of the soil, as well as solifluction, gelifluction and other cryogenic processes that model the geomorphology of the Andean regions.

The cryogenic rock glaciers are the expression of mountain permafrost (Barsch 1996) that correspond to landforms of periglacial genesis. The common definition corresponds to located bodies (in the slopes) or tongue-like (at the inner portion of the valleys) composed of unconsolidated materials, which are permanently frozen, supersaturated of interstitial ice, and which moves downslope by creeping as a consequence of the deformation of the ice contained inside (Barsch 1996).

There exists presently a debate about the periglacial or glacial origin of a rock glacier, which is closely related with its inner structure, a matrix of ice and rock or an ice core (Janke 2013). Those authors defending the periglacial model (cryogenic rock glaciers) believe that the interstitial cemented ice and/or the segregated ice (ice lenses) provokes a lithostatic pressure that allows the glacier movement by means of creeping (Wahrhaftig and Cox 1959; Barsch 1978, 1996; Wayne 1981; Haeberli 1985). Besides, those favouring a glacial model (glacigenic rock glaciers) suggest that an ice core would have its origin in an uncovered glacier (or “white”), covered and buried by debris (Potter 1972; Whalley 1974; Whalley and Azizi 1994; Potter et al. 1998). In this sense, the rock glaciers are frequently considered as part of a continuum in the landscape, that is, a cycle that describes the transition between uncovered glaciers, rock glaciers and slope deposits (Corte et al. 1987; Johnson 1984; Giardino and Vitek 1988). However, the present evidence support both models (Janke 2013).

It is accepted today that a rock glacier may have a cryogenic origin or a glacigenic origin.

Humlum (1988) developed a classification of debris ice components in mountainous areas based upon morphological criteria. In this classification, many landforms are recognized, such as valley glaciers, ice caps, cirque glaciers, rock glaciers (tongue shaped, valley floor, piedmont or spatulate), protalus rampart and extant or relict snow patches.

The degree of activity of a rock glacier is one of the most used classifications, proposed by Barsch (1996), who, starting from a set of dynamic characteristics, performed interpretations according to the presence of ice to the inner part of rock glacier, based upon the principle of the presence of sub-superficial ice in different cut sections (Haeberli et al. 2006). In this category, the rock glaciers are classified into active landforms, that means, in movement and with ice in the inner side, inactive landforms, without movement, but still with ice in its interior, and as fossil or relict, without movement and where the ice content has melted completely.

According to Barsch (1996), a rock glacier is considered as active when it shows signs or evidence of movement, when its front is strongly sloping (around 30–35°) or when in a vertical section it exhibits sedimentation processes and visible signs of rock fall. Their active surfaces present superimposed lobes, undulations and grooves characteristically caused by movement.

Besides, an inactive rock glacier is one that has stopped moving and it shows evidence of past but not recent movements. While its front presents a slope smaller than 30º, its surface is chaotic with depressions and collapse signals. The inactivity is generally the result of warming tendencies that have caused a diminution of the ice content of the ground (Trombotto et al. 2014). Both active and inactive forms are commonly grouped as intact rock glaciers, due to the difficulty in differentiation between active and inactive landforms by means of remote sensing (Barsch 1996).

Finally, a relic or fossil rock glacier is a mass of rock fragments and finer materials, in a certain slope, that shows evidence of the last movement, but it has lost all the soil and ground ice. Its surface is already vegetated and the fronts present slopes of less than 20° (Trombotto et al. 2014).

Concerning the lithology, this is an important factor in the shape and size of the debris that composes the rock glacier. According to studies done in the Alps, the larger clasts are associated with crystalline rocks, for instance, igneous rocks, conglomerates and limestones, whereas those of smaller size are related to foliated and/or high porosity rocks, such as slates (Ikeda and Matsuoka 2006). This has allowed to classify the rock glaciers into two types according to the size and shape of the debris: the pebbly rock glaciers and the bouldery rock glaciers (Ikeda and Matsuoka 2006). Figure 3.4 summarizes and defines what rock glaciers are, how they are formed and their classification according to the literature reviewed in this chapter.

