Você está na página 1de 11

PERMAFROST AND PERIGLACIAL PROCESSES Permafrost and Periglac. Process. (2012) Published online in Wiley Online Library (wileyonlinelibrary.

com) DOI: 10.1002/ppp.1736

Rock Glacier Activity in the Retezat Mountains, Southern Carpathians, Romania


Alfred Vespremeanu-Stroe,1* Petru Urdea,2 Rzvan Popescu1 and Mirela Vasile1
1 2

Faculty of Geography, Bucharest University, Bucharest, Romania Geography Department, West University of Timioara, Timioara, Romania

ABSTRACT Thermal conditions in seven rock glaciers (RGs) in the Retezat Mountains were examined using ground surface temperature, the bottom temperature of snow (BTS) and late-summer water temperature at springs. Direct current resistivity proles were used to detect ice-rich layers and to estimate RG volume. These parameters were supplemented by measurements of RG surface movement and morphological characteristics in order to develop a RG typology. RG activity is suggested based on BTS values and the mean annual surface temperature. The lower altitudinal limits for permafrost occurrence and the activity of RGs were found to be at 2000 m and 2100 m asl, respectively. The two highest and most active RGs are younger than 8 ka and developed following a cold climate event at 8.2 ka when a last isolated glacial episode occurred in their hosting cirques, while the remaining RGs date from 11.58.7 ka. Calculated rockwall denudation rates to produce these forms range from 0.81.1 mm/yr. Copyright 2012 John Wiley & Sons, Ltd.
KEY WORDS:

mountain permafrost; rock glacier; BTS; ground temperature; denudation rate; southern Carpathians

INTRODUCTION Information on permafrost conditions in the eastern and southeastern European mid-latitude mountain ranges is limited, but studies have shown perennially frozen ground to be present at sites in the High Tatras (Gdek and Kdzia, 2008) and the southern Carpathians (Urdea, 1992, 1993). Based on freezing and thawing indices and on-site investigations, the lower limit of permafrost is predicted to be at elevations between 1900 m and 2300 m above sea level (asl) in the Tatras, 2000 m asl in the southern Carpathians and 2350 m asl in the Musala and Rodopy Mountains (the Balkans) (Dobiski, 2005). It is difcult to make a precise determination of permafrost distribution, however, due to the inuence of local factors and the patchy nature of surfaces beneath which permafrost might occur. These include accumulations of frost-shattered rocks in rock glaciers (RGs), protalus ramparts, debris cones and scree slopes, which are particularly conducive to the development and preservation of mountain permafrost (Barsch, 1996). The southern Carpathians may represent the eastern boundary of mountain permafrost in Europe. Landforms there indicate extensive periglacial activity during the
* Correspondence to: Dr A. Vespremeanu-Stroe, Faculty of Geography, Bucharest University, 1 Bl. N. Balcescu, sector 1, Bucharest 01023, Romania. E-mail: fredi@geo.unibuc.ro

Pleistocene which is now restricted to elevations greater than 1800 m asl. Early publications (Ichim, 1978) identify three major areas (the Retezat, Parng and Fgra Mountains) with a frequent occurrence of RGs and describe their morphological characteristics. The presence of permafrost was underlined by Urdea (1992, 1993, 2000), who investigated the thermal characteristics of two RGs in the Retezat Mountains, and by Kern et al. (2004). Detecting and mapping of permafrost in RGs can be undertaken in a variety of ways including the bottom temperature of snow cover (BTS) method which has been widely used in alpine environments (Haeberli, 1973; Hoelzle, 1996; Ishikawa and Hirakawa, 2000; Lewkowicz and Ednie, 2004; Julin and Chueca, 2007). In order to verify BTS measurements and to establish the annual thermal regime of RGs, however, continuous monitoring is needed (Hoelzle et al., 1999). In fact, complete investigation of RGs requires a range of methods to determine the presence of permafrost including measurements of ground temperatures, spring temperatures, direct current (DC) resistivity, seismic sounding and RG dynamics. This paper assesses permafrost occurrence in RGs and their activity in the central sector of the Retezat Mountains by applying complementary eld techniques to seven RGs. A second aim is to establish the palaeoclimatic signicance of the evolution of RGs in the area, based on the assessment of morphometric parameters and denudation rates. Given that permafrost-related geomorphic landforms can be very
Received 2 March 2010 Revised 10 February 2012 Accepted 11 February 2012

Copyright 2012 John Wiley & Sons, Ltd.

A. Vespremeanu-Stroe et al.

sensitive to climatic changes, special attention was paid to the relationship between RG status and the present climate. STUDY AREA The Retezat Mountains are situated in the western part of the southern Carpathians (45 20 N, 22 23 E), having a general ENE-WSW orientation (Figure 1). Elevations higher than 2000 m asl represent an area of about 56 km2 and the highest peaks reach about 2500 m asl (Ppua 2504 m, Peleaga 2509 m, Bucura 2432 m, Retezat 2482 m asl). Given high Cenozoic surface uplift rates (Chalot-Prat and Grbacea, 2000; Reuther et al., 2007), the area may not have been glaciated until the Late Pleistocene (Urdea, 2000). Consequently, glacial valleys and cirques, moraines and glacial lakes are very well demarcated. During cold Pleistocene stages, the area subject to periglacial conditions

enlarged. Granites and granodiorites are the main bedrock components of the central sector and contribute, due to frost weathering, to the formation of medium and large blocks that accumulate in deposits such as RGs, debris cones, protalus ramparts or scree slopes, usually at elevations exceeding 1700 m asl. The timberline is situated at about 1800 m asl, and above this, vegetation (alpine meadows and shrubs with Pinus mugo extending up to 2200 m) is restricted because of the great extent of debris and block accumulations. The mean annual air temperature (MAAT) is 2.3 C (19592009) at the Vrfu Omu meteorological station (2505 m asl) and 0.5 C at the arcu station (2190 m asl). The 0 C MAAT isotherm is at about 2000 m asl on south-facing slopes and 2100 m asl on north-facing ones. The mean annual rainfall above 2000 m asl is 11001400 mm, with northern slopes being wetter. Snow cover is present around 200 days/year, reaching an average thickness of

