Лёд и Снег · 2020 · Т. 60 · № 3

УДК 551.324.63

doi: 10.31857/S2076673420030052

Change of Chalaati Glacier (Georgian Caucasus) since the Little Ice Age based

on dendrochronological and Beryllium-10 data

L.G. Tielidze^1-3, O.N. Solomina^4,5, V. Jomelli^6,7, E.A. Dolgova⁴, I.S. Bushueva⁴,

V.N. Mikhalenko⁴, R. Brauche⁶, ASTER Team^6,8

¹Vakhushti Bagrationi Institute of Geography, Ivane Javakhishvili Tbilisi State University, Tbilisi, Georgia; ²Antarctic Research Centre,

Victoria University of Wellington, Wellington, New Zealand; ³School of Geography, Environment and Earth Sciences, Victoria

University of Wellington, Wellington, New Zealand; ⁴Institute of Geography, Russian Academy of Sciences, Moscow, Russia;

⁵Higher School of Economic, Moscow, Russia; ⁶CNRS, UMR 34 Aix-Marseille University, CNRS-IRD-Coll, France;

⁷CNRS, LGP, Sorbonne University, Paris, France; ⁸A comprehensive list of consortium members is given at the end of the paper

^*tielidzelevan@gmail.com

Изменения ледника Чалаати (Грузинский Кавказ) с малого ледникового периода

по данным космогенных изотопов (¹⁰Be) и дендрохронологии

В.Н. Михаленко⁴, Р. Брошэ⁶, Команда АСТЕР^6,8

¹Институт географии им. Вахушти Багратиони, Тбилисский государственный университет им. Иване Джавахишвили, Тбилиси,

Грузия; ²Антарктический исследовательский центр, Университет Виктории, Веллингтон, Новая Зеландия; ³Школа географии

окружающей среды и наук о Земле, Университет Виктории, Веллингтон, Новая Зеландия; ⁴Институт географии РАН, Москва, Россия;

⁵Высшая школа экономики, Москва, Россия; ⁶CNRS, UMR 34 Экс-Марсельский университет - CNRS-IRD-Coll, Франция;

⁷CNRS, LGP, Университет Сорбонны, Париж, Франция; ⁸Полный список членов консорциума приведён в конце статьи

^*tielidzelevan@gmail.com

Received January 21, 2020 / Revised March 9, 2020 / Accepted June 7, 2020

Keywords: Cosmic Ray Exposure dating, dendrochronology, glacier variation, Greater Caucasus, Little Ice Age.

Summary

Glacier variations over the past centuries are still poorly documented on the southern slope of the Greater

Caucasus. In this paper, the change of Chalaati Glacier in the Georgian Caucasus from its maximum extent

during the Little Ice Age has been studied. For the first time in the history of glaciological studies of the Geor-

gian Caucasus, ¹⁰Be in situ Cosmic Ray Exposure (CRE) dating was applied. The age of moraines was deter-

mined by tree-ring analysis. Lichenometry was also used as a supplementary tool to determine the relative

ages of glacial landforms. In addition, the large-scale topographical maps (1887, 1960) were used along with

the satellite imagery - Corona, Landsat 5 TM, and Sentinel 2B. Repeated photographs were used to identify

the glacier extent in the late XIX and early XX centuries. ¹⁰Be CRE ages from the oldest lateral moraine of the

Chalaati Glacier suggest that the onset of the Little Ice Age occurred ~0.73±0.04 kyr ago (CE ~1250-1330),

while the dendrochronology and lichenometry measurements show that the Chalaati Glacier reached its sec-

ondary maximum extent again about CE ~1810. From that time through 2018 the glacier area decreased

from 14.9±1.5 km² to 9.9±0.5 km² (33.8±7.4% or ~0.16% yr^-1), while its length retreated by ~2280 m. The

retreat rate was uneven: it peaked between 1940 and 1971 (~22.9 m yr^-1), while the rate was slowest in 1910-

1930 (~4.0 m yr^-1). The terminus elevation rose from ~1620 m to ~1980 m above sea level in ~1810-2018.

Citation: Tielidze L.G., Solomina O.N., Jomelli V., Dolgova E.A., Bushueva I.S., Mikhalenko V.N., Braucher R., ASTER Team. Change of Chalaati Glacier

(Georgian Caucasus) since the Little Ice Age based on dendrochronological and Beryllium-10 data. Led i Sneg. Ice and Snow. 2020. 60 (3): 453-470.

doi: 10.31857/S2076673420030052.

Поступила 21 января 2020 г. / После доработки 9 марта 2020 г. / Принята к печати 7 июня 2020 г.

Ключевые слова: Большой Кавказ, дендрохронология, колебания ледников, космогенные датировки, малый ледниковый период.

Для реконструкции колебаний ледника Чалаати в Грузии использовались космические снимки,

старые карты, повторные фотографии, дендрохронология, лихенометрия и анализ космогенных

изотопов. Максимальное наступание ледника в начале малого ледникового периода произошло

в ~1250-1330 гг., второй максимум, когда ледник достиг почти такой же длины, датируется при-

мерно 1810 г. С этого времени до 2018 г. площадь ледника уменьшилась с 14,9±1,5 до 9,9±0,5 км²

(33,8±7,4%, или ~0,16% год^-1), а его длина сократилась на ~2280 м.

 453 

Палеогляциология

1. Introduction

one of the most frequently visited glaciers in Georgia

due its close location to the road, gentle valley incli

Francois E. Matthes [1] was the first scien

nation and low position of the terminus. Today this

tist to use the term Little Ice Age, which is global

valley is used for the construction of the hydropower.

ly known as one of the coldest periods of the last

Water and debris from Chalaati Glacier will be fill

millennia. The Little Ice Age is a documented cool

ing this reservoir, the rate of which is highly depends

ing event that began around the XIII or XIV centu

on the future state of Glacier, making this study es

ry and peaked between the mid-XVI and mid-XIX

pecially relevant; 4) in addition, Chalaati valley is

centuries [2]. The spatial extent of this cold condi

the best location on the southern slope of the Great

tion was mostly observed in the Northern Hemi

er Caucasus where the well-preserved Little Ice Age

sphere, although the cooling phase was also noted in

moraines are covered by old conifer forest and, which

other regions with different intensities and time peri

are useful for dendrochronological purposes. The

ods. Recent studies revealed strong spatio-temporal

goal of our paper is to: 1) reconstruct the spatial vari

variations in this cooling with synchronous or asyn

ations of Chalaati Glacier; 2) assess the age of glacier

chronous cool decades between regions suggesting a

deposits of the most recent centuries; 3) estimate the

complex pattern of change [3]. Quantifying glacier

length and area changes; and 4) compare the varia

change during the Little Ice Age is important when

tions of this glacier in the Little Ice Age with those

attempting to understand regional climatic chang

of similar type and size from northern side of the

es and can also help to improve the understanding

Greater Caucasus and other mountain regions (e.g.

of any predictions of future glacial changes. Glacier

European Alps). This will increase the knowledge of

variations themselves can be also used for the model

the more recent portion of the Holocene glaciation

ling studies [4].

from the Greater Caucasus.

Detailed records of variations spanning the past

few centuries for glaciers on the northern slope of

Greater Caucasus in the past few centuries were re

2. Study area

cently published [5-8]. These records were based

on the analyses of satellite imagery and maps, old

The Greater Caucasus is one of the world’s high

historical materials, lichenometry, tree-rings and

est mountain systems, containing over 2000 gla

¹⁴C dating. While some records of past glacier varia

ciers with a total area of about 1200 km². The Geor

tions [9] are available for the glaciers located on the

gian side of the Greater Caucasus contains about

southern slope of Greater Caucasus [10-12], the sta

700 glaciers with a total area of about 370 km² [11].

tus of Georgian glaciers during the Little Ice Age

These glaciers play a significant role in the ecology

period is so far more poorly documented [13] com

and economy of Georgia. They provide a freshwa

pared to their northern counterparts. For this pur

ter source that feeds rivers in the mountainous area.

pose, Chalaati Glacier was selected on the southern

This water is essential to river ecology, particularly

slope of the Greater Caucasus in order to assess its

during the summer months when rivers have lower

variations since the Little Ice Age. We use the same

flows. Georgian glaciers also have economic impor

approach as for the northern slope such as combining

tance as a major tourist attraction with thousands of

the information from the old maps and figures, aeri

visitors each year. Local economies and livelihoods

al images, historical data, lichenometry and dendro

are connected to glacial input in these ways. Glacier

chronology [7] along with ¹⁰Be in situ CRE dating.

