5. Low air temperatures
5.1. Definition of criteria and initial data
5.2. Average and extreme annual values for the number of days with very low temperatures
5.3. Seasonal variations in the number of days with very low temperatures
5.4. Interannual changes
Air temperatures that are unusually low for a particular region make general human activity more difficult, cause a decline in productivity, and make the construction and operation of dwellings and industrial buildings more expensive. Low air temperature has a strong impact on flora and fauna. When the duration of a low temperature episode is very long, accidents and even death may result, and various classes of animals and plants may suffer. For these reasons, low air temperature is classified as an HMP [51, 52, 53, 66].
5.1. Definition of criteria and initial data
There are a number of indicators of low air temperatures which may be used to refine the characterization of the temperature field. Among these are the absolute minimum value, the minimum temperature during a particular year or month, and the average temperature for the coldest days or 5-day intervals, etc. Indices used for estimating the duration period of low air temperature events include the number of days with temperatures equal to or below a specified value per month or per year, the duration of intervals with a particular low temperature, and the number of days with average daily temperatures below certain specified limits. The cumulative index, freezing degree-days, is often used for practical applications. It is usually calculated as a sum of average daily temperatures for a specified time period (10-day period, month, season etc.), or as the product of average monthly temperature and number of days per month, etc. Many of the parameters described above are interrelated, making it possible to define one parameter using another one.
In this section, the number of days with minimum and mean daily air temperatures of -30 degrees Celsius and below, -40 degrees Celsius and below, and -50 degrees Celsius and below are selected as the main parameters of severe low temperature. Each classification includes the entries for the more severe case or cases. These parameters are by nature complex and include both the temperature value itself and an estimate of its duration period. These are sufficiently informative that they give a general measure of the scale of the phenomena, focusing attention on steady and continuing low temperature episodes, especially those unfavorable for human activity. The choice of several criteria rather than a single one for the designation of "severe low temperature" is connected to the significant variability of the air temperature in various parts of the Arctic and Subarctic.
To calculate the parameters for severe low temperature events, we have included observational data from the stationary network of polar stations, the North Pole drifting stations and other expeditions, information from climatic handbooks and other publications [34, 44, 55, 60], and published data from the Norwegian, Danish, and Canadian climatic journals.
The number of days with a minimum air temperature which does not exceed the specified limits was defined by data from handbooks on climate. Using this type of parameter it was possible to obtain a general idea about the probable duration of both short-term and stable severe temperature events at the various stations during different months, and over the entire year. This analysis was extended to the boundaries of the Eurasian arctic and subarctic territories where severe temperatures are also observed.
The number of days with average daily air temperatures at or below the specified limits was calculated directly using data from the TM-1 standard meteorological tables for the period 1966-1985. A twenty year series of mean monthly and annual values were obtained. For some stations the series was extended to 1990. This characterizes the occurrence of stable severe temperature episodes with a duration of 24 hours or more. These statistics are also included as part of the total number of days with the lowest air temperatures.
The samples thus obtained were used to develop a variety of equations, some of which were used to define certain parameters of severe low temperatures by indirect methods. We describe three of these methods:
![]() |
(5.1) |
where t T is total duration in hours of air temperature of T or below per year, and nT is the number of days with an air temperature of T or below. Examples of these calculations are presented in Figure 30.
Figure 30. (a) The number of days with mean daily temperatures of -30 degrees Celsius and below (n-30) versus monthly mean temperatures in February; (b) The number of days with minimum temperatures of -30 degrees Celsius and below (nmin-30) versus n-30 for the Eurasian Arctic; and (c) The number of days with mean daily temperatures of -40 degrees Celsius and below (n-40) versus the corresponding number for -50 degrees Celsius (n-50).