Fig. 3.4
A classification chart of the definition and classification of rock glaciers. It is defined as a mass of rock fragments on a slope, depicts evidence of past or present, and produces creep and deformation of internal ice. Classification includes origin, morphology, location, activity, and grain size.

Definition and general classification of rock glaciers based on origin, morphology, location, activity and size of blocks in their composition

Full size image

3.4 Methodology

A geomorphological characterization performing a remote mapping was done. High resolution of imagery (1.7 m) of February 2015 obtained from the National Service of Geology and Mining was used by means of the base map Worldview from Arcgis. The main morphological features of the area were identified, and two rock glaciers were chosen to be studied.

Afterwards, a geomorphological map at 1:25.000 scale was prepared based upon the information taken in the field, determining common criteria to the associated morphologies and characterizing the unconsolidated deposits of the area. The elaboration of this map was performed using the ArcGIS 10.4 software and the Adobe Illustrator 2018 software. With these digital tools a morphometric analysis of slopes and solar exposition was completed.

Likewise, a lithological and sedimentary characterization of the rock glaciers was performed with the objective of identifying the rocks’ provenance and the patterns of compression, based upon the methodology used by Monnier et al. (2014). This methodology consists in the election of experimental points (sampling and measurements) within and around the landform, whose objective is to complement the surficial mapping and the delimitation of the main units.

To study the volumetric and morphological changes of the rock glaciers, the photogrammetric method was used. It allows the generation and comparison of Digital Elevation Models (DEM) constructed on different dates and, starting with them, it was possible to estimate the change of volume of the rock glaciers during a certain period (∆h/∆t). With this purpose, aerial photographs taken by the Instituto Geográfico Militar (IGM) of Chile were used as the basis of a Digital Elevation Model of high resolution (10 m) provided by the Servicio Nacional de Geología y Minería of Chile.

Control and elevation points through GNSS (GCPs, Ground Control Points) using two receptors of double frequency were taken to generate the DEM of the years 1955 and 1999, both in outcrops close to the study area, at the foot of the rock glaciers (Fig. 3.5). These control points have been taken to improve the precision of the generated elevation models, providing better precision. These points were chosen taking into consideration the conditions of the terrain such as climate, accessibility, logistics, as well as the proximity of rock glaciers.

Fig. 3.5
A satellite image of the study area highlights the control points. The control points are scattered on the slope with a series of control points extending between RG1 and RG2.

Control points in the terrain (GCPs), taken in outcrops of the Quebrada Caballos and at the foot of the rock glaciers 1 and 2. In yellow, the contours of the rock glacier 1 and rock glacier 2. Base map: Worldview ArcGIS, February 2015. Coordinate system UTM 19 S, Datum WGS 84

Full size image

Besides, the orthorectification of the aerial photographs HYCON 1955, GEOTEC 2000 and the generation of MDEs (each pair in a separate way) were generated using the software PCI GEOMATIC 2016. In this manner, the software provides a DEM of metric resolution and with a good level of detail, which may be improved by eliminating GCPs with high values of residuals in X, Y or in Z.

With the base of the results obtained from the calculation of the altimetric differences between the generated DEM, the behaviour of the rock glaciers was analysed for the period 1955–2018 corresponding to each difference, recognizing the main zones of loss or increase of volume and detecting the changes in thickness.

3.5 Geomorphological Context

As a first look, this section studies the geomorphological characteristics of the study area where the two rock glaciers of study are located in order to contextualize them, so as to understand their characteristics and historical variations.

Presently, the sub-basin of the Río Cochiguás is controlled by periglacial processes. This type of processes implies a readjusted and remodelled landforms and glacigenic deposits developed during a glaciation (Ballantyne 2002).

The existence of a large debris accumulation at its base indicates that they are being affected presently by cryoclastism, by mass-movement processes and fluvial action, whereas the good development of their morphology (U-shaped valleys, polished bedrock, well-preserved moraines) indicates a prolonged evolution under glacial-cryogenic conditions. Most of them are in contact with rock glaciers at their base.

Likewise, the basal moraines are the most abundant landforms in the study zone. They cover the base of the glacial valleys of Quebrada Caballos, Toro Muerto and Vallecillo. This type of moraine is the best preserved in the area, frequently presenting work by fluvial processes (Fig. 3.6a). The surface is characterized by a hilly topography of “hummock” type (Fig. 3.6b), and the deposit, of similar shape to the other types of moraines, is greyish, poorly sorted and matrix-supported, with blocks of more than 1 m in diameter faceted and polished in the higher sectors.

