Stability of Frozen Palsa at the Southern Border of its Distribution
on the Kola Peninsula

The Kola Peninsula has no genuine permafrost, but like in other Northern Fennoscandia regions with average annual air temperature of about – 1°C and lower, palsa bogs with ice core inside the mounds are an ordinary phenomenon there, as shown on the map of the Kola Peninsula vegetation in the “Atlas of Murmansk oblast” (Chernov, 1971). Their features and origin were studied in various parts of the world, such as Alaska, Northern Canada, Northern and North-Eastern Siberia, and especially Northern Fennoscandia including the Kola Peninsula (Sollid and Sorbel, 1974; Seppala, 1972, 1982a, 1983, 1994, 1997). The palsa issue attracts more and more attention in recent years in connection with the global warming hypothesis (Êukkonen & Safanda, 2001; Nelson et.al., 2001; Luoto and Seppala, 2002, 2003); palsa research project has recently started in Norway (Hofgaard, 2003).
There is much information on palsa of the Kola Peninsula. The first known description of palsa was published by Kihlman (1890). Anufriev conducted his expedition during the Civil War and described in detail some palsa bogs in Transactions of the Soviet Geographic Institute (Anufriev, 1922). The Permafrost Committee was active in the 30-s (Lavrova, 1934; Rikhter, 1934; Sumgin, 1934; Yegorov, 1938; Vlastova and Nikonov, 1940). According to the data published since 1890, bog mounds on the Kola Peninsula thaw to the depth of 30-35 cm by the end of summer.

The aim of the study was to evaluate the palsa stability on the Kola Peninsula for the last 65-80 years. The object of the study was not palsa in general, that covers 20-25% of the Kola Peninsula, but only the palsa at the southern border of its distribution. The palsa (bog mounds with permafrost) at the southern border of its distribution is susceptible to slightest change in temperature. Thus, the stable existence of palsa or its disappearance is a strong proof of temperature increase or lack thereof.

Southern border of the Kola Peninsula palsa lies about 67°50’ north latitude to the south of Monchegorsk town along the northern shore of Vite Gulf of Imandra Lake (Rikhter, 1934; Sumgin, 1934; Yegorov, 1938). Average annual air temperature was calculated with the help of the data from Reports of Monchegorsk Meteorological Station and was found to be –0.4?C for the last 70 years (1936-2006), with fluctuations from –3°C to +2.8°C (Figure 1). The sums of mean monthly winter temperatures (from October to April) for the period 1936 - 2002 ranged from –899°C to –2345°C; the average one was –1534 ±346°C (Figure 2). The sums of mean monthly summer temperatures (from May to September) for the same period varied from 1074°C to 1736°C; the average one was 1413 ±157°C (Figure 2).
Thirteen palsa bogs were investigated (see palsa layout scheme on Figure 3), two of them were investigated with the time difference of several decades: the palsa near Nyud Lake for 66 years (1936 and 2002), and the palsa near Yagel Bor halt for 82 years (1920 and 2002); information on two more palsas was provided by Yegorov (1938).
Fifteen palsa mounds were chosen in each bog using random numbers table; thickness of active layer inside the mounds in the second half of September (i.e. before the frost) was estimated with a steel measuring stick. Second half of September was chosen due to the peculiarity of transpolar climate; the seasons change very quickly, therefore the active layer of palsa has no time to grow. The palsa area was determined by terrestrial measurements in 1920/36 and 2002.

Table 1 presents some descriptions of investigated bogs.
The statistical key figures of the variants’ rows for thickness of active layer of palsa are given below. As mentioned above in the Methods section, 15 palsa mounds were chosen in every bog using random numbers table; calculation results are presented in Table 2.

Numbers and names of bogs:
¹1 – Travyanoye lake
¹2 – Nyud lake
¹3 – southern end of Moncha lake
¹4 – Lumbolka lake
¹5 – road to Yarva
¹6 – road M-18 – 1,267 km
¹7 - road Ì-18 – exit at 25 km
¹8 – Pustaya Gulf of Imandra Lake
¹9 – Yagel Bor halt
¹10 – Laplandia railway station
¹11 – Keresyavrench lake
¹12 – Schurie lake near Pulozero railway station
¹13 – to the north of Vite Gulf of Imandra Lake.