Figure 1 Location of the central sector of the Retezat Massif: (Up) Topographic map; (Down) orthophotomap. This gure is available in colour online at wileyonlinelibrary.com/journal/ppp

Copyright 2012 John Wiley & Sons, Ltd.

Permafrost and Periglac. Process., (2012)

Permafrost Research in the Southern Carpathians, Romania

1.52.0 m at high elevations during JanuaryMarch. More than 105 days exhibit diurnal freeze-thaw cycles at the 2500 m asl level (multiannual average) (Vespremeanu-Stroe et al., 2004) mainly during spring and autumn. Four of the RGs analysed (Valea Rea, Pietrele, Pietricele and tirbu) are located in northern valleys, while the other three (Judele, Lower Ana and Upper Ana) are situated on the southern side of the range, in the largest glacial complex of the southern Carpathians (Bucura), but oriented east and northeast. The average elevation of the RGs is 2110 m asl (Table 1), while the adjacent peaks are at 2340 m asl. The RGs vary in length from 200 m (tirbu) to 1430 m (Valea Rea) with a mean of 480 m. tirbu and Upper Ana are lobate RGs, Pietrele, Valea Rea, Judele are tongue-shaped and Lower Ana is spatulate (Figure 2). The ridge-furrow topography on all the RGs is due to their composition of coarse blocks (bouldery RG; cf. Ikeda and Matsuoka,

2006). Except for the Judele and tirbu RGs, however, their fronts and lower segments have sparse vegetation, indicating an absence of movement on much of their surfaces. METHODOLOGY Thermal Investigations Ground surface temperatures (GSTs) were measured every 2 h on all the RGs in 200809 and in addition on the Pietrele, tirbu and Judele RGs in 200910. DS 1922L iButton loggers with an accuracy of 0.5 C were used. In the rst year, 15 loggers were set out, with seven sensors on the Judele, tirbu, Valea Rea, Upper and Lower Ana RGs and eight sensors on the Pietrele RG, in order to examine temperature variability across the surface. Sensors were

Table 1 Elevation, aspect and thermal characteristics of the investigated rock glaciers. Rock glacier Judele tirbu Pietrele Valea Rea Pietricele Upper Ana Lower Ana Status Active Active Complex Complex Inactive Inactive Inactive/relict Elevation (m) 2175 2131 2042 2116 2157 2075 1996 Aspect 61 325 11 335 308 22 44 (ENE) (NW) (N) (NNW) (NW) (NNE) (NE) BTS ( C) 3.4 2.8 2.8 2.9 2.5 1.1 SWT ( C) 0.7 1.8 1.6 1.8 MAST ( C) 0809/0910 0.3/-1.2 0.3/-0.5 0.3/-0.6 1.2 0.9 2.2 Long-term MAST ( C) 0.9 0.3 0.3 0.6 0.3 1.6

Note: Complex status indicates varying degrees of activity within the same rock glacier. BTS = Bottom temperature of snow; SWT = spring water temperature; MAST = mean annual surface temperature.

Figure 2 Rock glaciers: (A) Valea Rea; (B) Pietrele; (C) and tirbu. This gure is available in colour online at wileyonlinelibrary.com/journal/ppp

Copyright 2012 John Wiley & Sons, Ltd.

Permafrost and Periglac. Process., (2012)

A. Vespremeanu-Stroe et al.

placed in holes drilled 2 cm beneath the top of a boulder or hung between boulders to avoid direct exposure to solar radiation. During the period of BTS sampling the two methods gave similar results (Hanson and Hoelzle, 2004), but the latter provides a better assessment of the mean annual surface temperature (MAST). Average temperatures recorded during the thermal stability period (late winterearly spring) were considered to be BTS data. The mean air temperature during the rst year of measurement, recorded by an automatic meteorological station placed in the Valea Rea valley, was 1 C warmer than the long-term average (1.3 C compared to 2.3 C). We estimated the long-term MAST (Table 1) by relating the upper soil temperature during the measurement period to the multiannual soil temperature recorded at Vrfu Omu, resulting in values 0.6 C lower than those measured. This adjustment is supported by MAST values obtained for the Pietrele, tirbu and Judele RGs in the following year (200910), which are 0.9 C lower. Water temperatures at springs located on the front of the Pietrele, Pietricele, Valea Rea and Judele RGs were measured with a portable thermometer ( 0.2 C accuracy). Temperatures in several water sources emerging from landforms other than RGs within the study area were also measured. BTS measurements were carried out on the Pietrele RG with a 3 m long manual probe with a thermistor in its tip with 0.1 C accuracy. The values were interpolated by kriging in Golden Surfer Software to develop a permafrost distribution map. RG Dynamics and Morphometry Measurements of movement were initiated on Pietrele and Valea Rea RGs in 2008 and 2009, respectively, using a high-precision Sokkia 610 total station. The magnitude and direction of horizontal movements were determined by means of successive topographic surveys. The main morphometric parameters of the RGs and surrounding slopes were computed in ArcGIS and GlobalMapper, based on a digital elevation model (with 20 m cell size) and orthophotomaps (0.5 m resolution), while the RG frontal lobe slope was computed on site with the total station. To evaluate the morphometry of the debris-source rockwalls, a threshold slope of 30 was used to differentiate between the rockwalls and surrounding debris cones and slopes. The rockwall surface used in the calculation was assumed to be continuously weathered up to present. The elevation growth of debris cones was computed to have occurred continuously and uniformly throughout the Holocene, thereby covering the base of the former debrisfree rockwall. DC Resistivity The electrical resistivity tomography (ERT) technique is based on differences in the electrical resistivity of various subsurface materials. In the present study, a PASI 16GS24N
Copyright 2012 John Wiley & Sons, Ltd.