Fig. 3.6
Two photos of hilly landscape from the south and north views. Top, the landscape under the mountain is labeled basal moraine with a wetland in the center. Bottom, the rocky mound is labeled basal moraine, hummocky surface, and a wetland in between.

Basal moraine belonging to the joint of the Toro Muerto and Vallecillo Valleys. a Basal moraine cut down by creeks, in whose sides wetlands have been developed. b View towards the north from the surface, where it may be observed the surface, where the hilly landscape of the deposit can be appreciated

Full size image

An orangish deposit is located at the foot of the rock glaciers 1 and 2 of the Quebrada Caballos, at 3,978 m a.s.l. with an area of 0,68 km2 and 1,229 km length (Fig. 3.7a). The shape of this deposit was probably generated by a rock avalanche coming from the SE. Due to its position in the valley, it is possible to infer that it corresponds to an event older than the formation of the two rock glaciers (Fig. 3.7b) which could occur due to a catastrophic rock flow of the pre-existing, weathered rock slopes. However, it is necessary to point out that due to its morphological and depositional complexity, it is difficult to characterize it as a rock avalanche.

Fig. 3.7
2 illustrations and 2 photos labeled a to d. a and b highlight rock avalanche deposit and 2 rock glaciers. c, the photo of the rock avalanche deposit slope from the east view. d, a mound is outlined and labeled as the front slope of the rock avalanche deposit.

a View in plant of the mass-movement deposit identified in the bottom of the Quebrada Caballos. b View towards the E of the scarp where it could have generated the movement, noticing also the two rock glaciers of the valley. c Hilly, hummocky surface of the avalanche deposit. d Front of the deposit, with greyish boulders exposed

Full size image

One of the important results of this work was the updating of the geomorphological map of the study area, based on all the observations presented in this chapter, considering geomorphologic, cryogenic and glacigenic features in the High Andean Cordillera, which is shown in Fig. 3.8.

Fig. 3.8
A geomorphological map and 2 maps of South America. Top, the geomorphological map highlights the landforms such as glacial, periglacial, cryogenic, landslide, and fluvial landforms, hydrography, and symbology. Two South American maps highlight the location of the Coquimbo region in Chile

Updated geomorphological map of the study zone considering periglacial and cryogenic landforms

Full size image

In addition, as a result of the geomorphological mapping and in order to complement the public inventory of glaciers of the DGA (2015), 11 rock glaciers have been identified and classified as active or inactive (Fig. 3.9). They have a total surface of 1.27 km2, concentrated above 4,000 m a.s.l. in the bottom valleys of Caballo and Vallecillo with abrupt rock scarps. Different rock glacier types are recognized taking into consideration such parameters as morphology and position.

Fig. 3.9
A study area map highlights the rock glaciers. The locations marked are Quebrada del Toro Muerto, Vallecillo, and Quebrada Caballos. They have moraine landforms. The map also highlights symbology and landforms in and around the study area.

Updated rock glacier inventory for the study area, and associated landforms and geomorphological processes

Full size image

The surface of these bodies is covered by coarse materials, and they present a surface with crests and curved transverse depressions. One of the more remarkable characteristics, especially observed in the rock glaciers 1 and 2, would correspond to surficial runoff coming from the foot of both landforms, including also in situ evidence of runoff/water flow to the inner portion of the body off the rock glacier 1.

In addition, extensive slopes characterized by the presence of lobes and wrinkles are predominantly located in higher zones (above 4,100 m a.s.l.). These slopes are thus observed because they are directly related to the soil grain size, which is generally of affine grained type. This happens because, if there is enough snow that can subsequently melt (and recrystallize to deepen as ice in winter) this surface becomes inflated in a perpendicular manner to the ice, whereas when the soil is seasonally unfrozen, the surface is almost parallel to the force of gravity, that is, it occurs vertically, showing such deformation (French 2007). The lobes are the resulting landforms, being along the length and width of the slope.