meltwater also supplies several hydroelectric power

We selected the Chalaati Glacier (43°07′ N,

stations. Understanding how glaciers in this region

42°42′ E) for several reasons: 1) Chalaati is an ex

are changing is therefore important for these local

ample of a well-documented glacier with a wealth of

considerations [10-13].

different historical sources (e.g. maps, photographs,

Chalaati is a compound-valley glacier and con

anecdotal evidences) [13] that allow reconstruction

sists of two flows, which are fed from the slopes of

of glacier length variations over the most recent cen

over 4000 meter-high peaks: Ushba, Chatini, Kavka

turies; 2) it is a typical valley glacier of relatively sim

si, and Bzhedukhi. Among the glaciers on the south

ple shape. This type of glacier is also climate sensitive

ern slope of the Greater Caucasus, this glacier has the

and often used for climate modelling; 3) Chalaati is

lowest terminus (1980 m a.s.l. in 2018) and intrudes

 454 

L.G. Tielidze et al.

into the forest zone. Three icefalls on the glacier indi

times too obscure to be used to accurately identify

cate a ledge under the glacier. The height of the larg

the position of the front. For example, according to

est icefall is ~300 m and its width is ~700 m. The two

Freshfield [14] Chalaati Glacier terminated at 5200

lower icefalls are relatively small. In the vicinity of the

feet (1585 m a.s.l.), and according the Déchy (1905),

icefalls, the glacier tongue is rugged by the various

at 1628 m a.s.l. respectively, i.e. below the confluence

fractures (serracs) going in different directions. The

to the Lekhziri River, ouside the Chaalaati valley.

edges of the glacier tongue are covered with debris of

This contradict to the image of Chalaati Glacier at

a variable thickness. It is likely that these two flows

the map of 1887 and other evidences (see discussion

will split in the near future. The middle part of the

below). This very detailed map, at a scale of 1:84 000,

glacier tongue is strongly inclined and cracked.

was used later by Podozerskiy [21] in his catalog of

The lateral moraines of Chalaati Glacier are well

the Caucasus glaciers but unfortunately, he did not

preserved. At their distal sides they are covered by

provide any indication of Chalaati frontal position.

forest, their proximal sides are bare and steep. The

Detailed analysis of this map shows some defects in

bottom of the valley stabilized in the XX century,

the shape of the glacier (see Chapters 4.4 and 5.4).

and is now covered by young birch forest. Below

Rutkovskaya [16] studied the dynamics, ice

~1750 m a.s.l. there are older moraine walls in coni

thickness and velocity of the Chalaati Glacier during

fer forest.

the 2^nd International Polar Year (Fig. 1). She found

that the glacier terminated at 1738 m a.s.l. and its

total area was 11.3 km² in 1933. In the central part of

3. Previous studies

the glacier tongue, the ice thickness was about 50 m

and the daily ice velocity was 15.3 cm during the

A number of scholars visited and described Cha

30 days period in August 1933. According to Tserete

laati Glacier and documented the position of its front

li, [17] Chalaati Glacier retreated over 1000 m from

in the XIX and XX centuries [14-20]. Freshfield [14]

the 1890s through the 1960s [17], and in 1959 ter

was one of the first scientists to visit the glacier in

minated at 1850 m a.s.l. and the snowline was locat

1868, provided the following description: «Chalaati

ed at 3050 m a.s.l. [18]. The mass balance observa

Glacier drains a double basin on either flank of Cha-

tions at Chalaati Glacier measured in summer 1959

tini. Owing to the steepness of the general inclination

at 1980 m a.s.l. found an ablation rate of ~0.7 cm/

of its bed, it attains to a lower point in the valley than

PDD (positive degree days) in July. Between July 31

the larger stream of the Lekhziri. It reaches, indeed,

and September 1 in 1961, daily ice velocity at Chalaa

a lower point than any other ice-stream on either side

ti Glacier was 18.2 cm [19].

of the Caucasus. Not many years ago the two glaciers

R. Gobejishvili [20] was leading glacial-geomor

met at their extremity, now the Lekhziri terminates at

phological studies between 1968 and 1995. The Cha

5600 feet (1706 m a.s.l.), and the Chalaati at 5200 feet

laati Glacier terminus was surveyed by the photo-

(1585 m a.s.l.). The descent from the Chatini Pass lies

theodolite method, and large-scale map (1:2 000)

over the northern névé of the Chalaati Glacier».

was created. According to his data the glacier ter

Déchy [15] also described Chalaati Glacier a few

minated at 1880 m a.s.l. in 1980. Later these stud

years later: «Due to the steepness of its bed, the Chalaa-

ies have continued by Tielidze [13]. The ablation

ti Glacier terminates lower into the valley than a larg-

rate was measured as ~0.6 cm/PDD (positive de

er stream of the Lekhziri Glacier. In fact, at 1628 m, it

gree days) in July and ~0.4 cm/PDD in August 2011

reaches the lowest point of all the glaciers of the Greater

at 2040 m a.s.l. Glacier terminated at 1950 m a.s.l. in

Caucasus on both sides, while the Lekhziri Glacier ter-

2011 (Table 1).

minates in 1734 m. Not many years back, the tongues

of the two glaciers must have touched each other. The

rocks of the moraines, which the Chalaati Glacier

4. Data and methods

brings down from its surroundings, are gneiss granites,

syenite, with fine grains consisting of feldspar and am-

4.1. In situ ¹⁰Be cosmic ray exposure dating. Three

phibolite, and pegmatites with quartz inclusions».

samples were collected with a hammer and chisel in

Unfortunately, the measurements provided by

September 2018 from the horizontal to sub-horizon

scholars in XIX and early XX centuries are some

tal uppermost surfaces of large boulders (> 60 cm

 455 

Палеогляциология

Fig. 1. Chalatai Glacier terminus position in 1933 [16] (modified by authors)

Рис. 1. Конец ледника Чалаати в 1933 г. [16] (с изменениями авторов статьи)

(250-750 µm); magnetic components were eliminat

Table 1. Front position of Chalaati Glacier according to vari-

ous sources

ed using a Frantz magnetic separator. Quartz was ex

Таблица 1. Положение конца ледника Чалаати по разным tracted by dissolution of undesirable minerals in HF/

данным

HNO₃ acid mixture and atmospheric ¹⁰Be removed

Years

Elevation, m a.s.l.

Reference

through sequential dissolution (~10% in mass) in di

1868

1585

Freshfield, 1896

luted HF. The obtained pure quartz was spiked with

1905

1628

Déchy, 1905

a commercial standard solution from the Scharlau

1911

Unknown

Podozerskiy, 1911

Company (1000 mg l^-1 of BeO) and then digested in

1933

1738

Rutkovskaya, 1936

48% hydrofluoric acid. Beryllium was extracted using

1959

1850

Tsereteli et al., 1962

anion and cation columns and alkaline precipita

1980

1880

Gobejishvili, 1995

tion. The obtained beryllium hydroxides were dried,

2011

1950

Tielidze, 2017

and finally oxidized for one hour at 800 °C. The final

BeO oxides were combined with niobium powder

high) located on the most external moraine on the

for AMS measurements at the French 5 MV AMS

left slope of the valley. The boulders were stable and

national facility ASTER (Aix-en-Provence) [22].

without evidence of loss due to denudation process

Measurements were calibrated against the in-house

es. Sample locations and elevations were recorded

standard STD-11 with an assigned ¹⁰Be/⁹Be ratio

using a handheld GPS, and topographic shielding

of 1.191±0.013 × 10^-11 [23] using a ¹⁰Be half-life of

was measured using a clinometer and a compass. The

1.387±0.0012 × 10⁶ years [24]. We calculated ¹⁰Be

samples were processed at CALM lab (Cosmonu

ages using the CREP online calculator [25] and using

cléides Au Laboratoire de Meudon - France) for

the Arctic production rate [26] and the Lal/Stone

in situ CRE dating. In situ-produced ¹⁰Be was mea

time corrected scaling scheme [27, 28]. We did not

sured in the quartz mineral fraction separated from

make corrections for snow cover and denudation of

the rock samples. Samples were crushed and sieved

the boulder surface.

 456 

L.G. Tielidze et al.