Statistical key figures of the variants’ row - thickness of active layer of palsa (cm)
¹ of lake	Arithmetic mean of the variants’ row	Sum of squares of conditional deviations	Dispersion of each row	Mean square deviation	Mean error of arithmetic mean	Significance test of arithmetic mean, Ì/ m M	Variation coefficient, %.
1	2	3	4	5	6	7	8
1	57	551	39	6	±1.5	38	10.5
2	64	562	40	6	±1.5	42.7	9.4
3	48	182	13	4	±1.1	43.6	8.3
4	55	462	33	6	±1.5	36.7	10.9
5	56	581	42	6	±1.5	37,3	10.7
6	50	145	10	3	±0.8	62.5	6
7	52	257	18	4	±1.1	47.3	7.7
8	44	555	40	6	±1.5	29.3	13.6
9	64	1447	103	10	±2.6	24.6	15.6
10	61	2118	151	12	±3.1	19.7	18.8
11	38	419	30	5	±1.3	29.2	13.2
12	61	2249	161	13	±3.3	18.5	20.3
13	63	1259	90	9	±2.3	27.4	14.3

Table 2 provides a clear insight into the statistical rows. Mean square deviation of each parameter varies from 3 to 6; it amounts to 10-13 only in three cases; the confidence interval is 0.95 and variation coefficient ranges from 6 to 20%. The obtained parameters are reliable; mean error of arithmetic mean varies from 1 to 3%, and accuracy figures – from 1 to 3%.
The Table 2 results were divided into two groups – bogs with live vegetation and bogs with dead vegetation on the surface – and presented in Table 3.

Comparative thickness of active layer of palsa with live and dead vegetation on its surface
¹¹ of palsa bogs	Bogs with dead vegetation (cm)	¹¹ of palsa bogs	Bogs with live vegetation (cm)
1	57 (from 49 to 66) ± 6	8	44 (from 35 to 53) ±6
2	64 (from 51 to 75) ±6	9	64 (from 45 to 75) ±10
3	48 (from 40 to 52) ±4	10	61 (from 54 to 84) ±12
4	55 (from 49 to 65) ±6	11	38 (from 30 to 46) ±5
5	56 (from 47 to 65) ±6	12	61 (from 49 to 84) ±13
6	50 (from 45 to 55) ±3	13	63 (from 45 to 73) ±9
7	52 (from 45 to 60) ±4

Vegetation on palsa surfaces Nos. 1, 2, 3, 4, 5, 6, and 7 (all close to Monchegorsk town) perished due to SO2 influence from smelter emissions.
The obtained data allows estimating statistical values of two rows of variants – mean thickness of active layer on palsas with live and dead vegetation on the surface.
Palsas with dead vegetation – Nos. of bogs: 1, 2, 3, 4, 5, 6, 7.
Palsas with live vegetation – Nos. of bogs: 8, 9, 10, 11, 12, 13.
Table 4 shows final statistical values of variants’ rows – mean thickness of active layer of palsas with live and dead vegetation.

Table 4
¹ of bogs	Arithmetic mean of the variants’ row	Sum of squares of conditional deviations	Dispersion of each row	Mean square deviation	Mean error of arithmetic mean	Significance test of arithmetic mean, Ì/ m M	Variation coefficient, %.
1	2	3	4	5	6	7	8
1	55	169	28	5	1.9	30	9
2	55	627	125	11	4.6	12	20

According to Railton and Sparling (1973), and Zuidhoff and Kolstrup (2005), absorption and loss of heat by palsa depends on the vegetation cover, but our measurements in 2002 and 2003 showed that the active layer in the mounds was 55±5 cm deep in bogs with dead vegetation and 55±11 cm deep in bogs with live vegetation (Table 4). It seems to contradict the published information, but it may be explained by the peculiarities of the vegetation cover on the palsas of the Kola Peninsula – it is very thin and sparse. It is difficult to imagine that such vegetation cover can alter significantly the heat exchange in the bogs.
The thickness of the active layer in the bog near Nyud Lake was 35-60 cm in 1936 and 45-70 cm in 2002; and in the bog near Yagel Bor halt it was 55-93 cm in 1920 and 40-75 cm in 2002.
It was found that the area covered by palsa, height and number of bog mounds, and depth of active layer in mounds had not changed for the last 80 years (Table 1).