Cryoweathering is a process that occurs in the rock outcrops, being more evident in the Vallecillo Valley (Fig. 3.10). It corresponds to the mechanical disintegration of the soil or rock, as a consequence of the processes of freezing and melting, ice pressure and hydration (Trombotto et al. 2014).

Fig. 3.10
A satellite image and 3 photos. Left, the satellite image highlights the cryogenic features on the slope. Top right, it highlights the basal moraine surface, and right center, it highlights cryogenic features from the north view. The bottom right highlights the cryogenic features from the south view.

Basal moraine of the Quebrada del Toro Muerto. Towards the left, satellite image of the deposit, where the surfaces in black are delimiting the surfaces with special features. To the right, photographs of the deposit in different angles, evidencing the sorting recurrence due to freezing in depressions of the deposit, which concentrate most part of the oxidized, orange clasts

Full size image

The selection of loose materials constitutes one of the most notorious evidences of the action of water freezing and melting of the ice in cryogenic environments. Evidence of landforms affected by these processes correspond to the southwestern lateral moraine located south of the Quebrada Caballos (MLW) and the basal moraine of the Quebrada del Toro Muerto.

According to the presence of ice, in the upper portion of the rock glacier 2, at almost 4,270 m a.s.l., a groove containing refrozen ice in its inner core has been dug up to 50 cm from the surface. The debris cover contained is coarse, the size of the blocks oscillates between 20 and 30 cm and even bigger ones, reaching sizes of up to 70 cm in some cases. The ice is massive and translucid, and it contains inside, clasts of smaller size, frozen and united by the ice. It is found in the sedimentary matrix of the debris cover cementing the clasts of smaller size, thus being classified as cement ice or matrix ice.

3.6 Rock Glaciers Characteristics

The delimitation of the basal and surface contours of rock glacier 1 and 2 was carried out from GPS points taken on site and by remote observations due to the complexity of the terrain, which made it impossible to take data in situ. The result of the contours obtained is shown in Fig. 3.11.

Fig. 3.11
A contour map delineates the study area, showcasing rock glaciers 1 and 2. The contours are widely spaced over the rock glaciers, whereas in the southern region, they are closely packed. Inferred contours and frontal scarps are evident surrounding the glaciers.

Contours of rock glaciers 1 and 2 delimited by GPS points taken in the field and in remote observations. Dashed line represents inferred areas due to terrain complexity. Base map: Worldview ArcGIS February 2015. UTM 19 S coordinate system, WGS 84 datum

Full size image

Thus, it is possible to point out that the rock glaciers have a multilingual surface shape, and their upper part is connected directly from the base of the rock walls (cirques) at the bottom (in the case of rock glacier 2) or very close to them (rock glacier 1).

On the other hand, both are approximately 1 km long and have an area of 0.22 km2 (rock glacier 2) and 0.26 km2 (rock glacier 1). They are oriented E-W (rock glacier 1) and SW (rock glacier 2) and have a similar frontal angle, between 35° and 37°. However, there are parts of the front that reach values greater than 40°, especially in rock glacier 2.

The rock glacier 1, located under the southernmost rock slope at the bottom of Caballos Creek, shows typical compression features such as transverse ridges along the maximum length of the body, as well as grooves or depressions towards the end of the body. The latter are the most visible in satellite images and are found between cords and ridges, also transversal (Fig. 3.12). These forms of compression are more recurrent in the most distal part of the glacier, in the area with the lowest surface slope. There are also longitudinal depressions and ridges, but these are not as recurrent as the transverse type, and are located in the central-high sector of the body, at the lateral ends of the body (Fig. 3.12).

Fig. 3.12
A satellite image highlights the study area with transverse ridges, longitudinal ridges, and rock glacier 1.

Plan view of rock glacier 1, where the ridges visible in the satellite image, which are concentrated in the lower limit of the body, have been marked with a black line. Base map: May 7, 2008 image from the Google Earth server

Full size image

On the other hand, the rock glacier 2 is located immediately under rock slopes with southern exposure and SW exposure, giving this landform a preferential orientation towards the SW. Morphologically, it is different from rock glacier 1, because its surface is characterized by pronounced and deep depressions both transversely and longitudinally, such as those shown in Fig. 3.13.