The curve labels correspond to the following regions
1 - the Laptev, Kara and East Siberian seas;
2 - the Chukchi Sea;
3 - the Canadian Arctic Archipelago and Greenland
We consider briefly a method suggested for indirect calculation. In Figure 30a, three separate curves are clearly delineated for the various regions showing the relationship between the number of days with average temperatures of -30 degrees Celsius and below, n-30, and the mean monthly temperature in February, one of the coldest months of the winter in the Kara, Laptev, East Siberian, and Chukchi Seas as well as in the coastal areas of the Canadian Arctic Archipelago. Figure 30b shows the correlation between the number of days with mean daily air temperature of -30 degrees Celsius and below in the Eurasian Arctic and the number of days with a minimum temperature of -30 degrees Celsius or below, nmin-30 . This relationship has been used to perform calculations for a series arctic and subarctic stations. The initial data used to perform calculations were taken from climate manuals for the USSR and for North America [34, 44]. While performing the analysis it was discovered that the relationship shown in Figure 30b can be used to calculate the number of days with average daily temperatures of -40 degrees Celsius and below as well. The empirical data that can be used for this are marked by crosses. In some cases, to improve the precision of the data, the relationship between the number of days and the average daily temperature of -40 degrees Celsius and -50 degrees Celsius and below was used (Figure 30c). It has a regional character similar to that shown in Figure 30a. Separate relationships are observed for the coastal zones of the Laptev Sea, the Canadian Arctic archipelago and the Greenland Sea.
Reference [29] contains data calculated from thermograph readings of the duration periods that had temperatures below the specified limits as recorded at several polar stations for 1951 through 1960. We recalculated these characteristics using equation (5.1) and obtained the calculated numbers of days for the period 1951-1960, which we then compared with our samples for the period 1966 to 1985 (Table 31).
The results of the comparison show satisfactory agreement between the calculated data for the period 1951 to 1960 using formula (5.1) and observed data from 1966 to 1985. As a result, we conclude that it is sufficiently accurate to use formula (5.1) to perform further calculations using additional data presented by Zykova (29).
Using several different methods, the annually averaged number of days with mean daily air temperature below the specified limits were obtained from data obtained at 81 Russian stations for the period 1966-1985. In addition, analogous calculations were performed using data from the following sources: seven Canadian stations for the period 1954 to 1977; five Norwegian stations for the period 1962-1979; the stations on Spitsbergen and islands in the North-European sector of the Arctic Ocean (Jan-Mayen, Hopen, and Bear Island); 15 stations on the Greenland coast from 1964 to 1971; and one high-altitude mountain station (3 years). In all, the number of days with mean daily air temperature at or below -30, -40, and -50 degrees Celsius was determined using data from 154 stations.
Table 31Comparison of calculated and observed values of the number of days with severe low temperatures. | |||
Station | Number of days with average daily temperature at | ||
-30degreesC and below | -40degreesC and below | -50degrees C and below | |
Dixon Island | 40/36 | 2.1/1.8 | 0/0 |
Khatanga | 81/81 | 26/25 | 0.4/1.3 |
Kyusyur | 89/90 | 37/38 | 3.7/3.5 |
Ambarchik | 54/59 | 3/5 | 0/0 |
Uelen | 13/10 | 0/0 | 0/0 |
Note: Calculations using equation (5.1) are in the numerator, and the observed results are in denominator.
The statistics for the duration of low temperature were determined for the central arctic basin using data from the North Pole drifting stations for each month of drift from 1954 though 1990. Thereupon the averages for each season were determined for each station and the average coordinates were defined to which the data were referred. In total, 351 monthly and 51 seasonal values were obtained for the arctic basin.
The calculated parameters for severe low temperatures were used to develop background climatic charts, to define annual variations and interannual changes, and to calculate statistics for estimating the correlations among the number of days with severe low temperature events at different stations and for other related purposes.
5.2. Average and extreme annual values for the number of days with very low temperatures.
The spatial scale and extent of severe temperature events can be estimated using the corresponding climatic charts. Background maps of the number of days with various minimum air temperatures, Tmin, were developed using the data from 215 stations (Figures 31 and 32). Temperatures of -30 degrees Celsius and below are observed over all areas of the Arctic and Subarctic from the Kola Peninsula, the Barents and White seas to the Chukchi Peninsula, the Bering Sea and the Sea of Okhotsk. The boundaries of the -30 degrees Celsius minimum-temperature isotherm in the Eurasian sector is located far from the Arctic and Subarctic boundaries to the south of the Russian border. To the west, the boundary of this isotherm passes across the Barents Sea. In the Asian Subarctic, the longest duration period with Tmin £ -30 degrees Celsius is 140 to 160 days. This occurs in the basins of the Yana and Indigirka rivers.