system, consisting of 32 electrodes, with a standard spacing of 5 m, was used for the Pietrele and Lower Ana RGs. Longitudinal and transverse resistivity proling was undertaken downslope of the roots of the two RGs and crossing their central parts (Pietrele 2043 m asl, 45 2213N, 22 521E; Ana 1989 m asl, 45 2113N, 22 5204E). Dipole-dipole and Wenner-Schlumberger arrays were used at the two sites. RES2DINV software (Locke, 1999) was used to invert the data. ERT proling was also performed in a previous study (Urdea et al., 2008a) in order to evaluate mean RG thickness in the Southern Carpathians. RESULTS BTS The BTS measurement data on Pietrele RG were compared with the continuous measurements obtained from the miniature temperature sensors placed on the RG in order to validate them. BTS measurements were performed at 127 points covering the entire RG surface of roughly 11 ha (Figure 3) at the end of March when the snow cover had an average thickness of 2.3 m on the RG. Values ranged from 0 C to 5.7 C. Temperatures near 0 C occurred in the frontal and central areas with sparse vegetation. A small area in the western central part of the RG exhibited negative temperatures down to 3.1 C. This can be accounted for by the high degree of shading of the western slope and the trench-like aspect of the surface, which facilitates the longer

Figure 3 Permafrost distribution map for the Pietrele rock glacier derived from bottom temperature of snow cover measurements. P1P8 are locations of the permanent temperature loggers. This gure is available in colour online at wileyonlinelibrary.com/journal/ppp

Permafrost and Periglac. Process., (2012)

Permafrost Research in the Southern Carpathians, Romania

duration of snow cover in summer. BTS values less than 3 C were measured in the upper third of the RG. Assuming that the rules of thumb developed for the Alps apply (BTS > 2 C permafrost improbable, BTS between 2 C and 3 C possible permafrost, BTS less than 3 C probable permafrost; cf. Haeberli, 1973; Haeberli and Patzelt, 1982), permafrost is probable at 66 of the sampling points. Temperatures between 3 C and 2 C were recorded at 20 points concentrated in adjacent areas where permafrost occurrence is possible. Interpolated temperature data indicate that probable permafrost may occupy 3.1 ha and possible permafrost 1.6 ha, together representing about 43 per cent of the surface of the Pietrele RG. Spring Water Temperature Late-summer spring water temperature is considered a good indicator of permafrost existence in RGs. If water originates from the melting of interstitial ice, the temperature should be no higher than 2 C (Warhafting and Cox, 1959; Haeberli and Patzelt, 1982). Moreover, the atmospheric temperature and its oscillations should not affect the thermal regime. In normal circumstances, all active and inactive RGs should display frontal springs because of interior ice melting during the summer. Individual measurements were made in August 2008 at four RGs where the springs were reachable (Valea Rea, Pietricele, Pietrele and Judele). Flowing subsurface water was also heard in the tirbu RG but could not be reached and it could not be detected in the other two (Upper and Lower Ana). Eight measurements of spring water temperatures were made at non-RG landforms in order to conrm the thermal threshold of ice-derived spring waters. Water temperatures differ between RG and non-RG springs at the same elevation (Figure 4). Non-RG springs have temperatures greater than 4 C and these are negatively correlated with elevation (R = 0.52). All springs originating

Figure 4 Individual measurements of spring water temperatures in the central part of the Retezat Mountains. Rock glacier (RG) frontal springs are represented by closed circles: P = Pietrele; PT = Pietricele; J = Judele; VR = Valea Rea. NonRG springs from neighbouring areas (the same glacial cirques) are represented by open circles. TLPS = Thermal limit of permafrost springs (+2 C).

from RG fronts display temperatures less than 2 C, ranging between 0.7 C and 1.8 C, which suggests the presence of permafrost. There was a signicant difference between these temperatures and the mean air temperature for the measurement period (6.1 C for 15 July31 August 2008), while non-RG spring temperatures (4.17.5 C) were similar to air. Thus, it is probable that the main source of RG springs is the thawing surface of the frozen core. The small differences between RG springs may relate to heat transfer with the atmosphere, active-layer conditions and the length of the subsurface water course. The Judele RG spring is suggestive of this process: its temperature was 0.7 C at the base of the RG front, 1.4 C at 90 m lower and 1.6 C at the entrance to the Florica glacial lake (130 m from the front). Therefore, there is no correlation between RG spring water temperature and elevation. Temperatures were measured continuously over the summer at springs issuing from the Pietrele and Judele RGs in 2008 and the Pietrele and Valea Rea RGs in 2009 (Figure 5). Pietrele and Valea Rea springs displayed mean

Figure 5 Spring water temperatures at Pietrele, Valea Rea and Judele rock glaciers compared to air temperatures and precipitation measured close to 2180 m in the Valea Rea valley in: (A) summer 2008; and (B) summer 2009. TLPS (thermal limit of permafrost springs) (+2 C) is marked with a dashed line.