Fig. 3.13
A satellite image of the study area with transverse ridges, longitudinal ridges, and rock glacier 2.

Satellite image of rock glacier 2 showing the main longitudinal and transverse ridges identified. Base map: May 7, 2008 image from Google Earth server

Full size image

However, contrary to what was observed in rock glacier 1, the grooves or depressions resulting from compressional processes are not so numerous or so close to each other, and are located mainly in the lower zone of the glacier. There are abundant longitudinal depressions located mainly in the centre of the body, which also mark an evident change in the surface coloration of the detrital cover. The surface shape is not as elongated as the neighbouring glacier, and its source of nourishment apparently comes directly from rock outcrops located near its upper boundary.

Important characteristics of this glacier include the presence of cryogenic features such as cryometeorization, circular depressions, an irregular cover characterized by deep longitudinal grooves up to 10 m deep, the presence of internal ice under the detrital cover in the upper zone of the rock glacier, snow forms such as penitentes and terraced snow covered by detritus.

The presence of snow was found in the upper portions of the rock glacier 2 in the shape of penitents, with a height between 10 and 50 cm. They have also been observed in nivation niches above the rock glacier 2, covering an area of small dimensions. Additionally, at the upper boundary of the glacier, surface snow has been found in the shape of terraces, which are presently covered by a thin layer of fine-grained debris, of approximately 5 cm in thickness. On these snow terraces, it is possible to observe debris of larger sizes, distributed in a chaotic manner. These debris of larger dimensions are also under snow benches. The height of each snowbank is approximately 40 cm and the snow that composes them is soft and dirty. This buried snow would correspond to multiannual snow, which could be related with the surficial ice, as noted before.

Concerning the historical variation and estimated volume of the rock glaciers of Quebrada Caballos, using the analysis of the digital elevation model (1955–2018) it was found that the average altimetric variation for the rock glacier 1 is −2.56 m with a standard deviation of 6.34 m, whereas for the rock glacier 2, the average value of the altimetric variation corresponds to −10.26 m, with a standard deviation of 5.60 m. Therefore, the average of the altimetric difference for this period for both glaciers is −6,41 m, which represents an average diminution of 10.17 cm per year. In this way, the total variation for the 1955–2018 period corresponds to −6,41 m, that is, −10 cm per year.

Assuming a highly conservative condition of total saturation of interstitial spaces in the ice, the annual variation is of approximately 27,000 m3. This is equivalent to a fluvial runoff of 0.002 m3/s considering four melting months (summer season), which is an insignificant value in terms of error ranges of discharge measuring methodologies.

3.7 Discussion

This high mountain landscape in the eastern slope of the Central Andes is the result of exogenic agents during the Quaternary, both of glacial and periglacial type. Moreover, the posterior modelling associated with the interaction of the periglacial system with gravitational processes, cryoclastism, seismicity, fluvial, debris transportation and of chaotic torrential character (such as mixed cones and saturated flows) rules the local morphodynamics at present.

The magnitude of the erosional processes associated with the ice in the past has been influenced by the structural characteristics of the zone, the relief orientation and by the rock competence due to its degree of fracturing. According to Paskoff (1970), those are key factors in the modelling of the relief. This would largely explain the magnitude of the erosional landforms as well as those of accumulation landforms such as moraines and talus cones distinguished both along the entire Cordillera del Elqui (Paskoff 1970).

The moraines of the study zone, especially those of the lateral type, are noticeable by their symmetry, similarities in the composition of their deposits and a very well-preserved morphology. This “perfection” in the symmetry of the moraines has been attributed by Paskoff (1970) to a pre-glacial condition, where the glaciers would have flown along valleys which were very wide already, since from a beginning due to the uplifting of the mountain chains in the third epoch of the Andean Cycle, in a manner that they were retouched by the action of the ice and later in-filled with an appropriate sedimentary load/discharge.

The glacial valleys that dominate the study area are mostly of short dimensions and quite wide at its base.

In relation to the landforms generated by mass-movement processes, the abundance of mixed talus cones has been originated from the fragmentation of the rocks due to cryoclastism, a process that increases with altitude and that generates abundant landforms of this type along the mountain systems (Paskoff 1970). In a general way, the slopes free of ice provide abundant debris for the development of cryogenic rock glaciers, being these of the talus or valley bottom types (Brenning et al. 2007).