Figure 31. Number of days per year with (a) Tmin £ - 40 degrees Celsius and (b) Tmin £ - 30 degrees Celsius.
Figure 32. Number of days per year with Tmin £ -50 degrees Celsius. The dots indicate the locations where Tmin £ - 60 degrees Celsius was recorded.
Extent of the area over which severe temperatures with Tmin £ -40 degrees Celsius are observed is significantly less than that for which the air temperature is -30 degrees Celsius and below. This decrease occurs, for the most part, as a result the influence of the Barents Sea, the Bering Sea and the Sea of Okhotsk. The number of days with temperatures of -40 degrees Celsius and below decreases considerably over the entire Eurasian Arctic and Subarctic, however, this number is also noticeably higher (110 to 120 days) in the Yana and Indigirka river basins.
The area over which very severe temperature conditions (-50 degrees Celsius and below) were observed shrinks further and the frequency of occurrence decreases sharply. This decrease occurs as a result of the marine influence of the northern and arctic seas. In the regions with the maximum occurrence of such severe temperatures, the Yana and Indigirka river basins, there are 40 to 60 days with minimum temperatures of -50 degrees Celsius and below.
The lowest air temperature category, -60 degrees Celsius or less, is observed in Eurasia in the region influenced by the Siberian High, and is limited to selected areas in the basins of Anabar, Lena, Yana, Indigirka, Enisey and Kolyma rivers (Fig. 32). In the northern hemisphere, such temperatures are also observed at high elevations in the mountainous areas of Greenland and in some valleys of northern North America.
We now consider the number of days with low mean daily temperatures as a subset of the total number of days with severe low temperatures. The percentage of such days with stable temperatures of -30 degrees Celsius and below is 40 to 55 percent for the western and eastern arctic coastal zones of Eurasia and 70 percent or more for the continental parts of central Eurasia.
The annual number of days with severe low temperatures for the Arctic basin is presented in Table 33. These data, which include information about the Subarctic, and information from Russian and foreign stations, were used to develop corresponding background climatic maps of the Arctic (Figs. 33, 34, and 35).
Figure 33. The number of days per year with mean daily air temperatures of -30 degrees Celsius or below.
Figure 34. The number of days per year with mean daily air temperatures of -40 degrees Celsius or below.
Figure 35. The number of days per year with mean daily air temperatures of -50 degrees Celsius or below.
As noted above, these charts show the stable part of severe low temperatures and, on one hand, complement the maps of number of days with minimum air temperature developed earlier (Figs. 31, 32) and, on the other hand, include a more extended area.
The distribution of the new indices for severe low temperatures for the Eurasian Arctic and Subarctic as a whole is consistent with previous results. The greatest values of n-30 (in excess of 100 days) are observed in the Asian Subarctic and in the circumpolar region. These conditions are seldom encountered in the Atlantic section of the Arctic and Subarctic. Greenland and the Canadian Arctic archipelago also show local maxima in the n-30 field. In the majority of regions, the southern boundary of n-40 passes across the Arctic. Only in Asia does it dip down to subarctic and midlatitudes.
Average daily air temperature of -50 degrees Celsius and below, n-50, is observed relatively rarely in the Eurasian Arctic, mainly in the Asian continental area. Such temperatures are very unusual for almost the entire oceanic part of the arctic basin. As is evident in Figure 31, the maximum values of n-50 in the northern hemisphere are observed in the Asian Subarctic, Greenland, and Ellesmere Island in the Canadian Arctic archipelago.