Copyright 2012 John Wiley & Sons, Ltd.

Permafrost and Periglac. Process., (2012)

A. Vespremeanu-Stroe et al.

temperatures of 1.7 C and 1.6 C with pronounced isothermal regimes indicated by standard deviations of only 0.04 C and 0.07 C, respectively, re-conrming the impact of permafrost. Rainfall events less than 30 mm had a minor impact, with spring water temperatures rising less than 0.5 C for a few hours. Only a rainfall event of 50 mm (2223 July 2008) and the rst snowfall (13 October 2009) raised the spring water temperature to values of 23 C. Temperature change was less for the larger RG (Valea Rea occupies 0.43 km2, 3.8 times larger than Pietrele) than the smaller one, but the impact lasted longer (Figure 5). GST Regime GST monitoring illustrates the annual thermal regime of RGs and provides information on the likelihood of permafrost maintenance (Ribolini and Fabre, 2006). The GST annual cycle for each of the RGs was divided into six main periods for 200809 (Figure 6). Period A represents snow-free time in autumn when ground surface cooling is intense and the lowest temperatures are reached. During this period, the supercial active layer (as measured by implanted sensors Upper Ana and tirbu in Figure 7) experiences large temperature uctuations but these diminish with depth (hanging sensors Pietrele and Judele in Figure 7). During period B (November to mid-January),

Figure 6 Annual ground surface temperature (GST) cycle at a typical Retezat rock glacier divided into six periods and compared with air temperature (measured every 2 h) in the Pietrele glacial cirque: (A) ground surface cooling during snow-free interval; (B) shallow snow cover (< 60 cm) allowing ongoing heat transfer between the active layer and the atmosphere; (C) bottom temperature of snow cover window corresponding to stable temperatures; (D) zero curtain when snowmelt and inltration raise ground surface temperatures to 0 C; (E) active-layer thaw period; and (F) air temperature oscillates around 0 C.

the shallow snow cover (< 60 cm) initially causes an increase in GST due to homogenisation of the active-layer temperature caused by cold air inltration through snow funnels (chimney effect) (Hanson and Hoelzle, 2004; Delaloye and Lambiel, 2005). This process is favoured by the blocky structure of the RGs. Subsequently, GST shows moderate oscillations as a function of measurement depth. Thus, temperatures rise slowly at the surface as heat moves

Figure 7 Ground surface temperatures of six rock glaciers. The temperature threshold for possible permafrost during the bottom temperature of snow cover period (2 C; Haeberli, 1973) is shown by a dashed line.

Copyright 2012 John Wiley & Sons, Ltd.

Permafrost and Periglac. Process., (2012)

Permafrost Research in the Southern Carpathians, Romania

from the interior through the displacement of warm air by colder air inltration. Period C includes the BTS window when snow depths are at their greatest. Mean temperatures for period C indicate permafrost is probable for the Pietrele and Judele RGs, possible at Valea Rea, tirbu and Upper Ana RGs and improbable at Lower Ana (Figure 7; Table 1). Period D, which follows, is the snowmelt and zero curtain period. The BTS window ends suddenly at all the RGs except for Pietrele and Upper Ana and the zero curtain lasts 5070 days due to signicant amounts of superimposed ice formed between the boulders by refreezing of melt water, leading to a delay in warming of the active layer. Period E represents thawing and warming of the active layer, when warm air masses and rainfalls (812 C) directly impact the permafrost. Period F is the early-autumn transition period when air temperatures oscillate around 0 C. Data collected in 200910 for the Pietrele and Judele RGs showed similar thermal regimes in terms of pattern, but there was a delay of 2030 days due to a later start of winter and a shorter period C. The annual regime enables permafrost maintenance (cf. Ishikawa, 2003), because intense heat loss can occur during period A, there is progressive cooling of the active layer in period B and a prolonged period D preserves low temperatures at the ground surface, in contrast to much higher air temperatures. As mentioned above, MAST values in 200809 (when most of the RGs were investigated) were generally positive due to a warmer than normal year (Table 1). In comparison, 200910 values for the Pietrele and Judele RGs were almost 1 C lower.

ERT Proles The inverted DC resistivity proles of the Pietrele and Lower Ana RGs (Urdea et al., 2008a, 2008b) exhibit resistivities ranging across several orders of magnitude (Figure 8), indicating near-surface layers with unfrozen water and deeper high resistivity values which can be interpreted as permafrost bodies (see Kneisel et al., 2008). The Pietrele RG resistivity prole ran across ridge and furrow topography in the upper part of the RG where most BTS values were less than 3 C (see Figures 3 and 9). An undulating layered structure was revealed using a dipole-dipole survey, with an apparent limit to the supercial layers at a depth of approximately 7 m and a possible thickness of the deposits between 15 and 20 m. It may be that the resistivity prole is imaging bedrock at its base, but the high root mean square error and lack of corroborating information do not allow a rm conclusion to be reached in this regard. The prole for the Lower Ana RG (Figure 8B) ran longitudinally within the upper third of the RG and shows an anomaly with high resistivity values (approximately 700 km) at depth beneath the upper and central parts and lower values in the lower part of the prole. This suggests the existence of permafrost at the upper end of the prole and possibly lower ice contents or liquid water at the lower end. The unfrozen openwork boulders forming the active layer explain the lower resistivities of the surface layers. The high resistivity anomalies (> 400 km) indicate the presence of ice-rich cores in these bouldery RGs.