Table 2. List of the maps and satellite images scenes used in this study

Таблица 2. Список карт и снимков со спутников, использованных в настоящем исследовании

Date

Resolution, m

Type of imagery/map

Scene ID

1887

1:84 000 topographic map

X_13_4

1960

1:50 000 topographic map

k_38_26_v

20.09.1971

Corona

DS1115-2154DF070_d

06.08.1986

LT51710301986218XXX02

Landsat 5 TM

12.08.2000

LT51710302000225AAA02

22.09.2018

Sentinel 2B

L1C_T37TGH_A008069_20180922T080212

17.11.2011

ASTER GDEM

ASTGTM2_N42E041

4.2. Dendrochronology. For moraine dating the

Glacier dates back to the late 19th century. The mo

standard procedures of measuring, cross-dating and

raine deposits outside this contour clearly show that

indexation routinely used in tree-ring analysis [29]

a set of advances occurred before the Chalaati Gla

were applied. We cored the trees at the moraines and

cier reached this position. The lichenometry can be

glacier forefields either at the breast height (1.5 m) or

used to roughly assess the age of these deposits [31].

at the lowest possible level near the soil surface. In the

Although this dating tool was recently criticized by

first case, in order to assess the real age of the tree, the

Osborn et al. [32], lichenometry is still in use in geo

number of rings in the core was corrected by the ad

morphic studies as an express method of a relative

dition of 15 years and samples cored near the surface

dating of landforms (for further discussion see [7]).

this correction was 5 years. When the pith in the core

Here we used this method only for the relative

was missing we used CooRecorder software which

dating and an approximate assessment of the age of

has a tool to measure «distance to pith». The number

moraines [33]. We measured the maximum diame

of missing rings was assessed using curvature of the

ters of Rhizocarpon geographicum senso lato lichens

last five rings. The trees growing on the moraines pro

at the surfaces, when there were enough large boul

vide minimum age of the surface stabilization. Over

ders at these surfaces and considered the largest (old

all, we collected twelve tree ring samples within the

est) specimen as an indicator of the age of moraines.

two days (27-28 September 2018). Three long living

This method was criticized by Naveau et al. [34] and

species (Scots pine (Pinus sylvestris L.), Nordmann fir

Jomelli et al. [33] as statistically incorrect and the

(Abies nordmanniana (Steven) Spach.), spruce (Picea

GEV approach was suggested instead as a more sta

orientalis) are growing at the Little Ice Age moraines

tistically robust. Unfortunately we do not have origi

of the Greater Caucacus glaciers and they can poten

nal data from the growth curve constructed by Sere

tially provide ages of the surfaces up to five hundred

bryanny et al. [9] for the southern slope of Greater

years old. The eccesis (time lag for colonization) for

Caucasus where single maximum diameters of li

pine used for the dating at the northern slope of the

chens for five moraines from 14-15 years old up to

Greater Caucasus is 10-20 years as estimated from

3360±90 years BP (TA-1233) in Khalde valley have

the analyses of the aerial photos and direct survey

been used for this purpose (see Discussion for further

(e.g. [30]). The colonization by fir and spruce takes a

details). For this reason, we use the model suggested

few years longer. Thus, a minimum of 20 years should

by Beschel [31] to obtain a rough age estimate.

be added to the tree ring date to assess the minimum

4.4 Maps and Satellite Imagery. Old topograph

age of the surface (moraine) stabilization.

ic map (1:84 000) from the second Caucasus topo

In Chalaati valley we used the oldest conifer trees

graphic survey (1887) was used to evaluate the Cha

growing on the moraines or between the ridges. In a

laati Glacier outline (Table 2) along with the replaced

few cases when the sampling (coring) was impossible

and co-registered topographic map (1:50 000 scale)

due to the weak rotten wood we roughly counted the

from the 1960s [11]. The projection of the 1887 map

number of annual rings that would give us a rough

is different from the modern UTM system, as it was

estimate of the minimum age of the surface.

created by plane table survey using the old Russian

4.3. Lichenometry. Like at most other glaciers in

unit of length such as the Verst (1 Verst = 1.07 km).

the Greater Caucasus, the earliest map of Chalaati

In addition, the relief of this map is distorted (in

 457 

Палеогляциология

cluding the glacier shapes) compared to the map

for the mapping. We also crosschecked our mapped

from the 1960s and satellite imagery (see 5.4. chapter

glacial extents with high-resolution data from Digital

for more details).

Globe and Google Earth images. Glacier areas were

Corona image, dating from 1971, was obtained

also measured from the maps and all satellite imag

from Earthexplorer website (http://earthexplorer.

ery by manual digitization. The time series of maps

usgs.gov/). We co-registered the 1887 map and Co

and satellite imagery resulted in glacier area values for

rona image using the 12 August 2000 Landsat 5 TM

each corresponding date since the 1880s (1887). The

(Thematic Mapper) image as a master [10]. Offsets

glacier boundary in the accumulation area for earlier

between the images and the Corona/archival map

glaciers than 1880s was taken from the 1986 ice divide

were within one pixel (30 m) based on an analysis of

and was kept constant over time. Based on the out

common features identifiable in each dataset. We re

lines of the different survey years, corresponding areas

projected Corona image and 1887 map to Universal

and area changes were calculated.

Transverse Mercator (UTM), zones 38-north on the

The length of glaciers was determined accord

WGS84 ellipsoid, to facilitate comparison with mod

ing to Global Land Ice Measurements from Space

ern image datasets (ArcGIS 10.2.1). Together with

(GLIMS) recommendations (www.glims.com). The

Landsat imagery, these older topographic maps and

longest glacier's flow was determined manually as

Corona image enabled us to identify century-long

perpendicular to the altitude contours. Front vari

glacier change.

ation measurements were conducted by using the

Two Landsat 5 TM georeferenced images dated

glacier outlines for each date, along the ice front -

6 August 1986 and 12 August 2000, were supplied by

perpendicular to the flow. We have determined gla

the US Geological Survey’s Earth Resources Obser

cier area uncertainty with a buffer method similar to

vation and Science (EROS) Center and downloaded

Granshaw and Fountain [35], and adopted by Tielid

using the EarthExplorer tool (http://earthexplorer.

ze [10], Tielidze and Wheate [11]. This generated an

usgs.gov/). We used a color-composite scene for both

uncertainty of the mapped glacier area of ±5.6% for

images - bands 5 (short-wave infrared), 4 (near in

1960, ±6.3% for 1971 and 1986, and ±6.7% for 2000

frared), and 3 (red).

(buffer size ±30 m), while the uncertainty was ±5.1%

High-resolution satellite instruments such as

for 2018 (buffer size ±20 m).

Sentinel 2B, with 20 m horizontal resolution avail

Due to the different projection and distorted gla

able since March 2017, help in recognizing glacier

cier shape, the map from 1887 was co-registered and

margins. Cloud free image from 22 September 2018

digitized a second time by a different operator similar

was downloaded using the REMOTE PIXEL tool

to Tielidze [10] in order to accurately access the un

(http://remotepixel.ca). For Sentinel image, we used

certainty of the old glacier from 1887. To determine

pan-sharpened tools in ArcGIS 10.2.1 software to

the precision of the digitizing, we used the Nor

enhance pixel size (20 m) to 10 m. Advanced Space

malised Standard Deviation (NSD - based on delin

borne Thermal Emission and Reflection Radiometer

eations by multiple digitalization divided by the mean

(ASTER) Global Digital Elevation Model (GDEM,

glacier area for all outlines). The difference between

30 m) version 2 (http://asterweb.jpl.nasa.gov/gdem.

these two manually mapped outlines was ±10.3%.

asp) was used to assess the Chalaati Valley longitudi

For glacier outlines earlier than 1880s and outlines

nal profile. Other datasets used in this study include

for 1910-1940s we used a buffer size of ±60 m that

the GPS measurements (terminus position) from a

generated an average area uncertainty of ±9.7% for

2018 field survey.

1810, ±10.2% for 1840, ±10.2% for 1910, ±10.4% for

4.5. Glacier mapping and uncertainty assessment.

1930 and 1940.

The Little Ice Age glacial extent was digitized manu

ally, proceeding from the 1986 glacial extent, based

on clear visible morphological evidence, e.g. terminal

5. Results and Discussion

and lateral moraine systems. To avoid an overestima

tion of the Little Ice Age extent, larger rock outcrops

5.1. ¹⁰Be CRE ages. We collected all samples at

(as included in the scenes from 1986) were preserved.

the left side of the Chalaati valley while the right side

Due to its higher resolution compared to Landsat 5

was not accessible with no bridge over the river. We

(30 m), data from Sentinel-2 (10 m) were also used

report individual in situ ¹⁰Be CRE ages with their as

 458 

L.G. Tielidze et al.