The geomorphological map obtained is mostly coincident with those shown by Paskoff (1967). However, there exist several differences that keep relationships with the scale at which the mapping has been done (1:25.000 vs. 1:250.000), as well as the omission of other morphologies due to their relationship with other processes. In this sense, Paskoff (1967) indicated that the area is dominated by landforms and deposits of glacial or nival origin, as these processes are the principal modelling agents of the landscape. Notwithstanding, the presence of other morphologies linked to gravitational, fluvial or periglacial is not indicated but they are widely recognized in the study area.

The glacial evidence presented indicates that colder climates occurred in the past, allowing the development of valley glaciers whose tongues would have reached up to the confluence of the Río Claro with the Río Cochiguás, partly evidenced by the basal moraines located in Toro Muerto, Vallecito and Quebrada Caballos valleys (Figs. 3.6 and 3.10). Additionally, the location of the rock glaciers of the Quebrada Caballos and some of the Quebrada Vallecillo allows their classification as valley bottom, rock glaciers, whose origin is associated with glacial processes (Humlum 1982). This permits to suggest that these landforms are morphogenetically related to glacial processes in its origin, and to periglacial processes in their evolution, constituting a continuous geomorphic system (Soto et al. 2002). In this manner, it is possible to remark that the main rock glaciers studied (rock glaciers 1 and 2) are glacigenic rock glaciers, whose origin would be associated with ice bodies existing in the past in the semiarid Andes (Riquelme et al. 2011). The geomorphological sequence of the construction of the landscape in the Quebrada Caballos may be proposed as follows:

  1. 1.

    A first step related to a glacial advance responsible for the deposition of the lateral moraines of the valley (MLW and MLE) and of the latero-frontal moraine. These moraines are located at high elevations and are morphologically well preserved that would allow to assign them to the Last Glacial Maximum (32 ka approximately, Zech et al. 2007) or to an episode of glacial advance reported for the semiarid Andes between 17 and 12 ka B.P. (Zech et al. 2006, 2007, 2008).

  2. 2.

    A second phase of glacier retreat due to more arid conditions during the Holocene (Ammann et al. 2001; Grosjean et al. 1998; Riquelme et al. 2011) gave path to periglacial conditions, increase in the instability of the slopes due to deglaciation at the same time of an increase in the debris production and valley in-filling forming dejection cones and debris cones, due to the transition towards a more arid, warm environment, conditioned to freezing process and seasonal melting.

  3. 3.

    A last phase of complete transition towards a full periglacial environment with the predominance of cryogenic features and landforms and rock glaciers, both glaciogenic and cryogenic. Under these conditions, the rock glaciers accumulated snow meltwater that refreezes forming ice in the superficial ice/rock matrix, afterwards liberated, at least partially, during the summer (Corte 1988).

The identification of 11 rock glaciers with varied morphological differences among them and different placement sites allows to infer the existence of landforms of different origin, coeval with the same morphoclimatic domain, obeying singular topoclimatic and geomorphological conditions that have favoured their formation and development.

This number of active and inactive rock glaciers differs with the number indicated by the National Glacier Inventory (DGA 2015) (Fig. 3.14). In this case, only three active rock glaciers were found in the studied valleys, corresponding to the rock glaciers 1, 2, and 4 of the inventory performed in this present study (Fig. 3.9). In the terrain, it was possible to verify that the remaining landforms numbered as 3, 5, 6, 7, 8, 9 and 10 (Fig. 3.9) actually correspond to rock glaciers, making it very probable that the methods used in the former inventory from DGA would not be the most appropriate ones, for instance, the use of old satellite imagery of low resolution makes impossible a good definition of the landforms.

Fig. 3.14
A geological map of the study area highlights rock glaciers. Rock glaciers in the south are active, and the contour lines on the map are widely spaced.