Figure 8 Longitudinal direct current resistivity proles on rock glaciers: (A) Pietrele; and (B) Lower Ana. The location of the prole in (A) is shown in Figure 9.

Copyright 2012 John Wiley & Sons, Ltd.

Permafrost and Periglac. Process., (2012)

A. Vespremeanu-Stroe et al.

Figure 9 (A) Map of Pietrele rock glacier showing locations of the dynamics monitoring (dashed line box), iButton digital thermometers (P1P8) and direct current (DC) resistivity prole; and (B) annual movement rates in the investigated sector.

Dynamics of the Pietrele RG In the upper third of the Pietrele RG, the numerous vegetationfree ridges alternating with furrows and micro-depressions suggest the possibility of creep. The results of the eld measurements revealed annual average horizontal velocities of 28 cm/yr (200810) in the upper part of the RG and no detectable trend in the lower part (Figure 9). Signicant differences in movement rates and orientations are present over small distances. This suggests that the movements are independent on each of the lobes constituting the upper sector of the RG. The areas subject to movement generally accord with locations where BTS measurements indicate that permafrost is probable (see Figure 3). Given the relatively low mean RG slope of 8 , we presume that the movements represent horizontal and/or vertical boulder movements (re-arrangement of the RG structure) due to creep and melting of the ice core (Haeberli, 1985). DISCUSSION Permafrost Occurrence and RG Activity in the Retezat Mountains The RGs in the study area are located close to the lower mountain permafrost boundary based on their elevations and mean annual temperatures. Most show possible or probable permafrost conditions (Table 1), but the inference of their activity status requires a more thorough assessment. Based on our results, the Judele and tirbu RGs display temperature values which are most suited to permafrost
Copyright 2012 John Wiley & Sons, Ltd.

existence and they have favourable topographic conditions for permafrost-related movement. Both RGs display relatively high average slope values, are completely vegetation-free and the negative long-term MAST values indicate that permafrost can be preserved under the present climate. The major favourable conditions for permafrost preservation in these two RGs are the high mean elevations (> 2100 m asl), their orientation (NW for tirbu and NE for Judele) and the surrounding topography, which reduces the radiation input and helps preserve snow. For example, snow patches are present on the Judele RG until mid-July. The topography of the surrounding cirque also creates a network of avalanche areas that converge on the RG surface and create a very thick, dense snow cover. Consequently, the Judele RG has a very short period E when the active-layer temperature rises above 0 C. The Pietrele and Valea Rea RGs exhibit similar morphologic and topographic characteristics. They have the greatest lengths and areas, very pronounced tongue-shaped forms and ow-microforms covering their entire surfaces. Temperatures in the upper sectors indicate probable permafrost. The data obtained from the sensors on the Pietrele RG show considerable variation in temperature: MAST was 0.3 C/-0.6 C (200809/200910) at P6 and 1.6 C at P8 (Figure 9), while BTS varied from 4 C (P4) to 0 C near the front (P2). The preliminary movement rates (28 cm/yr) along with the moderately low long-term MAST (Table 1) indicate a transition from active to inactive on the upper third of the RG, while the lower segment is generally relict with no permafrost and the two lower thirds are partially vegetated. The Pietrele RG and probably Valea Rea RG are therefore complex forms where the three activity statuses coexist (Figure 10).
Permafrost and Periglac. Process., (2012)

Permafrost Research in the Southern Carpathians, Romania

Palaeoclimatic Implications The extensive surfaces of the Valea Rea and Pietrele RGs are due to the existence of rockwalls about 200 m high, which provide their debris and shade their upper parts (Table 2). Both RGs are tongue-shaped, with several crescentic compression ridges and a complex system of creep lobes which partially override each other or interact laterally, indicating climatic oscillations (Frauenfelder and Kb, 2000). They were probably fully active during the cold phases of the Holocene (Subboreal and LIA) when the MAAT was up to 1 C lower than at present (Ribolini and Fabre, 2006). Consequently, the Valea Rea and Pietrele RGs are not only the largest but were probably also the most dynamic RGs in the Retezat Mountains in the past. The post-LIA warming episode slowed their movements, with the middle and lower sectors becoming inactive or relict. The approximate ages of the Pietrele and Valea Rea RGs can be estimated from their lengths, current movement rates in their active parts (28 cm/yr) and mean RG velocities reported internationally (515 cm/yr; Konrad et al., 1999; Serrano et al., 2006). For the 560 m long Pietrele RG, mainly fed by the rockwall corresponding to the central crest and assuming mean values of horizontal displacements of 58 cm/yr (due to irregular movement rates inside active RGs, it is more appropriate to use mean rates closer to the lowest velocities recorded), an approximate age of 11.58.7 ka results. The same procedure was applied to Valea Rea RG, using a mean RG length of 690 m due to the additional debris source area represented by the western lateral crest. The resulting age ranged between 13.88.7 ka. Considering that this overlaps with the Younger Dryas when glaciers were present in the area (Reuther et al., 2007), we assume the common age of RG formation to be 11.58.7 ka. The estimated volumes of RGs and adjacent debris cones (deposit porosity estimated at 50%), obtained with the help of morphometric parameters and DC resistivity proles, indicate mean rockwall denudation rates of 0.41.1 mm/yr if we assume a theoretical common age of 10 ka since the

Figure 10 Classication of rock glacier activity status in the Retezat Mountains based on measured thermal conditions (mean annual surface temperature (MAST) and bottom temperature of snow cover (BTS)).