Permafrost in palsa on the Kola Peninsula is not a relic of the last Great Ice Age (10-12 thousand years ago).
G.A. Blagoveshchensky dates palsa formation back to approximately 3,000 years ago on the basis of pollen analysis of peat near Nyud Lake (Monchegorsk town) (Yegorov, 1938). Lavrova (1934), Bradley & Jones (1993), and Grove (1988) hold the same point of view.
Several publications describe the process of preservation and contemporary formation of palsa (Tyrtikov, 1969; Seppala, 1982 a,b, 1994, 1998). The palsa near Nyud Lake can serve as an example – the freeze-thaw action there is connected with the topographic features of the place – the bog is open and exposed to the winds that blow into the Nyud Lake valley through a passage between Nyuduayvench and Sopchuayvench mounts (Figure 3) (Yegorov, 1938). Strong cold winds blowing along Nyud Lake in wintertime are the main factor of preservation and contemporary formation of palsa there; the winds blow snow off the peat mounds. The same meteorological conditions can be found in other places near Monchegorsk town and areas located 15 km, 22 km and 36 km away from Nyud Lake. All those bogs are located approximately along the same north-north-east line. Other authors adduce the same process of preservation and contemporary formation of palsa (Sollid and Sorbel, 1998; Matthews et al, 1997; Zuidhoff, 2002; Zuidhoff and Kolstrup, 2000).

Observations demonstrate various stages of palsa evolution (Seppala, 1988, 1998). Palsa formation begins when the snow cover is so thin at some places that winter frost penetrates deep enough to prevent the summer heat from thawing it out. Then, the surface of the bog grows (Figures 4,5,6,7). But no embryos of new palsa I discovered on investigated bogs. Yegorov and Seppala conducted an experiment on new palsa formation (Yegorov, 1938; Seppala, 1982b).

During subsequent winters the frost penetrates still deeper, the formation process accelerates and the hump continues to grow due to freezing of pore water and ice segregation. As the surface grows, the wind becomes ever more effective in drying the surface peat and keeping it clear of snow. Degradation starts now, and peat blocks collapse, and during later stages the palsa surface is exposed to deflation and rain erosion (Figure 8). Old palsas are partially destroyed by thermal karst.

Ò.å. î÷åâèäíî, ÷òî òîëùèíà àêòèâíîãî ñëîÿ èçìåíÿëàñü â ïîñëåäíèå 70-80 ëåò, â ñîîòâåòñòâèè ñ êîëåáàíèÿìè òåìïåðàòóðû îêðóæàþùåãî âîçäóõà. Íåêîòîðûå áóãðû, ïî-âèäèìîìó, èñ÷åçàëè (ðàñòðåñêèâàëèñü, óïëîùàëèñü), íî â öåëîì áîëîòà ñ áóãðàìè ñîõðàíÿëèñü. Ðàlsa ïîðàçèòåëüíî óñòîé÷èâû – ÷åðåç íèõ ïðîêëàäûâàëè êîëåè òÿæåëûõ ãðóçîâèêîâ, â áóãðû âêàïûâàëè ñòîëáû ýëåêòðîïåðåäà÷ – à ìåðçëîòà ïîñëå ýòîãî âñ¸ ðàâíî ñîõðàíÿëàñü â òå÷åíèå äåñÿòêîâ ëåò.

The frozen core of peat mound is blocked by a thick peat layer, which is a good thermal insulator; and positive summer temperatures affect only the thawing layer that can reach several dozen centimeters by the end of summer; the core remains frozen for centuries, though positive average annual temperature at the southern border of palsa distribution could lead to thawing of the ice core.

The information in this article shows that the palsa at the southern border of its distribution on the Kola Peninsula has not thawed out for over the last 80 years.