Table 32Number of days with mean daily temperature -30, -40, & -50 degrees Celsius and below in the arctic basin | ||||||||||
Station | Observing Season | Number of days | Station | ObservingSeason | Number of days | |||||
n-30 | n-40 | n-50 | n-30 | n-40 | n-50 | |||||
NP-4 | 1955/56 | 112 | 23 | 0 | NP-18 | 1969/70 | 69 | 0 | 0 | |
NP-5 | 1955/56 | 86 | 17 | 0 | 1970/71 | 75 | 20 | 0 | ||
NP-6 | 1956/57 | 65 | 18 | 0 | NP-19 | 1968/69 | 70 | 15 | 0 | |
1957/58 | 100 | 14 | 0 | 1971/72 | 99 | 22 | 3 | |||
1958/59 | 107 | 18 | 0 | 1972/73 | 98 | 26 | 0 | |||
NP-7 | 1957/58 | 114 | 20 | 0 | NP-20 | 1970/71 | 72 | 12 | 0 | |
1958/59 | 114 | 19 | 0 | 1971/72 | 95 | 6 | 0 | |||
NP-8 | 1959/60 | 69 | 5 | 0 | NP-21 | 1972/73 | 91 | 10 | 0 | |
1960/61 | 89 | 12 | 0 | 1973/74 | 79 | 3 | 0 | |||
1961/62 | >101 | 20 | 0 | NP-22 | 1973/74 | 69 | 6 | 0 | ||
NP-9 | 1960/61 | 87 | 15 | 0 | 1974/75 | 112 | 24 | 1 | ||
NP-10 | 1961/62 | 66 | - | 0 | 1975/76 | 112 | 18 | 0 | ||
1962/63 | 117 | 14 | 0 | 1976/77 | 136 | 36 | 0 | |||
1963/64 | 115 | 25 | 0 | 1977/78 | 76 | 5 | 0 | |||
NP-11 | 1962/63 | 97 | 19 | 0 | NP-23 | 1976/77 | 99 | 1 | 0 | |
NP-12 | 1963/64 | 103 | 28 | 0 | 1977/78 | 109 | 9 | 0 | ||
1964/65 | 102 | 19 | 0 | NP-24 | 1978/79 | 116 | 24 | 0 | ||
NP-13 | 1964/65 | 98 | 12 | 0 | 1979/80 | 120 | 16 | 0 | ||
1965/66 | 90 | 8 | 0 | NP-26 | 1984/85 | 90 | 8 | 0 | ||
1966/67 | 107 | 35 | 0 | NP-27 | 1984/85 | 97 | 4 | 0 | ||
NP-15 | 1966/67 | 100 | 18 | 0 | 1985/86 | 104 | 20 | 0 | ||
NP-16 | 1968/69 | 127 | 10 | 0 | 1986/87 | 97 | 10 | 0 | ||
1969/70 | 96 | 3 | 0 | NP-28 | 1987/88 | 120 | 15 | 0 | ||
1970/71 | 105 | 31 | 0 | NP-29 | 1987/88 | 116 | 12 | 0 | ||
1971/72 | >102 | 27 | 0 | NP-30 | 1987/88 | 75 | 5 | 0 | ||
NP-17 | 1968/69 | 116 | 33 | 0 | NP-31 | 1988/89 | - | 6 | 0 | |
NP-18 | 1968/69 | 67 | 0 | 0 | 1989/90 | 113 | 11 | 0 |
Note: In the winter of 1976/1977, station NP-22 drifted into the vicinity of 83 degrees North, 127 degrees West.
The spatial distribution of the annual duration of severe low temperatures agrees well with atmospheric circulation, location of the land and ocean, especially warm sea currents, geographical zone and height above sea level. The maximum duration period of very severe temperature conditions is observed on the continents but not in the circumpolar region.
During most of the year, the Arctic Basin is shrouded in a relatively thin layer of cold air that is isolated from the upper levels of the atmosphere to a significant extent due to the atmospheric inversion. Air temperature near the ground depends principally on the temperature and detailed characteristics of the snow-ice surface, and on the oceanic heat flux. Influence of the ocean can be clearly identified by considering the differences between air temperatures of -30 degrees Celsius, -40 degrees Celsius and -50 degrees Celsius. The number of days per season with an average daily air temperature of -40 degrees Celsius was 15 percent of n-30, while the corresponding value for - 50 degrees Celsius was less than 0.1 percent on n-30. During this same period, these proportions for valleys of Yana and Indigirka rivers were 75 percent and 33 percent respectively.