The slightly positive long-term MAST values and the BTS data (Table 1) suggest that Upper and Lower Ana RGs are inactive, or in transition from inactive to relict for the latter. Although the DC resistivity proling suggested the presence of ice-rich materials at depth, the overall aspect and conguration of Lower Ana indicate that it is relict. In turn, this suggests that it may be in disequilibrium with the current climate and is slowly thawing. Based on the spring water temperatures, we assume that the Pietricele RG probably contains permafrost but is also inactive. Relict RGs are common below 2000 m asl in the Retezat Mountains, indicating that mountain permafrost was previously more extensive. Relict RGs such as Slveiu, Rade and Pelegua, close to the study area, could prove useful for palaeoclimatic reconstruction. There are several cases when RGs seem to be in transition from active to inactive (the upper parts of complex RGs) or inactive to relict (Lower Ana) (Figure 10), indicating that about half of RGs that still preserve ice are in thermal disequilibrium with the present climate, which initially renders them inactive and subsequently causes them to thaw. These changes likely started in the post-Little Ice Age (LIA) warming phase and are currently accelerating due to the rising MAAT (Figure 11). The MAAT for 19902009 was 2.1 C at the Vrfu Omu meteorological station, 0.4 C higher than for the195990 period.

Table 2 Morphometric characteristics and mean denudation rates of rockwalls adjacent to the rock glaciers. Rock glacier Judele tirbu Pietrele Valea Rea Pietricele Upper Ana Lower Ana Status Active Active Complex Complex Inactive Inactive Inactive/ relict RGS (km2) 0.039 0.036 0.108 0.415 0.017 0.038 0.081 RWS (m2) 0.220 0.117 0.278 0.523 0.084 0.098 0.134 DR (mm/yr) 0.37 0.55 0.79 1.10 0.89

Figure 11 Mean annual air temperature for the last 50 years at the Vrfu Omu meteorological station (elevation 2500 m asl).

Note: Complex status indicates varying degrees of activity within the same rock glacier. RGS = Rock glacier surface; RWS = rockwall surface; DR = apparent denudation rate computed relative to an assumed initiation date of 10 ka.
Permafrost and Periglac. Process., (2012)

Copyright 2012 John Wiley & Sons, Ltd.

A. Vespremeanu-Stroe et al.

last glacial evacuation of debris. The estimated rockwall denudation rates apparently differ between the two active RGs (0.40.6 mm/yr) and the rest (0.81.1 mm/yr) (Table 2). The frost-weathering control factors (lithology, microtectonics, exposure, rockwall relative elevation) are similar and do not account for these contrasting denudation rates. These differences suggest shorter periods for debris accumulation within the two active RGs (Judele and tirbu), which are probably younger. Over the Holocene as a whole, therefore, denudation rates of 0.81.1 mm/yr apply to the central sector of the Retezat Mountains. Recent studies document the maximum Wrmian extension of glaciers in the Retezat Mountains at 1716 ka, and the presence of high erratic blocks (18001870 m asl) indicates both their altitudinal development and the age of the last major glacial advance (13.611.4 ka; Reuther et al., 2007). The latter corresponds to the Younger Dryas spread of alpine meadows and to tree cover reduction between 12.911.5 ka in the Romanian Carpathians (Farca et al., 1999). Signicant debris accumulation started only with the disappearance of glaciers 11.510 ka, so permafrost could have developed when boulder deposits were large enough to sustain the development of perennial ground ice. Within glacial cirques surrounded by high rockwalls and large shaded areas (e.g. Valea Rea), small retreating glaciers may have gradually become debris-covered and later on evolved into RGs. However, we interpret the cold climate event of 8.2 ka ago that lasted a few hundred years (Mayewski et al., 2004) as the ideal enlargement phase for most of the RGs in the Retezat Mountains. A different scenario is likely for the two currently active RGs, whose smaller volumes indicate lower denudation rates (Table 2) and, probably, younger ages. Their simple morphology and the topographic characteristics of their hosting cirques which are both the highest (above 2100 m asl) and shadiest in the Retezat Massif (Tables 1 and 2) support this hypothesis. These conditions could have led to the isolated development of cirque glaciers during the climatic event 8.2 ka ago so that these active RGs are younger than 8 ka. CONCLUSIONS What is probably the easternmost mountain permafrost in Europe is present within RGs in the southern Carpathians. Four different types of RGs were identied based on their spatial distribution and characteristics such as MAST, BTS, morphometry and vegetation cover: (i) active RGs