Table 3. Sample details. Cosmogenic nuclide concentrations and calculated cosmic ray exposure ages from the valleys. ¹⁰Be/⁹Be

ratios were corrected for a process blank value of 9.11 × 10^-15 (¹⁰Be/⁹Be uncertainty: 8%) (samples Geo 1,3) and 5.31 × 10^-15

(¹⁰Be/⁹Be uncertainty: 15%) (sample Geo 2)

Таблица 3. Сведения об образцах. Концентрация космогенных нуклидов и рассчитанное время воздействия космиче-

ских лучей. Соотношения ¹⁰Be/⁹Be были скорректированы для начальных точек процесса: 9,11 × 10^-15 (ошибка опре-

деления соотношения ¹⁰Be/⁹Be - 8%) (образцы Geo 1 и Geo 3) и 5,31 × 10^-15 (ошибка определения соотношения

¹⁰Be/⁹Be - 15%) (образец Geo 2)

¹⁰Be

Erosion

Sample

Latitude

Longitude

Elevation,

Shielding

Density,

Thickness,

Age,

1σ,

1σ without

conc,

uncert,

rate,

m a.s.l.

factor

g cm^-3

kyr

PR error

at g^-1

at, g^-1

mm yr^-1

Geo 1

43.114097

42.737215

1843

10848

1657

0.9801

2.75

0.67

0.1

Geo 2

43.113806

42.738317

1820

59942

9913

0.9612

2.75

4.2

0.65

Geo 3

43.113630

42.738794

1812

12429

1284

0.9801

2.75

0.79

0.08

sociated uncertainties that include the standard devi

0.67±0.1 kyr to 4.2±0.65 kyr. Geo 2 dated at 4.2 kyr

ations of both analytical (reported Table 3) and pro

was rejected and considered as an outlier (possibly

duction rate uncertainties.

due to inheritance) compared to the other samples

The three rock samples (Geo1-3) were col

and rejected from the analysis. Geo 1 and 3 yield a

lected on the external moraine ridge M8 to docu

mean ¹⁰Be age of 0.73±0.04 kyr and suggest a forma

ment the maximum extent of Chalaati Glacier dur

tion of moraine M8 at the onset of the Little Ice Age.

ing the last millennia in this upper part of the valley

Because of the very limited number of samples it is

(Fig. 2). These samples show ages that range from

still difficult to assess the age of this moraine. Here

Fig. 2. Moraine ridges located on the left slope of the Chalaati valley; Locations of lichen measurements (diameter in

mm) and dendrochronological sampling along with the in situ ¹⁰Be CRE ages (kyr). Sizes of lichens and ages of trees

by dendrochronology are given in Table 4. Google Earth image (21.10.2019) is used background

Рис. 2. Моренные гряды, расположенные на левом борту долины Чалаати; местоположения измеренных ли

шайников (диаметр в мм) и дендрохронологических образцов, а также возраст обломков по космогенным

нуклидам ¹⁰Be (тыс. лет). Размеры лишайников и возраст деревьев приведены в табл. 4. Изображение Google

Earth (21.10.2019 г.) использовано в качестве фона

 459 

Палеогляциология

Table 4. Minimum moraine age of the Chalaati Glacier based on dendrochronological measurement and maximum lichen

diameters at the moraines

Таблица 4. Минимальный возраст морен ледника Чалаати по дендрохронологическим данным и размер максималь-

ных лишайников на моренах

Name

Dendrochronological measurement

Eleva

Mo

(Number)

Latitude

Longitude

Lichens

the year of

last

presence

correction

tion,

raine

on the map

(maximum diameter, mm)

the first mea

ring,

of the

(year) for the

m a.s.l.

(see Fig. 2)

sured ring

year

core

missing pith

43.11533

42.72925

1829

40, 40, 40, 35, 30

М1

45, 45, 40, 40, 39, 39, 36, 35, 34,

43.11516

42.87320

1822

33, 32, 32, 31, 31

М2

43.11230

42.73430

1760

45, 40, 38, 35, 35

М3

43.11213

42.77353

1783

1965

2018

43.11209

42.73576

1752

1944

2018

М4

43.11209

42.73576

1752

1928

2018

43.11176

42.73611

1739

1939

2018

8-9

43.11355

42.73796

1812

1885

2018

10-11

43.11357

42.73771

1820

65, 60, 55 50, 47

1857

2018

12-13

43.11336

42.73876

1820

1879

2018

43.11338

42.73906

1792

1887

2018

55, 52, 50, 50, 49, 45, 42, 40, 40,

М7

15-16

43.11247

42.74066

1737

1865

2018

40, 39, 36, 35, 32, 32, 31, 22

43.11180

42.74117

1717

57, 52, 51, 50, 50, 42

54, 52, 51, 50, 50, 47, 46, 46, 45,

43.11112

42.74171

1700

45, 45, 43, 42, 41, 41, 40, 34, 35, 32

43.11012

42.74222

1670

60, 48, 45

43.06830

42.44266

1838

1825

2018

43.11346

42.73917

1800

1859

2018

М8

43.11332

42.73957

1790

53, 52, 50

23-24

43.11316

42.74040

1698

65, 62, 62, 62, 60

1868

2018

the mean ¹⁰Be age of 0.73±0.04 kyr has to be consid

2. The three moraine ridges - M2-M4 are locat

ered as a very preliminary estimate.

ed at an approximately 1730 m a.s.l., i.e. where the

5.2. Dendrochronological measurements. Fig. 2

termination of the glacier was recorded in 1933 by

shows the locations of dendrochronological sam

Rutkovskaya [16] (see Fig. 1).

pling sites and their position relative to moraine ridg

3. At an altitude of about 1710 m a.s.l., anoth

es. Minimum ages of moraines based on tree-ring

er, relatively older stage M5 is distinguished, which

data are presented in Table 4. The highest lateral mo

is also visible on the map of Rutkovskaya as a ridge

raines are clearly expressed in the Chalaati valley -

going down to the river. Between the two moraine

unsodded in its proximal part and covered with for

complexes M4 and M5, the distance is approximately

est - in the distal one. Their corresponding terminal

20 m vertically and 100 m horizontally.

moraine complex is more poorly preserved, but nev

4. The lower moraine ridge M6 is covered by forest

ertheless it is discernible on both sides of the valley.

and not very well shaped, especially in its lower part.

At the Fig. 2 we mapped the moraine ridges from the

5. Two older lateral moraines in this complex M7

M1 to M8 (from the youngest to the oldest) marking

and M8 stretching parallel to the slope of the valley.

individual stages of the advances or stationary stages

They are partially destroyed by debris flows and ava

of the glacier.

lanches coming from the left side of the valley. The

1. Moraine M1 at an altitude of about 1820 m a.s.l.

slope processes keep renewing the surface of these mo

marks a location of the glacier position in 1960s (In

raines, erode them and damage the vegetation growing

strumental measurement by Gobejishvili [20]).

on their surface. These tracks are clearly marked at the

 460 

L.G. Tielidze et al.

Google Earth image (see Fig. 2) by the lighter color of

vegetation in the lower parts of these debris-flows.

So, the best undisturbed portion of moraines

older than CE 1887 that we studied using bioindi

cation approaches is located between the two de

bris flows at the left side of the valley. The map of

1887 shows that the space left unoccupied by the ice

at that time was quite narrow. One moraine ridge

descending to the Lekhziri River is marked on this

map. In fact, there are at least two moraine ridges

outside the glacier limits indicated on the 1887 map,

which was not accurate enough to determine if it cor

responds to M7 or M8.

The first tree-ring sample that we got from the

moraine 3 that is located very close to the margins

Fig. 3. Lichen growth curve developed:

of the glacier marked at the map of Rutkovskaya for

1 - for the southern [9] (brown) and 2 - northern [7] (blue)

the year 1933 dates back to 1959. Taking into account

slopes of Greater Caucasus

the correction for the height of coring (approximate

Рис. 3. Кривые роста лишайников:

1 - для южного [9] (коричневый) и 2 - северного [7] (си

ly 5 years) and the time lag between the stabiliza

ний) склонов Большого Кавказа

tion of the surface and its colonization by forest (10-

15 years) the tree-ring minimum date of this surface is

early 1930s, i.e. closely fit to the observations. At the

5.3. Lichenometric measurement. Although, for

foot of the M4 moraine ridge, at its distal side, two

the reasons discussed above, we did not produce any

pines grow, which settled here in 1926 and in 1943.

systematic lichenometric studies in this valley, some

I.e. glacier terminated here in early 1900s and the mo

data on the size of lichens seem to us worthy of dis

raine M4 (1750 m a.s.l.) was deposited slightly later,

cussion in this work. As it was described previously

most likely in the 1910s (see Fig. 2, M4).

by Solomina et al. [7], currently there are serious rea

Unfortunately, the preservation of the moraine

sons to discard all control points constraining the li

ridges in the lower part of the valley is poor. They

chen’s growth in Caucasus that are older than two to

are flattened and lost their clear outline, so it is very

three hundred years old. In general, the rough esti

difficult to trace the contours of the glacier framed

mate of the rate of Rhizocarpon geographicum growth

by them. It is likely that these surfaces are also con

in the past 100-150 years at the northern slope of the

stantly affected by slope processes and are partial

Greater Caucasus is around 0.25-0.30 mm per year.

ly eroded. We could not find old trees on the sur

The past century period characterizes the stage of the

face between moraine M6 and terminal moraines M7

fast growth of the lichens, however in the earlier time

and M8. However, fragments of moraines M7 and

the growth rate of lichens should be slower [36], but

M8 that have survived the erosion, support that old

there are no reliable control points to constrain the

trees are useful to estimate the minimum age of these

older part of the growth curve (Fig. 3). In Fig. 3 we in

generations. The oldest tree (live pine), settled on

dicated control points for the moraines of the north

moraine M7 in CE 1854. Several other old trees grew

ern slope of Greater Caucasus [7] and four points that

on this moraine in CE 1862, 1878, 1883 and 1884.