Very severe climatic conditions occur during extended intervals of extremely severe low temperatures.
The greatest number of days with mean daily air temperatures of -30 degrees Celsius and below varies from a few days up to 100 to 200 days on the arctic coast of Eurasia and adjacent islands (Andrey Island, Zemlya Bunge, Pronchishcheva Bay, Tyumeti, Kusyur etc.). It was measured at 46 stations in 1979. The maximum number of days with average daily air temperatures of -40 degrees Celsius and below varies along the Eurasian Arctic coast from 1 to 2 days (Amderma, Viktoriya Island, Wrangel Island, Cape Netten) to 20 to 40 days (Andrey Island, Dunay, Cape Kosistyi etc.), but in the more continental regions of Siberia n-40 can be 50 to 60 days or more (Tyumeti and Kusyur). An extreme value of n-40 was recorded in 1979 at 42 stations, in 1966 at 16 stations, and in 1978 at 5 stations. Along the coast of Siberia and on the islands, extreme cases of severe low temperatures of -50 degrees Celsius and below last 1 or 2 days. With increasing distance inland from the coast, the duration of such severe temperatures increases to 15 to 20 days and more (for example, at Tyumeti and Kusyur).
In the Barents Sea, the southern area of which is influenced by warm ocean currents, severe conditions with average daily temperatures of -30 degrees Celsius or below are observed only in the northern and eastern regions, which are regularly ice-covered. The largest values of n-30 (50 to 56 days) were recorded on the islands of Franz Josef Land (Rudolf and Hayes islands), where temperatures of -40 degrees Celsius and below were observed 2 to 4 days. The maximum number of days with a temperature of -30 degrees Celsius and below in the southeastern Barents Sea was observed at Cape Belyi Nos (27 days). On Novaya Zemlya corresponding temperature conditions were recorded at Maliye Karmakuly (7 days) and Russkaya Gavan' (21 days) stations. On Hope Island and Spitsbergen (Ny-Ålesund) the maximum was 5 to 6 days. In the seas of the North-European basin, mean daily temperatures of -50 degrees Celsius or below were never observed.
In the arctic basin the maximum number of days with temperatures of -30 degrees Celsius and below for the years covered by this study was in excess of 100 days. The maximum of 136 days observed during the winter of 1976/77 was recorded during the drift of station NP-22 (see Table 33). The maximum number of days (36) with temperatures of -40 degrees Celsius or below was observed at this particular station during this same season. More than 30 days with temperatures of -40 degrees Celsius or below were recorded during the drift of NP-13 during the winter of 1966/67, at NP-16 during the winter of 1968/69, and at NP-18 during the winter of 1970/71. During the winter of 1971/72, three days with average daily temperatures of -50 degrees Celsius or below were reported at station NP-19. At that time, the station was located near 87 degrees North latitude, 114 degrees East longitude. In addition to the observations of severe low temperatures at specified stations, it is also of considerable interest to examine situations where low air temperatures are recorded simultaneously at several locations in a particular region. Particularly in cases of this sort, such a phenomenon can be classified as hazardous with considerable justification.
As a specific example, we consider the region of the European Arctic. To develop a chronology of the periods of severe temperature coincidence at two or more stations, two regions were selected: the northern part of the Barents and Kara seas (the region of Franz Josef Land and of the northern part of Novaya Zemlya) and the area around Vaygach Island in the southern part of the Barents and Kara seas. Data were used from 5 stations in the former region and from 4 stations in the latter.
The dates of severe temperature events for which average daily temperatures were -30 degrees Celsius or below at stations inside each region were selected and corresponding periods were compared for the two areas. Data from the following five years were used: 1967, 1969, 1978, 1979, and 1982. It was found that as many as 17 instances were observed each year at several of the stations on Franz Josef Land and Severnaya Zemlya where an air temperature of -30 degrees Celsius or below extended over 40 to 100 percent of the area. The total number of days with such conditions during those years ranged from 41 to 53. In the larger southern region around Vaygach Island, the number of days with average daily temperatures of -30 degrees Celsius and below varied from 8 to 29 days. The date and time intervals for which the temperatures were lower than -30 degrees Celsius at more than 5 stations at the same time in the two regions specified above are presented in Table 33.