which may be in thermal equilibrium with the present climate (e.g. Judele and tirbu); (ii) large, complex RGs where all permafrost activity states coexist within the same RG body (e.g. Valea Rea and Pietrele); (iii) inactive RGs with no current movement (e.g. Upper Ana and Pietricele); and (iv) relict RGs at elevations below 2000 m asl, which are the most common, with at least one above timberline in almost every formerly glaciated valley. The annual ground thermal regime on the RGs studied can be divided into distinct periods corresponding to changes in heat ow between the ground and the atmosphere. Most important for permafrost maintenance are periods of intensive heat loss in autumn and long insulation of the surface from positive air temperatures by snow in early summer. The transition from active to complex RGs and the restricted permafrost areas of Pietrele and Valea Rea suggest their disequilibrium with the current climate. The lower limit of mountain permafrost (the elevation between inactive and relict RGs) is therefore identied as being at 2000 m asl, and the RG activity limit is 21002130 m asl on northern slopes. Both limits are 100 m higher on southern slopes. Denudation rates of the source rockwalls calculated for the last 10 ka range from 0.8 mm/yr to 1.1 mm/yr. We infer two categories of age for RGs in the region: (i) current complex, inactive and relict RGs above 1800 m asl with ages estimated at 11.58.7 ka; and (ii) two currently active RGs which started their development after 8000 BP, probably due to the presence of glacier ice in their high cirques during the climatic cooling event 8.2 ka ago.

ACKNOWLEDGEMENTS This study was partially supported by the National University Research Council (CNCSIS research grant 738 MEDALP). We thank Nicolae Cruceru, Florin Ttui, Luminia Preoteasa, Alexandru Onaca and Florina Ardelean for their technical support during the eldwork campaigns and their useful discussions on permafrost topics. We are also grateful to Mircea Ardelean, Olga Pavel, Nataa Videanu, Cezar Radu-Buturez, Loredana Bizgan and Mdlin Oprea for their kind assistance in the eld. Special thanks are addressed to Antoni Lewkowicz for his contribution to the paper quality improvement.

REFERENCES
Barsch D. 1996. Rockglaciers: Indicators for the Present and Former Geoecology in High Mountain Environments. Springer: Berlin. Chalot-Prat F, Grbacea R. 2000. Partial delamination of continental mantle lithosphere, uplift-related crust-mantle decoupling, volcanism and basin formation: a new model

for the Pliocene-Quaternary evolution of the southern East-Carpathians, Romania. Tectonophysics 327: 83107. DOI: 10.1016/ 50040.1951(00)00155-4 Delaloye R, Lambiel C. 2005. Evidence of winter ascending air circulation throughout talus slopes and rock glaciers situated in the lower belt of alpine discontinuous permafrost (Swiss Alps). Norsk Geogrask TidsskriftNorwegian Journal of

Geography 59: 194203. DOI: 10.1080/ 00291950510020673 Dobiski W. 2005. Permafrost of the Carpathians and Balkan Mountains, eastern and southeastern Europe. Permafrost and Periglacial Processes 16: 395398. DOI: 10.1002/ppp.524 Farca S, de Beaulieu J-L, Reille M, Coldea G, Diaconeasa B, Geory C, Jull T. 1999. First 14C datings of Late Glacial and

Copyright 2012 John Wiley & Sons, Ltd.

Permafrost and Periglac. Process., (2012)