Rock glaciers (active and inactive) identified in this study. The glaciers also identified in the National Glacier Inventory (Inventario Nacional de Glaciares, DGA 2015) are in red. Base Map: hillshade developed from DEM SRTM 2000, contours of IGM. System of coordinates UTM 19 S, datum WGS 84

Full size image

It was also observed that in both glaciers there is an important supply of debris from the lateral slopes of both rock glaciers, which is especially evident in the upper zone of rock glacier 2, due to the presence of talus cones, debris flows and rockfalls. In this sense, these glaciers play an important role in the system of debris transportation (Brenning 2005), since the presence of these landforms not only stabilizes the debris material in itself, but also constitutes a base level for fast mass-movement processes as avalanches, rock falls and even small landslides (Brenning 2005).

Concerning their thinning, rock glacier 2 would be characterized by a gradual but continuous reduction of the inner ice, where the amount of internal debris would become gradually proportional to the ice quantity. It will be completely covered by debris and ice would not be observable at the surface (only buried), something that was corroborated in the terrain and typically holding values of 25–45% of ice. Among the additional characteristics that are coincident with those observed in the rock glacier 2, there are pronounced crests and grooves both transversal and longitudinal, in addition to the longitudinal groove. This last one would indicate the reduction of the ice core in this rock glacier. According to Janke et al. (2015), the consequence due to an atmospheric warming in this type of glaciers is the continuous reduction of the ice under the surface, which would make slow the movement and would reduce its capacity as a water reserve. In a segregated or interstitial manner, the front appears more or less pronounced and elongated, whereas the transversal chains are observed to be more of a lineal type.

To be able to establish the quantitative supply of the rock glaciers and the actual ice content in their inner portion, it is needed to make later studies that would include, for instance, coring combined with geophysical and geochemical studies, so as to obtain real data related to the inner structure, water provenance and full water supply.

Although the model performed to evaluate the rock glacier variations along time is not the most precise, it indicates tendencies that were already estimated from time ago in those studies related to mountain permafrost (Mollaret et al. 2019). Therefore, to improve even more the precision and results, it is needed to perform more studies in the semiarid zone of Chile, following the vertical variations using for that purpose more sophisticated methodologies, as drones, high-resolution imagery, techniques of Structure from Motion (SfM), etc.

The values obtained indicate that there is a mean thinning of the ice of 10 cm per year in the period between 1955 and 2018. However, the loss rates per year vary for each glacier. In fact, rock glacier 1 has experienced a thinning of 4 cm per year, whereas rock glacier 4 has experienced a thinning of 16 cm per year. These results are very different in between them, and they differ in different periods, especially in the period 1999–2018, when it has been observed a mean increase in thickness in the 1 m vertical of rock glacier 1, and a drastic thinning of the mean elevation of rock glacier 2, with a value close to—22 cm per year. These differences in the mentioned period may be explained mainly by: (1) a larger debris cover in rock glacier 1 which allows for a better protection (Lambrecht et al. 2011), (2) by methodological mistakes in the photogrammetric process or 3) because the period comprised (1999–2018) is not enough to find reliable results.

According to the literature, the active rock glaciers are preferably located in zones where the mean annual temperature is lower than −1 or −2 °C and where the precipitation does not exceed 2,500 mm/year (Haeberli 1985; Barsch 1996). In the case of the studied valleys, the origin area of all rock glaciers is located at a mean elevation of 4,400 m a.s.l., with a mean annual temperature equal or lower than −2 °C, coincidently with that stated above, the frontal zone is approximately located at 4,100 m a.s.l., in zones where the mean annual temperature is slightly above 0 °C. Consequently, the distribution of active rock glaciers over this elevation occurs with a mean annual temperature of at least 0 °C. At this latitude, these morphologies are common, even in the cirques exposed towards the NW or NE, with a high input of solar radiation (Brenning 2005).

From these observations, it is possible to state that the active and inactive landforms between 3,900 and 4,400 m a.s.l. in this high mountain sector may reflect past thermal conditions, due to the large time response scale of the rock glaciers and of the permafrost in general (in the order of decades and centuries, after Haeberli (1990). This would suggest an increase of at least 1 °C due to climatic changes, which would not impact in great manner on the rock glaciers situated under the −2 °C isotherm, since they would be kept under cold conditions. However, the rock glaciers located in isotherms between 0 and −1 °C (three of them and the front of the rock glaciers 1 and 2) could develop a greater sensitivity to temperature increase and due to this warming could cause more melting (Azócar 2013).