Table 33Chronology of severe low temperatures (-30degreesC or below) for the regions of Franz Josef Land, Novaya Zemlya and Vaygach Island for specific years | |||
Date | No. of Days | Number of stations with mean daily temperatures of -30degreesC or less | Percentage of study area covered by the phenomenon |
26-29 Jan 1967 | 4 | 6 - 9 | 70 to 100 |
13-15 Jan 1969 | 3 | 9 | 100 |
14 Feb 1978 | 1 | 9 | 100 |
27-30 Dec 1978 | 4 | 8 - 9 | 90 to 100 |
2 & 3 Jan 1979 | 2 | 9 | 100 |
1-10 Feb 1979 | 10 | 7 - 9 | 80 to 100 |
14-16 Feb 1979 | 3 | 5 - 9 | 50 to 100 |
17 & 18 Jan 1982 | 2 | 9 | 100 |
26-28 Feb 1982 | 3 | 9 | 100 |
Despite these considerations, temperatures lower than -30 degrees Celsius are observed every year at various stations and such events can hardly be classified as HMP. Rather, they are an integral indicator of the general severity of the Arctic climate. However, for the southwestern part of the Barents Sea, an air temperature of -30 degrees Celsius or below can be considered an HMP.
Table 34Chronology of severe temperatures (-40 degrees C and below) for the regions of Franz Josef Land and Vaygach Island for specific years | ||
Date | No. of Days | Amount of stations with average daily air temperature -40 degrees Celsius and below |
Franz Josef Land | ||
14 Jan 1967 | 1 | 2 |
26 Feb 1967 | 1 | 2 |
5 Mar 1969 | 1 | 2 |
1 Mar 1978 | 1 | 2 |
16 Feb 1979 | 1 | 2 |
21 Feb 1982 | 1 | 2 |
Vaygach Island Region | ||
3 Feb 1979 | 1 | 2 |
8 Feb 1979 | 1 | 2 |
Note: The stations at Franz Josef Land are Nagurskaya and Rudolph Island and in Vaygach Island region Amderma and Belyi Nos.
Average air temperatures of -40 degrees Celsius or less are not observed every year and not at every station in the Eurasian Arctic. Specifically, they are not observed along the Barents Sea coast of the southern island of Novaya Zemlya. A chronology of severe temperatures at or below -40 degrees Celsius for Franz Josef Land and Vaygach Island is presented in Table 34. From these results it is evident that average daily temperatures of -40 degrees Celsius or less are rarely observed at several stations simultaneously in the European Arctic. In the Vaygach Island region this was recorded only in February, 1979 although the observation period extended from 1966 to 1989. It follows from this that air temperatures below -40 degrees Celsius in the European Arctic are observed over a considerably smaller area than temperatures of -30 degrees Celsius and below, and their duration is less than 24 hours. Thus for the area investigated here, such temperature events are short-term and localized, but there is no guarantee that their duration could not become longer or that their area could not increase. A temperature of -40 degrees Celsius and below in the European Arctic can probably be classified as a hazardous phenomenon if it is observed in a region where a stable air temperature decrease to these values is observed simultaneously at two or more stations.
On average, air temperatures below -30 degrees Celsius are observed for 500 to 1000 hours per year in the western and eastern sectors of the Eurasian Arctic and for 2000 hours or more in the central part. The maximum continuous duration of -30 degrees Celsius and below temperatures can range from 160 to 400 hours per year in the western and eastern regions up to 650 hours or more in the central region. In the central region, the greatest duration of -40 degrees Celsius and below temperatures are observed. The largest value, at Khatanga, is 750 hours. The maximum continuous duration of temperatures below -40 degrees Celsius recorded in Khatanga was 300 hours in January, 1979.