Serebryanny et al. [9] identified for the southern slope

Thus, the minimum age of this moraine should be

of the Greater Caucasus. We discarded the oldest one

approximately 180 years, i.e. the moraine was formed

for the moraine of Khalde Glacier (ca 3500 years old)

in the CE ~1840. According to the dendrochrono

that almost certainly supports the secondary gener

logical data, moraine M8 is older, where the oldest

ation of lichens and therefore is too old to be used

tree (spruce) grew in CE 1822. Other pine dates are

for lichenometry. Moreover, it has been demonstrat

CE 1858 and 1867. However, as we learnt from the

ed that the mean of five largest lichens does not fit

¹⁰Be date of this moraine this minimum age estimate

with the statistical extreme value theory and thus the

is very far from the real one and tree rings in this case

mean value is not robust [33, 34]. Fig. 3 shows that the

strongly underestimate the real age of the surface.

growth rates of the Rhizocarpon geographicum at the

 461 

Палеогляциология

Fig. 4. Chalaati Glacier terminus change:

a - 1887 (1:84 000 scale topographic map); b - 1932 [16], 1960 (1:50 000 scale topographic map); c - 1971 (Corona, 20.09.1971);

d - 1986 (Landsat 5 TM, 06.08.1986); e - 2000 (Landsat 5 TM, 12.08.2000); f - 2018 (Sentinel 2B, 22.09.2018)

Рис. 4. Изменения конца ледника Чалаати:

a - 1887 г. (топографическая карта масштаба 1:84 000); b - 1932 г. [16], 1960 г. (топографическая карта масштаба

1:50 000); c - 1971 г. (Corona,, 20.09.1971 г.); d - 1986 г. (Landsat 5 TM, 06.08.1986 г.); е - 2000 г. (Landsat 5 TM,

12.08.2000 г.); f - 2018 г. (Sentinel 2B, 22.09.2018 г.)

southern slope is similar, but slightly higher than at the

the growth rate according to two independent studies is

northern one. This figure can be used for a rough esti

comparable. On the moraine M7, where it was possible

mate of the age of the moraines of the past one to two

to measure a sufficient number of lichens, their maxi

centuries that we are studying at Chalaati Glacier.

mum sizes were up to 65 mm (see Fig. 2). According to

On the moraines M2 and M3, which were depos

the extrapolated linear growth curve for the southern

ited about 90-100 years ago, judging by cartographic

slope, shown in Fig. 3, lichen reaches 65 mm in about

and dendrochronological data, the maximum size of

150 years. According to our dendrochronological data,

lichens is 45 mm (see Fig. 2). According to Serebryan

the minimum age of this surface is 180 years. Serebry

ny et al. [9], the lichens as large as 45 mm on the Khal

anny et al. [9] using the logarithmic curve claimed that

de Glacier moraines are also about 90 years old. Thus,

the moraines supporting the lichens of 61-63 mm on

 462 

L.G. Tielidze et al.

Fig. 5. Chalaati Glacier degradation between 1884 (a and c) (photos by V. Sella) and 2011 (b and d) (photos by L. Tielidze)

Рис. 5. Деградация ледника Чалаати между 1884 г. (a и c) (фото В. Селлы) и 2011 г. (b и d) (фото Л. Тиелидзе)

the southern slope dated back to the XVIII century in

lateral moraine ridge at the left side of the valley that

terval of CE 1770-1780 years. Due to the lack of re

can be seen at the map. Thus, most moraines located

liable data on the growth rate of lichens in this time

in the valley and covered by conifer forest by now were

range, no more accurate data have yet been obtained.

deposited later, after 1880s. The second attempt to co-

5.4. Old maps and repeated photos. The map of 1887

(Fig. 4, a) along with the moraine samples is the oldest

and satellite images (see Fig. 4, c-f) (see also chap

records that we can use for the reconstruction of glacier

ter 4.4), showed that the shape of the glacier terminus

fluctuations in XIX century. According to the map the

of the XIX century (1887) does not fit into the valley.

glacier was terminating very close (about 200 m away)

Moreover, is significantly higher than it is shown on

to the confluence of two rivers (Lekhziri and Chalaa

the map of 1887, namely in the region of the younger

ti). Its tongue occupied almost entirely the lower part

lateral moraine. This is also clearly visible in Fig. 5, b,

of the valley covering most of moraines except for one

the treeless right side of the valley. Thus, we discuss

 463 

Палеогляциология

Table 5. Chalaati Glacier area and terminus elevation since

Table 6. Chalaati Glacier area and length change since the

the ~1810 to the different years

~1810 according to the time periods between the dated posi-

Таблица 5. Площадь ледника Чалаати и высота его фрон-

tions of glacier terminus (the average error terms for length

та с ~1810 г. за разные годы

change are ±15 m)

Таблица 6. Изменение площади и длины ледника Чалаати

Area,

Area uncer

Terminus,

Year

Moraine

с ~1810 г. за периоды времени между датированными

km²

tainty, %

m a.s.l.

положениями его фронта (средние значения погрешно-

~1810

14.9±1.5

±9.7

1620

сти для изменения длины составляют ±15 м)

~1840

14.3±1.5

±10.2

1630

Time

Area

Area change,

Terminus

1887 (~1880s)

14.1±1.5

±10.3

1650±100

periods

change, %

~% yr^-1

change, m

change, m yr^-1

~1910

13.8±1.4

±10.3

1720

1810-1840

4.4±9.9

0.15

215

7.2

~1930

13.6±1.4

±10.4

1730

1840-1880

1.1±10.2

0.03

185

4.6

~1940

13.6±1.4

±10.4

1740

1880-1910

2.5±10.3

0.08

470

15.7

~1960

12.8±0.7

±5.6

1800

1910-1930

1.2±10.3

0.06

4.0

1971

12.3±0.8

±6.3

1860

1930-1940

0.2±10.4

0.02

110

11.0

1986

11.6±0.7

±6.3

1900

–

1940-1960

5.6±8.0

0.28

440

22.0

2000

11.0±0.7

±6.7

1920

1960-1971

3.9±6.0

0.35

270

24.5

2018

9.9±0.5

±5.1

1980

1971-1986

5.8±6.3

0.39

135

9.0

1986-2000

5.0±6.5

0.36

5.7

whether to take as the basis the assumption that 1) the

2000-2018

10.2±5.9

0.57

295

16.4

terminus of the glacier on the 1887 map is correct

1810-2018

33.8±7.4

0.16

2280

11.0

ly reflected (200 m away from the two rivers conflu

ence) but the shape of the glacier tongue is distorted,

or, 2) the glacier terminus was not 200, but 500 m away

ellite imagery, which allowed us to define glacier

from the two rivers confluence. We suppose that the

change much precisely over the last half century (see

topographers of the XIX century accurately mapped

more results in the 5.5. chapter).

terminus of the glacier relative to the two rivers conflu

5.5 Glacier change since the XIII century. ¹⁰Be

ence (i.e. 200 m away), while the shape of the tongue of

CRE ages from the surface of the oldest Moraine

the glacier could be distorted.