It is especially important the presence of fresh snow terraces covered by debris, the penitents of ice and snow identified in the upper part of rock glacier 2. This snow is probably coming from avalanches in the higher portions, and it would allow to explain, partially, the incorporation of snow inside this part of the glacier due to the melting dynamics and inner refreezing, also providing evidence of the fast debris production on the slopes that occur at the heads of the valleys and which determines the dynamics of these landforms. This indicates the importance of snow precipitation that would favour the supply of summer discharge in the zone, forming part of the ice contained in the interior of the debris cover, which freezes and refreezes seasonally.

Besides, in the surroundings of both rock glaciers, and particularly in rock glacier 1, continuous or discontinuous snow was not observed, but the presence of inner runoff in the glacier was seen, what allows to infer that due to a greater incident solar radiation with respect to the rest of the study zone, there is a greater rate of snow melting that runs along the inner body of the rock glacier.

Taking into consideration the studied period, the thinning rates observed in the rock glaciers 1 and 2 would be responding, partly, to historical tendencies of atmospheric warming and diminution of precipitation, especially stressed during the last decades due to the ENSO phases and the aridification of the Central Andes (Santibañez 1997; Le Quesne et al. 2006; Vuille and Milana 2007).

However, if the precipitation deficit close to 30% since the year 2010 is considered related to drought events of diverse severity, intensity and duration, a phenomenon that has been named as the “megadrought of Central Chile” by Aldunce and González (2009) and Garreaud et al. (2017). According to it, a smaller quantity of snow would have influence in small recharge and thermal isolation for the underlying ice, which, together with a regular increment of the temperatures, would affect at least the more superficial layer of intrinsic ice of the glacier. This cause/consequence relationship is applicable to precipitation, therefore, to the negative historical variation of both rock glaciers in the studied period, understanding that these climatic tendencies have been observed since past decades.

Clearly, the rock glaciers respond in a different manner depending upon the thermal régime, the microclimate conditions, the debris supply, the topography and geological parameters (Arenson and Jakob 2010). In spite of the fact that the Central Andes seem to offer optimal conditions for the development of rock glaciers, it is also susceptible to atmospheric warming so a pronounced and generalized change is expected in the rock glaciers existing in this sector. The morphological changes of these glaciers in a warmer climate are still unknown, although more of them would become relics once that their boundaries stay below the 0 °C isotherm (Bodin et al. 2012; Iribarren Anacona et al. 2014).

Finally, it is estimated that in the near future, the atmospheric warming would elevate the 0 °C isotherm, around 500 m above the present position in the worst scenario, even more than during the Holocene. Therefore, it is expected that in the Central Andes, close to the 95% of the active ice-content landforms would have its terminal position below the 0 °C isotherm around 2070 (Drewes et al. 2018).

3.8 Conclusions

The geomorphological characteristics of the high basin of the Río Cochiguás clearly show how the different morphogenetic agents have had influence in different time scales on the present configuration of the landscape. This is reflected in the existence of landforms directly attributed to the glacial action, as cirques and glacial valleys, that presently show in their surface some of the most common features of the erosion systems of periglacial environments.

It is also recognized the importance to have available a precise geomorphological map at the time to suggest the uses of the territory in the mountain areas. Particularly, because it was determined that in the same morphoclimatic domain or altitudes, there may exist a great variety of processes depending on the mixing of multiple factors.

The previous information cited here, as well as the relationship between morphological, geological, topoclimatic and regional characteristics in the evolution and present historical variation of the rock glaciers, are theoretical self-supported. Nevertheless, it is needed the exploration and study of these landforms in mountain areas of large dimensions and with great geological variations to validate these results at the scale of a semiarid region. In this sense, there is still much to do about the real water implication of these cryoforms and the impact that the climatic change has on them, as well as about their dynamics and possible mechanisms of feeding/melting, what has a clear relationship with the interaction between solid precipitation, the resulting warming and the role of the different thickness of the debris covers, as they act as a thermal transmitter or a thermal separation.

3.9 Limitations and Updatings

This research was performed in 2018. Until today (2022), the International Permafrost Association (IPA) has developed new lines of research and technical handbooks for the development of inventories of rock glaciers under new morphological standards that may differ from what has been presented in this chapter.