Anonymous. 1936-2006. Reports of Monchegorsk Meteorological Station. (In Russian).
Anufriev, G.L. 1922. On bogs of the Kola Peninsula. The works of the Kola Soil and Botany Group of the North Scientific and Production Expedition. Geographical Institute. Issue III. SPB, 35-65. (In Russian).
Bradley, R.S. and Jones, P.D. 1993. “Little Ice Age” summer temperature variation: their nature and relevance to recent global warming trends. Holocene 3:367-376
Chernov, V.G. 1971. The map of vegetation. Atlas of Murmansk oblast. Map # 17. (In Russian).
Grove, J.M. 1988. The Little Ice Age. Methuen, London
Hofgaard, A. 2003. Effects of climate change on the distribution and development of palsa peatlands: background and suggestions for a national monitoring project. NINA (Norwegian Institute for Nature Research) Project Report 21. Trondheim, Norway
Kihlman, A.O. 1890. Pflanzenbiologische Studien aus Russishc Lappland. Acta Societatis pro Fauna et Flora Fennica, VI, # 3, 57-60 Lavrova, M.A. 1934. On permafrost research in Volchiya (Wolf) and Monche tundras on the Kola Peninsula. Transactions of the Permafrost Committee, Issue III, 117-120 (In Russian).
Matthews, J.A., Dahl, S.-O., Berrisfjord, M.S. and Nesje, A. 1997. Cyclic development and thermokarstic degradation of palsas in the mid-alpine zone at Leirpullan, Dovrefjell, southern Norway. Permafrost and Periglacial Processes 8:107-122
Railton, J.B. and Sparling,J.H. 1973. Preliminary studies on the ecology of palsa mounds in northern Ontario. Can. J. Bot. 51:1037-1044
Rikhter, G.D. 1934. Information on peat mounds near Nyud Lake. Transactions of the Permafrost Committee, Issue III, 121-126 (In Russian).
Seppala, M. 1972. The term “palsa”. Zeitschrift fur Geomorphologie. 16, 4:463
Seppala, M. 1982à. The origin of palsas. Geogr. Ann, 63 A (3): 141-147
Seppala, M. 1982b. An experimental study of the formation of palsas. Proceedings 4th Canadian Permafrost Conference, National Research Council of Canada, Ottawa, 36-42
Seppala, M. 1983. Seasonal thawing of palsas in Finnish Lapland. International Permafrost Conference, 4th Proceedings, pp. 1127 – 1132
Seppala, M. 1988. Palsas and Reated Forms. In: Clark, M.J. (ed) Advances in Periglacial Geomorphology, 247-278
Seppala, M. 1994. Snow Depth Controls Palsa Growth. Permafrost and Periglacial Processes, v. 5: 283-288
Seppala, M. 1997. Distribution of Permafrost in Finland. Bull.Geol.Soc.Fin., 69, part 1-2, 87-96
Seppala, M. 1998. New permafrost formed in peat hummocks (pounus), Finnish Lapland. Permafrost and Periglacial Processes, 9: 367-373
Sollid, J.L. and Sorbel, L.1998. Palsa bogs as a climate indicator – examples from Dovrefjell, Southern Norway. AMBIO 27: 287-291
Sumgin, M.I. 1934. On permafrost in peat mounds on the Kola Peninsula. Transactions of the Permafrost Committee, issue III, 107-115 (In Russian).
Tyrtikov, A.P. 1969. Influence of vegetation cover on the ground freezing and thawing. Moscow. (In Russian).
Vlastova, N.V. and Nikonov, M.N. 1940. The peat deposit on the shore of Nyud-yavr Lake and prospects of its use. Report of the Peat Expedition work, # 1. Moscow. (In Russian).
Yakovlev, B.A. 1961. Climate of Murmansk province. Murmansk publishing house. 200 pp. (In Russian).
Yegorov, A.N. 1938. Permafrost in peatbogs on the Nyud Lake shores in Monche tundra (the Kola Peninsula). Transactions of the Permafrost Committee, Academy of Sciences of USSR, Issue VII, 113-125 (In Russian).
Zuidhoff, F.S. and Kolstrup, E. 2000. Changes in palsa distribution in relation to climate change in Laivadalen, northern Sweden, especially 1960-1997. Permafrost and Periglacial Processes 11: 55-69
Zuidhoff, F.S. 2002. Recent decay of a single palsa in relation to weather conditions between 1996 and 2000 in Laivadalen, northern Sweden. Geografiska Annaler 84 A: 103-111