(M8) suggest that the age of the upper glacial unit and

Photographs of Vittorio Sella of 1884, taken at

of the corresponding glacial advance is at least CE

about the same time as the map of 1887, could shed

~1300. Neither lichen nor tree ring were able to pro

light on this issue, but unfortunately, we cannot see

vide a reasonably close age estimate of this surface.

the terminus of the glacier from these positions due

The trees were too young (up to 2 hundred years old)

to the bushes and heavy debris cover (see Fig. 5,

and we did not find any lichen larger than 65 mm on

a, c). However, if we suppose that the white spot

this surface (except the one single lichen 85 mm on

that is visible though the bushes is ice (see Fig. 5, a),

the M7). The reason for this is either the lack of suit

the glacier should have been descend at least up to

able material on the surface of the moraines, specific

the lowest moraine that can be identified in a 2011

conditions for lichens slow growth (shade) or partly

photo (see Fig. 5, b). This moraine is connected to

renovated surface of the moraines. If indeed the mo

the main crest of the lateral moraine and is located

raine is 600-700 years old as we identified by ¹⁰Be the

about 450 m away from the confluence of two riv

lichens should be almost 200 mm.

ers. However, we cannot rule out that the end of the

Since the ~1810 Chalaati Glacier area decreased

glacier in 1884 was located even below this position,

by 33.8±7.4% or ~0.16% yr^-1 from 14.9±1.5 km² to

as we cannot see it in the 1884 photos. Thus, the his

9.9±0.5 km² while its terminus elevation rose from

torical material from 1880s does not provide precise

~1620 m to ~1980 m during the same time (Table 5;

evidence where the Chalaati Glacier was exactly ter

Fig. 6). While area was decreasing, Chalaati Gla

minated at that time. However, as no other data exist

cier retreated steadily over the past two centuries.

from this time, this map is the only source for this re

The most intense retreats occurred in ~1880-1910,

search to establish century-long trend glacier change.

1940-1971, and 2000-2018, while the slowest re

Unlike the XIX century map, the large-scale top

treats have been recorded in ~1840-1880, 1910-

ographic map from 1960 is consistent with all sat

1930, and 1986-2000 (Table 6). According to the

 464 

L.G. Tielidze et al.

Fig. 6. Chalaati Glacier change (а) since the Little Ice Age (GeoEye 2014 image is used as background).

Different colors of outlines show glacier margins in different years. Moraine M8 corresponds to the CE 1810; M7 - 1840; M4 -

1910; M3 - 1930; M2 - 1940; M1 - 1960.; b - Chalaati valley longitudinal profile (based on ASTER GDEM, 2011) and Chalaati

Glacier terminus positions according to the different years; c - yellow dotted line shows Chalaati valley longitudinal profile.

Рис. 6. Сокращение ледника Чалаати (a) после малого ледникового периода (изображение GeoEye 2014 ис

пользовано в качестве фона).

Различные цвета контуров показывают края ледника в разные годы. Морена М8 соответствует CE 1810 г.; М7 - 1840 г.;

М4 - 1910 г.; М3 - 1930 г.; М2 - 1940 г.; М1 - 1960 г.; b - продольный профиль долины Чалаати (по дан

ным ASTER GDEM, 2011 г.) и положение конца ледника Чалаати в разные годы; c - жёлтая пунктирная линия показы

вает продольный профиль долины Чалаати

previous detailed field measurement of the termi

with retreat measurement by Tsereteli [17], accord

nus position of Chalaati, the retreat in 1990s was in

ing to which, the Chalaati Glacier has retreated over

terrupted by an advance. Microstadial moraines in

more than 1000 m from 1890s to 1960s.

front of Chalaati Glacier confirm ~20 m glacier ad

Unlike the other investigated periods, glacier area

vance during 1990-1993 [11] that gives a confidence

loss rates over the last three decades (since the 1986)

in our measurements (small retreat rates between

are much higher compared to the relative terminus

years 1986 to 2000). Moreover, Chalaati Glacier re

retreat rates (see Table 6), suggesting that the glacier

treat between 1880s and 1960s (~1100 m) is in line

area decrease over this time period progressing not

 465 

Палеогляциология

Fig. 7. Comparison of cumulative curves of terminus changes:

1 - for the Chalaati Glacier; 2 - for the Tsey Glacier in the northern Greater Caucasus [7]; 3 - Mer de Glace Glacier in the

French Alps [40]

Рис. 7. Сравнение кумулятивных кривых изменения длины ледников:

1 - ледник Чалаати; 2 - ледник Цей на северном макросклоне Большого Кавказа [7]; 3 - ледника Мер-де-Глас во Фран

цузских Альпах [40]

only the terminus expense but also upper bodies of

of the Greater Caucasus. e.g. according to Solomi

the glacier. This can be the result of the supra-glacial

na et al. [7] general glacier retreat on the northern

debris cover area and thickness increase near the ter

slope started in the late 1840s. Four to five minor re-

minus similar to some glaciers of the Greater Cau

advances occurred in the period between CE 1860s

casus (e.g. Djankuat Glacier, [37]) which became

and 1880s and three re-advances or steady states in

more pronounced after 2000 [12]. We note that this

1910s, 1920s and 1970s-1980s, which again are agree

requires more detail investigation and it can be the

ment with lowest retreat rates or steady states be

topic of the future study. Even though the century

tween ~1840-1880 (~4.63 m yr^-1) and 1910-1930

long trend of glacier retreat is global, and the rate of

(~4.0 m yr^-1) of the Chalaati Glacier. In addition,

this retreat has increased in the past few decades, the

these recorded length changes of Chalaati Glacier are

retreat trend between 1940 and 1971 is unusual for

quite similar to changes observed in similar size Tsey

Chalaati Glacier in the context of the past two centu

Glacier (northern counterpart) over the past two cen

ries and it requires additional research.

turies [7], while these are quite different to chang

5.6. Comparison with other glaciers. The ¹⁰Be

es observed in the Mer de Glace Glacier (France)

CRE ages are in line with global trends [2] of gla

with several advance phases since the beginning of the

cier advances in the second half of the Holocene and,

XIX century [40] (Fig. 7). However, to be more ro

in particular, during the Little Ice Age. The early

bust, all these comparisons at the regional scale of the

advance of Chalaati Glacier 600-700 years ago is

Georgian Caucasus need more data collection, obser

broadly consistent with minimum ¹⁴C age for a mo

vations, and enhanced chronology that had to be con

raine in the Bezengi Valley on the northern slope of

fronted to those obtained in other mountain ranges.

the Greater Caucasus (CE 1245-1428) [9]. Chalaati

advances are also in phase with other studies from the

European Alps, where the general advance is docu

6. Conclusions

mented in the late XIII century that culminated be

tween CE ~1350 and ~1385 at Great Aletsch, Gorner

We present the first Chalaati Glacier variations

(Switzerland) [38], and Mer de Glace (France) [39].

analysis including multitemporal data sets covering

The dendrochronological data of this study also

the time period since the Little Ice Age. In situ ¹⁰Be

fits to other investigation from the northern slope

CRE ages, dendrochronology, lichenometry, along

 466 

L.G. Tielidze et al.

with the manual digitization from 1887 and 1960s

ers, Dmitry Petrakov and anonymous reviewer 2, for

topographic maps and satellite imagery from 1971

detailed comments which clearly enhanced the qual

(Corona), 1986, 2000 (Landsat 5) and 2018 (Senti

ity of the paper. Special thanks to Lauren Vargo and

nel) were used to map the glacier surface area to sub

Regis Braucher for proofreading the paper.

stantially improve the regional existing knowledge.

The main study findings can be summarized as

Благодарности. Проект выполнен в рамках Гос

follows:

задания № 0148-2019-0004 и российско-фран

1) based on in situ ¹⁰Be CRE ages, the Chalaati

цузского (CNRS-PICS) совместными програм

Glacier reached its maximum extent in the past mil

мами (DECAU; IPAGCAUA) и программой

lennium probably at the onset of an early Little Ice

LIA DEGLAC. Исследование с грузинской сто

Age 0.73±0.04 kyr ago (CE ~1250-1330), even if

роны поддержано Национальным научным

more data are needed to get a better constrain on the

фондом Грузии им. Шота Руставели

age of this advance. This maximum extent recorded

(SRNSFG) [YS17_12]. Измерения ¹⁰Be прово

in this paper corresponds to moraine M8;

дились на национальном объекте ASTER AMS

2) according to minimum tree-ring dates, the

(CEREGE. Экс-ан-Прованс), который поддер

same oldest terminal moraine (M8) date back to the

живается INSU/CNRS, ANR в рамках програм

very beginning of the 19th century (~1810), while the

мы «Projets thématiques d'Excellence» для

second oldest moraine formed in the ~1840 (M7).

«Equipements d'excellence» ASTER - CEREGE

Moraines from the 20th century were dated as ~1910

и IRD. В состав команды ASTER входят Жорж

(M4) ~1930 (M3), ~1940 (M2), and ~1960 (M1);

Ауматр, Дидье Л. Бурлес, Карим Кеддадуш.

3) since the ~1810, Chalaati Glacier decreased

Мы благодарны двум рецензентам, Дмитрию

by 33.8±7.4% (~0.16% yr^-1) with highest decrease

Петракову и А.М. Грачеву, за подробные ком

rates in 1971-1986 (~0.39% yr^-1) and 2000-2018

ментарии, которые значительно улучшили ка

(~0.57% yr^-1), while the lowest rates in 1840-1880

чество статьи. Выражаем также благодарность

(~0.03% yr^-1) and in 1910-1940 (~0.04% yr^-1);

Лорен Варго и Реджису Браухеру за помощь в

4) over the last two centuries glacier terminus re

правке текста.

treated by about 2280 m with highest retreat rates

in 1940-1971 (~22.9 m yr^-1) and in 2000-2018

(~16.4 m yr^-1), while the lowest rates in ~1840-1880

Расширенный реферат

(~4.6 m yr^-1) and in 1986-2000 (~5.7 m yr^-1).