5.3. Seasonal variations in the number of days with very low temperatures
In the European Arctic, the period of severe frosts with temperatures of -30 degrees Celsius and below extends, as a rule, from November through April (Table 35). At some stations such temperatures are observed in October and, very rarely, in May. On coasts and islands of the arctic seas about 90 percent of the days with temperatures of -30 degrees Celsius and below are observed from December through March. Over the continental Asian Arctic about 80 percent of such days fall within this period. The maxima of the mean monthly number of days with these temperatures are smoothed out over the course of the year.
The distribution of the number of days with air temperatures of -40 degrees Celsius and below has a more clearly defined monthly maximum in January and February. Along the coasts and islands of the Eurasian Arctic 70 to 79 percent of these days occur in these two months compared with about 60 percent for the continental part of the region.
Distribution of the number of days with severe temperatures of -50 degrees Celsius or below in the central continental region is characterized by a clear maximum in January. In the marine regions of the Arctic, such low temperatures can be observed only during the winter (January-March) and they occur very infrequently (less than 1 case every 10 years).
The seasonal progression of the number of days with severe low temperatures for the Arctic basin is presented in in Figure 35. Monthly mean values were obtained by averaging the observational data from the North Pole drifting stations (55-60 values for each month) and represent an estimation of the general patterns. The volume of data is not enough to perform a more detailed evaluation. The intra-annual distribution of n-30 in the arctic basin is characterized by the absence of clearly defined peaks. From December through March approximately 20 percent of the yearly total is observed each month. This distribution is close to that observed along the coasts and islands of the arctic seas, but is more smoothed out. The maximum of the intra-annual distribution of severe temperatures of -40 degrees Celsius and below is observed in February.
Table 35Seasonal distribution of the number of days with severe low temperatures, % | |||||||||
Region | Air temperature degrees C | Month | |||||||
Jan | Feb | Mar | Apr | May | Oct | Nov | Dec | ||
The Eurasian Arctic | |||||||||
Coastal Zone and Islands | |||||||||
Western Region | -30 | 25 | 24 | 24 | 5 | 0 | 0 | 6 | 16 |
-40 | 35 | 37 | 18 | 0 | <1 | 9 | |||
Central Region | -30 | 25 | 24 | 21 | 4 | 0 | 0 | 7 | 19 |
-40 | 41 | 38 | 11 | 1 | 0 | 0 | 9 | ||
Eastern Region | -30 | 23 | 28 | 24 | 4 | 0 | <1 | 4 | 16 |
-40 | 32 | 38 | 13 | 0 | <1 | 16 | |||
Continental Area Central Region | |||||||||
-30 | 23 | 20 | 17 | 5 | 0 | 1 | 14 | 20 | |
-40 | 33 | 26 | 10 | <1 | 0 | 8 | 22 | ||
-50 | 53 | 25 | 2 | 2 | 18 | ||||
Arctic Basin | |||||||||
NP drift station region | -30 | 21 | 21 | 22 | 5 | 0 | 2 | 10 | 19 |
-40 | 27 | 31 | 22 | 1 | 3 | 16 |
To estimate the interannual changes and climatic trends of severe low temperatures, the number of days with average daily air temperatures of -30 degrees Celsius and below for a year was selected. An archive of annual data was assembled including results from 38 stations in the Eurasian Arctic for a 20-year period and from 5 stations of the Canadian Arctic for an 18-year period. Selected statistical characteristics and sets of extreme values for the Eurasian Arctic are presented in Table 36. From these data it is evident that the duration of severe temperature events varies over a wide range. The coefficient of variation (Cv = rms deviation/mean No. of severe temperature days) varies from 0.15 to 0.20 for the regions where severe temperatures are a characteristic feature of the climate to 0.7 to 0.8 for those regions where low temperatures are observed relatively infrequently. Figure 32 shows the dependence of the coefficient of variation for the duration of low temperature events versus the corresponding mean values. The location of the observed points shows the general relationship for different arctic regions. This dependence can be used to estimate the variability of the duration of stable severe temperature conditions for regions where the density of observation is not as high.