Permafrost Research in the Southern Carpathians, Romania


Holocene pollen sequences from Romanian Carpathes. Comptes Rendus de lAcadmie des Sciences de Paris, Sciences de la vie 322: 799807. Frauenfelder R, Kb A. 2000. Towards a palaeoclimatic model of rock-glacier formation in the Swiss Alps. Annals of Glaciology 31: 281286. DOI: 10.3189/ 172756400781820264 Gdek B, Kdzia S. 2008. Winter ground surface temperature regimes in the zone of sporadic discontinuous permafrost, Tatra Mountains (Poland and Slovakia). Permafrost and Periglacial Processes 19: 315321. DOI: 10.1002/ppp.623 Haeberli W. 1973. Die Basis-Temperature der winterlichen Schneedecke als mglicher Indikator fr die Verbreitung von Permafrost in den Alpen. Zeitschriftfr Gletcherkunde und Glazialgeologie 9: 221227. Haeberli W. 1985. Creep of mountains permafrost: internal structure and ow of alpine rock glaciers. Mitt. Der Versuch. F. Wasserbau, Hydrologie und. Glaziologie: 77, 142. Haeberli W, Patzelt G. 1982. Permafrostkartierung im Gebiet der HochebenkarBlockgletscher, Obergurgl, tztaler Alpen. Zeitschriftfr Gletcherkunde und Glazialgeologie 18, 2:127150. Hanson S, Hoelzle M. 2004. The thermal regime of the active layer at the Murtl rock glacier based on data from 2002. Permafrost and Periglacial Processes 15: 273282. DOI: 10.1002/ppp.499 Hoelzle M. 1996. Mapping and modelling permafrost distribution in the Alps. Norsk Geogrask Tidsskrift 50: 1115. DOI: 10. 1080/00291959608552347 Hoelzle M, Wegmann M, Krummenacher B. 1999. Miniature Temperature Dataloggers for mapping and monitoring of permafrost in high mountain areas: First experience from the Swiss Alps. Permafrost and Periglacial Processes 10: 113124. DOI: 10.1002/(SICI)1099-1530(199904/06)10: 2<113::AID-PPP317>3.0.CO;2-A Ichim I. 1978. Preliminary observations on the rock glacier phenomenon in the Romanian Carpathians. Revue Roumaine de Gologie, Gophysique, Gographie, serie Gographie 23(2): 295299. Ikeda A, Matsuoka N. 2002. Degradation of talus-derived rock glaciers in the Upper Engadin, Swiss Alps. Permafrost and Periglacial Processes 13: 145161. DOI: 10. 1002/ppp.413 Ikeda A, Matsuoka N. 2006. Pebbly versus bouldery rock glaciers: Morphology, structure and processes. Geomorphology 73: 279296. DOI: 10.1016/j.geomorph.2005.07.015 Ishikawa M. 2003. Thermal regimes at the snow-ground interface and their implication for permafrost investigation. Geomorphology 52: 105120. DOI: 10.1016/S0169555X(02)00251-9 Ishikawa M, Hirakawa K. 2000. Mountain permafrost distribution based on BTS measurements and DC resistivity soundings in the Daisetsu Mountains, Hokkaido, Japan. Permafrost and Periglacial Processes 11: 109123. DOI: 10.1002/1099-1530(200004/ 06)11:2<109::AID-PPP343>3.0.CO;2-O Julin A, Chueca J. 2007. Permafrost distribution from BTS Measurements (Sierra de Telera, Central Pyrenees, Spain): Assessing the importance of Solar Radiation in a Mid-elevation Shaded Mountainous Area. Permafrost and Periglacial Processes 18: 137149. DOI: 10.1002/ppp.576 Kern Z, Balogh D, Nagy B. 2004. Investigations for the actual elevation of the mountain permafrost zone on postglacial landforms in the head of Lpunicu Mare Valley, and the history of deglaciation of Ana Lake Judele Peak region, Retezat Mountains, Romania. AnaleleUniversitii de Vest din Timioara, Seria Geograe 14: 119132. Kneisel C, Hauck C, Fortier R, Moorman M. 2008. Advances in Geophysical Methods for Permafrost Investigations. Permafrost and Periglacial Processes 19: 157178. DOI: 10.1002/ppp.616 Konrad SK, Humphrey NF, Steig EJ, Clark DH, Potter N, Pfeffer WT. 1999. Rock glacier dynamics and paleoclimatic implications. Geology 27(12): 11311134. DOI: 10.1130/0091-7613(1999)027<1131: RGDAPI>2.3.CO;2 Lewkowicz AG, Ednie M. 2004. Probability mapping of mountain permafrost using the BTS method, Wolf Creek, Yukon Territory, Canada. Permafrost and Periglacial Processes 15: 6780. DOI: 10.1002/ppp.480 Locke MH. 1999. Electrical imaging survey for environmental and engineering studies a practical guide to 2-D and 3-D surveys. M. H. Locke, Penang, Malaysia. Mayewski PA, Rohling EE, Stager JC, Karln W, Maasch KA, Meeker LD, Meyerson EA, Gasse F, van Kreveld S, Holmgren K, LeeThorp J, Rosqvist G, Rack F, Staubwasser M, Schneider RR, Steig EJ. 2004. Holocene climate variability. Quaternary Research 62: 243255. DOI: 10.1016/j.yqres.2004. 07.001 Reuther AU, Urdea P, Geiger C, Ivy-Ochs S, Niller HP, Kubik PW, Heine K. 2007. Late Pleistocene glacial chronology of the Pietrele Valley, Southern Carpathians constrained by 10 Be exposure ages and pedological investigations. Quaternary International 164165: 151169. DOI: 10.1016/j.quaint.2006.10.011 Ribolini A, Fabre D. 2006. Permafrost existence in rock glaciers of the Argentera Massif, Maritime Alps, Italy. Permafrost and Periglacial Processes 17: 4963. DOI: 10.1002/ppp.548 Serrano E, San Jos JJ, Agudo C. 2006. Rock glacier dynamics in a marginal periglacial high mountain environment: Flow, movement (19912000) and structure of the Argualas rock glacier, the Pyrenees. Geomorphology 74: 285296. DOI: 10.1002/j. geomorph.2005.08.014 Urdea P. 1992. Rock glaciers and periglacial phenomena in the Southern Carpathians. Permafrost and Periglacial Processes 3: 267273. DOI: 10.1002/ppp.3430030317 Urdea P. 1993. Permafrost and periglacial forms in the Romanian Carpathians. In Proceedings of the Sixth International Conference on Permafrost, Wushan G (eds). South China University of Technology Press: Beijing; 1: 631637. Urdea P. 2000. Munii Retezat. Studiu geomorfologic. Editura Academiei: Bucureti. Urdea P, Ardelean F, Onaca A, Ardelean M, Trok-Oance M. 2008a. Application of DC resistivity tomography in the alpine area of Southern Carpathians (Romania). In Ninth International Conference on Permafrost, Kane DL, Hinkel K (eds). Institute of Northern Engineering, University of Alaska: Fairbanks; 323333. Urdea P, Mihai B, Popa I, VespremeanuStroe A, Trok-Oance M, Ardelean A, Onaca A, Ttui F. 2008b. Noi metode de studio aplicate la zona alpin a Carpailor Romneti. Programul Cercetare de Excelen 20052008 MENER, 47 September, Sinaia. Universitatea Politehnic Bucureti: 383395. Vespremeanu-Stroe A, Mihai B, Cruceru N, Preoteasa L. 2004. The freeze-thaw cycles frequency in the Romanian Carpathians. Revue Roumaine de Gographie 48: 147155. Warhafting C, Cox A. 1959. Rock glaciers in the Alaska Range. Bulletin Geological Society of America 70: 383436.

Copyright 2012 John Wiley & Sons, Ltd.

Permafrost and Periglac. Process., (2012)

Você também pode gostar