Future studies can be focused in collection of

Лихенометрические исследования морен ма

more data from other glacier valleys in order to bet

лого ледникового периода у ледников южного

ter define the Little Ice Age glacier extend in region

макросклона Большого Кавказа, выполненные

al context.

Н.А. Голодковской в 1980-х годах [9], факти

чески, единственный источник информации о

Acknowledgements. The project is supported by the

времени наступаний этих ледников в прошлом.

fundamental scientific research № 0148-2019-0004

В этой работе авторы приводят первые датиров

and Russian-French (CNRS-PICS) collaborative

ки морен ледника Чалаати в Сванетии, возраст

programs (DECAU; IPAGCAUA) and LIA program

которых определён с помощью космогенных

DEGLAC. The study from Georgian side was sup

изотопов бериллия ¹⁰Be. Кроме того, для опре

ported by Shota Rustaveli National Science Founda

деления времени и масштабов колебания лед

tion of Georgia (SRNSFG) [YS17_12]. The ¹⁰Be

ника Чалаати (Грузия) в прошлом были исполь

measurements were performed at the ASTER AMS

зованы космические снимки (Corona, Landsat 5

national facility (CEREGE. Aix-en-Provence) that is

TM и Sentinel 2B), старые карты (1887, 1933,

supported by the INSU/CNRS, the ANR through

1960 гг.), повторные фотографии, дендрохро

the «Projets thématiques d’excellence» program for

нология и лихенометрия. Исторические свиде

the «Equipements d’excellence» ASTER-CEREGE

тельства о положении конца ледника имеют

action and IRD. ASTER Team includes Georges

ся для 1868 г. [14] (1585 м), 1905 г. [15] (1628 м),

Aumaître, Didier L. Bourlès, Karim Keddadouche.

1933 г. [16] (1738 м), 1959 г. [18], 1980 г. [20]

We gratefully acknowledge the support of two review

(1880 м), 2011 г. [13] (1950 м).

 467 

Палеогляциология

Три пробы на ¹⁰Be были отобраны ручным

ны на горизонтальной поверхности дисталь

молотком и долотом в сентябре 2018 г. с гори

ной левой береговой морены М8 (см. рис. 2) на

зонтальных и субгоризонтальных верхних по

анализ ¹⁰Be, имеют возраст 4,2±0,65, 0,79±0,08

верхностей крупных валунов высотой более

и 0,67±0,1 тыс. лет. Возраст 4,2 тыс. л.н. мы

60 см, расположенных на самой внешней море

сочли аномальным (возможно, из-за наследо

не на левом борту долины. Валуны были устой

вания сигнала). Образцы Gео 1 и Gео 3 дают

чивыми и не имели признаков поверхностной

средний возраст ¹⁰Be в интервале 1250-1330 гг.

эрозии. Координаты и высоты образцов реги

и означают образование морены M8 в нача

стрировались с помощью портативного GPS, а

ле малого ледникового периода. Однако из-за

топографическую экранировку (линию горизон

очень ограниченного числа проб средний воз

та) определяли с помощью клинометра и компа

раст морены по ¹⁰Be должен рассматриваться

са. Образцы обрабатывали в лаборатории CALM

как предварительный. По своим масштабам это

lab (Cosmonucléides Au Laboratoire de Meudon-

наступание в малом ледниковом периоде было

France). Длина ледников определялась в соот

максимальным. Второй максимум, когда лед

ветствии с рекомендациями GLIMS (www.glims.

ник достиг почти такой же длины, датируется

com). Самый длинный поток ледника опреде

примерно 1810 г.

лялся вручную как перпендикуляр высотным

Таким образом, ледник Чалаати на южном

контурам. Измерения положения фронта ледни

склоне Кавказа наступал примерно в то же

ка проводили с использованием контуров лед

время, что и ледники в Альпах и других гор

ника для каждой даты, вдоль ледового фронта -

ных странах Северного полушария. С этого

перпендикулярно течению.

времени до 2018 г. площадь ледника уменьши

Мы оценили неопределённости вычисления

лась с 14,9±1,5 км² до 9,9±0,5 км² (33,8±7,4%,

площади ледника буферным методом. Ошиб

или ~0,16% год^-1), а его длина сократилась

ки определений площади ледника составили

на ~2280 м. Скорость отступания была не

±5,6% для 1960 г., ±6,3% для 1972 и 1986 гг. и

равномерной: она достигла максимума между

±6,7% для 2000 г. (размер буфера ±30 м), в то

1940 и 1971 г. (~22,5 м год^-1), самая медлен

время как неопределённость составила ±5,1%

ная скорость отступания была в 1910-30-х

для 2018 г. (размер буфера ±20 м). Три образ

годах (~4,0 м год^-1). Высота конца поднялась с

ца породы (см. табл. 3), которые были отобра

~1620 м до ~1980 м в период с 1810 по 2018 г.

References

tuations. Climate Dynamics. 2012, 38: 1065e1079.

http://dx.doi.org/10.1007/s00382-011-1145-7.

1. Matthes F.E. Report of the committee on glaciers, 1939-

5. Bushueva I.S. Kolebaniya lednikov na Tsentralnom i Za-

40. Transactions of the American Geophys. Union.

padnom Kavkaze po kartograficheskim, istoricheskim i

1940, 1: 396-406. http://glaciers research.pdx.edu/

bioindikatsionnym dannym za poslednie 200 let Fluctua

Report-Committee-Glaciers-1939-40.

tions of glaciers on the Central and Western Caucasus

2. Solomina O.N., Bradley R.S., Jomelli V., Geirsdottir A.,

using cartographical, historical and proxy data over the

Kaufman D.S., Koch J., McKay N.P., Masiokas M., Mill-

last 200 years). PhD Thesis. 2013. Moscow: Institute of

er G., Nesje A., Nicolussi K., Owen L.A., Putnam A.E., Wan-

Geography Russian Academy of Sciences, Russia [In

ner H., Wiles G., Yang B. Glacier fluctuations during the

Russian].

past 2000 years. Quaternary Science Reviews. 2016, 149:

6. Solomina O.N., Bushueva I.S., Kuderina T.M., Mats-

61-90. https://doi.org/10.1016/j.quascirev.2016.04.008.

kovsky V.V., Kudikov A.V. Holocene history of the Ul

3. Neukom R., Gergis J., Karoly D., Wanner H., Curran M.,

lukam Glacier. Ice and Snow. 2012, 1 (117): 85-94.

Elbert J., González-Rouco F., Linsley B.K., Moy A.D.,

https://doi.org/10.15356/2076-6734-2012-1-85-94.

Mundo I., Raible C.C., Steig E., van Ommen T.,

[In Russian].

Vance T., Villalba R., Zinke J., Frank D. Interhemi

7. Solomina O.N., Bushueva I., Dolgova E., Jomelli V.,

spheric temperature variability over the past millenni

Alexandrin M., Mikhalenko V., Matskovsky V. Gla

um. Nature climate change. 2014, 4: 362-367. https://

cier variations in the Northern Caucasus compared

doi.org/10.1038/nclimate2174.

to climatic reconstructions over the past millenni

4. Leclercq P.W., Oerlemans J. Global and hemispheric

um. Global Planetary Change. 2016, 140: 28-58. doi.

temperature reconstruction from glacier length fluc

org/10.1016/j.gloplacha.2016.02.008.

 468 

L.G. Tielidze et al.

8. Solomina O.N., Bushueva I.S., Polumieva P.D., Dolgo-

20. Gobejishvili R.G. Present day glaciers of Georgia and

va E.A., Dokukin M.D. History of the Donguz-Orun

evolution of glaciation in the mountains of Eurasia in

Glacier from bioindication, historical, cartographic

late Pleistocene and Holocene. PhD. Tbilisi, Institute

sources and remote sensing data. Ice and Snow. 2018,

of Geography, Georgian National Academy of Scienc

58 (4): 448-461. doi.org/10.15356/2076-6734-2018-4-

es. 1995: 320 p. [In Georgian].

448-461. [In Russian].