Table 36Statistical characteristics for the number of days per year with average daily temperatures of -30degreesC and below. | |||||
Station | Average No. of days | Standard deviation | CV | Minimum | Maximum |
Rudolph Island | 28 | 11.2 | 0.40 | 9 (1984) | 50 (1981) |
Cape Zhelaniya | 11 | 9.7 | 0.88 | 0 (1975) | 32 (1979) |
Amderma | 10 | 8.8 | 0.88 | 0 (1982, 1983 ) | 32 (1979) |
Belyi Island | 31 | 12.5 | 0.40 | 8 (1984) | 60 (1979) |
Tadibeyakha | 43 | 13.1 | 0.30 | 19 (1983) | 69 (1979) |
Dixon Island | 36 | 13.7 | 0.38 | 9 (1983) | 62 (1979) |
Golomyannyi Island | 54 | 13.5 | 0.25 | 15 (1985) | 75 (1969) |
Pravdy Island | 57 | 12.1 | 0.21 | 34 (1983) | 84 (1979) |
Khatanga | 81 | 14.4 | 0.18 | 53 (1983) | 108 (1979) |
Tiksi | 68 | 15.0 | 0.22 | 37 (1975) | 99 (1966) |
Kazachye | 91 | 16.6 | 0.18 | 62 (1981) | 115 (1979) |
Chokurdakh | 90 | 13.1 | 0.15 | 53 (1985) | 111 (1967) |
Pevek | 45 | 11.3 | 0.25 | 26 (1970) | 71 (1976) |
Wrangel Island | 16 | 6.5 | 0.41 | 4 (1982) | 26 (1966) |
Uelen | 11 | 7.9 | 0.72 | 0 (1967, 1982, 1983) | 28 (1977) |
Note: The year when the maximum and minimum number of days were recorded are shown in parentheses.
Interannual variations in the number of days with severe low temperatures in various Arctic regions are presented in Figure 37. Visual analysis shows that there are short periods of oscillations in these series, but no significant unidirectional trends over the 20-year period.
Figure 36. The coefficient of variation of the mean duration of severe low temperatures with air
temperatures of -30 degrees Celsius or below:
1- the Eurasian Arctic; 2- the Canadian Arctic
Figure 37. Interannual variation in the number of days with severe low temperatures of (a) -30 degrees Celsius and below and (b) -40 degrees Celsius and below at the following stations:
1 - Tiksi, 2 - Pevek, 3 - Belyi Island, 4 - Tyumeti, 5 - Khatanga
Interannual variations in n-30 are interrelated at different stations because they are mostly part defined by atmospheric circulation patterns in the arctic and midlatitudes. The interdependence becomes weaker as the distance between the locations increases, which can be expressed quantitatively with correlation coefficients. The correlation coefficients were calculated using 20-year time series for 24 stations located along the coast and on the islands in the Kara Sea and on the coast of the Chukchi Peninsula. Correlation functions r(r ) calculated as a function of the distance between the locations (r ) are presented on Figure 38. The results show that the distribution of correlation isolines is anisotropic. This is well represented by the example from the Kara Sea in which correlation length (the e-folding distance of CV) is more than 2000 kilometers along the coast but only 1100 kilometers along the meridian. The minimum correlation length (900 kilometers) was determined for the arctic coast of Chukchi Peninsula. This is related to the more compact characteristic dimensions of the region and with the structure of the atmospheric circulation in the Pacific Ocean sector.
Figure 38. Spatial correlation function of n-30: 1 - for the zonal direction (in the coastal region) and 2 - for the meridional direction, both in the Kara Sea region, and 3 - for the arctic coast of the Chukchi Peninsula.
To solve a variety of practical problems, it is important to know the absolute minimum temperature values ever recorded including the short-term temperature decreases. The corresponding map for the Russian Arctic and Subarctic (Figure 35) was developed using data from the Climatic Handbook of the USSR [44] over a 50-year period, from North Pole drifting station data covering the last 40 years, and from selected foreign and Russian publications [32, 33, 34, 42].
Figure 39. Absolute minimum values of air temperature (degrees Celsius) observed at the surface from 1936 to 1986.