21. Podozersky K.I. Glaciers of the Caucasian Range. Za-

9. Serebryanyi L.R., Golodkovskaya N.A., Orlov A.V.,

piski Kavkazskogo otdela Russkogo Geograficheskogo

Malyasova E.S., Ilves E.O. Kolebaniya lednikov i

Obshchestva. Proc. of the Caucasian Branch of the

protsessy morenonakopleniya na Tsentral’nom Kavka-

Russian Geographical Society. 1911, 29 (1): 200 p. [In

ze. Fluctuations of glaciers and processes of moraines

Russian].

formation in the Central Caucasus. Moscow: Nauka,

22. Klein M.G., Gottdang A., Mous D.J.W., Bourlès D.L.,

1984: 216 p. [In Russian].

Arnold M., Hamelin B., Aumaître G., Braucher R., Mer-

10. Tielidze L.G. Glacier change over the last centu

chel S., Chauvet F. Performance of the HVE 5MV AMS

ry, Caucasus Mountains, Georgia, observed from

system at CEREGE using an absorber foil for isobar

old topographical maps, Landsat and ASTER satel

suppression. Nuclear Instruments and Methods in

lite imagery. The Cryosphere. 2016, 10: 713-725. doi.

Physics Research Section B: Beam Interactions with

org/10.5194/tc-10-713-2016.

Materials and Atoms. 2008, 266: 1828-1832. https://

11. Tielidze L.G. and Wheate R.D. The Greater Cauca

doi.org/10.1016/j.nimb.2007.11.077.

sus Glacier Inventory (Russia, Georgia and Azerbai

23. Braucher R., Guillou V., Bourlès D.L., Arnold M., Au-

jan). The Cryosphere. 2018, 12: 81-94. https://doi.

maître G., Keddadouche K., Nottoli E. Preparation of

org/10.5194/tc-12-81-2018.

ASTER in-house ¹⁰Be/⁹Be standard solutions // Nu

12. Tielidze L.G., Bolch T., Wheate R.D., Kutuzov S.S.,

clear Instruments and Methods in Physics Research

Lavrentiev I.I., Zemp M. Supra glacial debris cover

Section B: Beam Interactions with Materials and

changes in the Greater Caucasus from 1986 to 2014.

Atoms. 2015, 361: 335-340. https://doi.org/10.1016/j.

The Cryosphere. 2020, 14: 585-598. https://doi.

nimb.2015.06.012.

org/10.5194/tc-14-585-2020.

24. Chmeleff J., von Blanckenburg F., Kossert K., Jakob D.

13. Tielidze L.G. Dynamics of the Glaciers of Georgia.

Determination of the ¹⁰Be half-life by multicollec

Glaciers of Georgia. Geography of the Physical Envi

tor ICP-MS and liquid scintillation counting // Nu

ronment. Springer, Cham., 2017. doi.org/10.1007/978-

clear Instruments and Methods in Physics Research

3-319-50571-8_5.

Section B: Beam Interactions with Materials and

14. Freshfield D.W. The Exploration of the Caucasus.

Atoms. 2010, 268: 192-199. https://doi.org/10.1016/j.

1896. V. II. Edinburgh: T. and A. Constable, printers to

nimb.2009.09.012.

her majesty. London and New York.

25. Martin L.C.P., Blard P.-H., Balco G., Lavé J., Delunel R.,

15. Déchy M. von. Kaukasus Reisen und Forschungen im

Lifton N., Laurent V. The CREp program and the ICE-D

kaukasischen Hochgebirge (Travel and research in the

production rate calibration database: A fully parameter

Caucasian high mountains). 1905. Berlin, Band 1:

izable and updated online tool to compute cosmic-ray

313-314. [In German].

exposure ages. Quaternary Geochronology. 2017, 38:

16. Rutkovskaya V.A. Sections: Upper Svaneti Glaciers.

25-49. https://doi.org/10.1016/j.quageo.2016.11.006.

Transactions of the glacial expeditions. 1936, 5: 404-

26. Young N.E., Schaefer J.M., Briner J.P., Goehring B.M.

448.

A ¹⁰Be production-rate calibration for the Arc

17. Tsereteli D. Glacier change in the southern slope of the

tic: A ¹⁰Be production-rate calibration for the Arc

Greater Caucasus during the last 20-25 years). Works

tic. Journ. of Quaternary Science. 2013, 28: 515-526.

of Georgian National Academy of Sciences (Moam

https://doi.org/10.1002/jqs.2642.

be). 1959, XII (6). [In Georgian].

27. Lal D. Cosmic ray labeling of erosion surfaces: in situ

18. Tsereteli D., Khazaradze R., Lomtatidze G., Inashvi-

nuclide production rates and erosion models. Earth

li Sh., Lashkhi T., Kurdghelaidze G., Kalandadze G.

Planetary Science Letters. 1991, 104: 424-439. https://

and Chekurishvili R. Glaciological observations on the

doi.org/10.1016/0012-821X(91)90220-C.

Chalaati and Lechziri Glaciers (Upper Svaneti) in the

28. Stone J.O. Air pressure and cosmogenic isotope pro

spring of 1959. Georgian National Academy of Scienc

duction // Journ. of Geophys. Research. 2000, 105:

es. Works of Vakhushti Institute of Geography. 1962,

23753-23759. https://doi.org/10.1029/2000JB900181.

XVIII: 223-256. [In Georgian].

29. Stokes M.A., Smiley T.L. An Introduction to Tree-Ring

19. Shengelia R. Chalaati and Lekhziri glaciers regime in

Dating. University of Chicago Press. Chicago. 1968,

the summer of 1961. Georgian National Academy of

II: 73 p.

Sciences. Works of Vakhushti Institute of Geography.

30. Bushueva I.S., Solomina O.N. Kashkatash Glacier

1964, XX: 233-244. [In Georgian].

fluctuations in the XVII-XI centuries from carto

 469 

Палеогляциология

graphic, dendrochronological and lichenometric

35. Granshaw F.D., Fountain A.G. Glacier change (1958-

data. Ice and Snow. 2012, 52 (2): 121-130. https://

1998) in the North Cascades National Park Complex,

doi.org/10.15356/2076-6734-2012-2-121-130. [In

Washington, USA. Journ. of Glaciology. 2006, 52:

Russian].

251-256. doi: 10.3189/172756506781828782.

31. Beschel R.E. Flechten als Altersmaßstab rezenter

36. Innes J.L. Lichenometry. Progress in Physical Geogra

Moränen. Zeitschrift für Gletscherkunde und Ge

phy. 1985, 9 (2): 187-254.

ologie. 1950, 1: 152-162 (In German, translated by

37. Popovnin V.V., Rezepkin A.A., Tielidze L.G. Superficial

Barr W., Lichens as a measure of the age of recent mo

moraine expansion on the Djankuat Glacier snout over

raines. Arctic and Alpine Research. 1973, 5: 303-309).

the direct glaciological monitoring period. Earth Cryo

32. Osborn G., Menounos B., Ryane C., Riedel J.,

sphere. 2015, XIX (1): 79-87.

Clague J.J., Koch J., Clark D., Scott K., Davis P.T.

38. Holzhauser H., Magny M., Zumbühl H.J. Glacier and

Latest Pleistocene and Holocene glacier fluctuations

lake-level variations in west-central Europe over the

on Mount Baker, Washington. Quaternary Sciences

last 3500 years. Holocene. 2005, 15 (6): 789-801.

Review. 2012, 49: 33-51. doi.org/10.1016/j.quasci

https://doi.org/10.1191/0959683605hl853ra.

rev.2012.06.004.

39. Le Roy M., Nicolussi K., Deline P., Astrade L., Ed-

33. Jomelli V., Grancher D., Naveau P., Cooley D. As

ouard J.L., Miramont C., Arnaud F. Calendar-dated

sessment study of lichenometric methods for dating

glacier variations in the Western European Alps during

surfaces. Geomorphology. 2007, 86: 131-143. doi.

the Neoglacial: the Mer de Glace record, Mont Blanc

org/10.1016/j.geomorph.2006.08.010.

massif. Quaternary Sciences Review. 2015, 108: 1-22.

34. Naveau P., Jomelli V., Cooley D., Grancher D. Rabatel A.

https://doi.org/10.1016/j.quascirev.2014.10.033.

Modelling uncertainties in lichenometry studies with

40. Zumbühl H.J., Steiner D., Nussbaumer S.U. 19th century

an application: The Tropical Andes (Charquini Gla

glacier representations and fluctuations in the central

cier in Bolivia). Arctic, Antarctic and Alpine Research.

and western European Alps: an interdisciplinary ap

2007, 39: 277-288. https://doi.org/10.1657/1523-

proach. Glob. Planetary Changes. 2008, 60 (1): 42-57.

0430(2007)39[277:MUILS]2.0.CO;2.

https://doi.org/10.1016/j.gloplacha.2006.08.005.

 470 