Note: This data set was first published on the 1998 CAPS CD.
The text for this document was taken unchanged from that CD.

See also Description of Data Files.

Soil temperatures from historical climate station records, Former Soviet Union

David Gilichinsky
Institute of Soil and Photosynthesis, Russian Academy of Sciences
Pushchino, Moscow Region 124292, Russia, gilichin@issp.serpukhov.su

1. Description of the program and observing methods for soil temperature at Russian meteorological stations.

The program of observations at the Russian meteorological stations includes the following characterisations of soil thermal regime:

• Temperature of soil surface, specifically - the temperature of the soil surface in summer (usually bare soil), and in winter, the snow surface;
• Temperature of soil at depths of 0.05-0.20 m (during the warm season; usually on a plot without vegetation cover) using Savinov's bent-stem thermometers;
• Temperature of soil at depths of 0.20-3.20 m (usually under natural cover), using insertion thermometers;
• The depth of the soil freezing- thawing layer.

Temperature measurements at the soil surface are carried out at nearly all meteorological stations, the other measurements are taken at a selected network of stations.

The most important information about the conditions under which measurements are taken is given here. A complete description of the methods of measurement is given in "Instruction to meteorological stations and posts. Issue 3, Part 1. Editions 1946, 1954, 1958, 1969, 1985." For information about this referenced document, contact the data provider, David Gilichinsky, at gilichin@issp.serpukhov.su

1.1. Periods of observations.

During the period up to 1936, meteorological observations were carried out at 07, 13, 19 hours. From 1936 to 1965 they were made at 01, 07, 13, 19 hours local (mean solar) time, and from 1966 onwards the observations were carried out at 00, 03, 06, 09, 15, 18, and 21 hours Moscow time (Greenwich time + 3 hours).

Observations of the temperature of the soil surface and soil temperature at the depths up to 0.6 m carried out at each of the specified hours. The measurement at depths of 0.8 m or more - is made once a day, at noon; before 1965 at 13 hours local time, from 1966 at the closest observation hour to 14 hours (zone +1 hour) time.

In winter, when snow cover is deeper than 15 cm, or when the temperature in the layer 0-40 cm. is zero degrees, the observations at bent-stem thermometers are taken once a day at all depths, while the surface temperature is measured at all specified hours.

1.2. The place of observations

Plots for the observation of soil temperature are chosen in the south part of each meteorological enclosure, on a level, unshaded surface.

Shallow and bent-stem thermometers are usually installed on 4 x 6 m plots that is bare of vegetation. If observations by bent-stem thermometers are not carried out, the size of the plot for shallow thermometers is decreased to 3 x 4 m.

Insertion thermometers are usually installed on 6 x 8 m plots with natural vegetation, situated to the east of the plot for shallow thermometers.

In cases where the conditions of installation of equipment differed from the standard, it is noted in the resulting data files or tables.

1.2.1. Natural surface

Natural surface means grass cover, characteristic of the territory surrounding the station. However, the grass is regularly mowed on the meteorological enclosure, so that its height will not be more than 20 cm. Mowed grass is removed immediately from the plot. Snow cover is preserved in its natural condition.

1.2.2. Bare surface

"Bare surface" can refer to one of two types:

1. A cultivated area of surface free of vegetation ("bare fallow", "farm field", "black fallow", "arable land"). Modern standard methods of temperature measurement of surface and upper soil layers refer to this type. Every year in early spring the area is cultivated to a depth of 25-30 cm, using a rake. The surface is kept at the same level as that of the meteorological plot. If thawing and subsidence has occurred (especially in regions with permafrost), soil of the same composition is added to the surface of the plot. This cultivation and leveling is done systematically. Consolidation of soil, and crust and crack formation is permitted. Snow cover is preserved in its natural condition.

2. Bare area, free of vegetation and snow. On such areas the observations are carried out at very limited number of stations, usually in parallel with observations under the natural surface. The distinction from the previous variant is that soil is not hoed or raked, but only cleared of vegetation in summer and snow in winter.

1.3. Specific methods of measuring thermal characteristics:

1.3.1. Temperature of soil and snow surfaces

The temperature of soil and snow surface is measured by mercuryv and spirit meteorological thermometers: instantaneous, minimum and maximum thermometers. Minimum and Maximum thermometers are installed on areas bare of vegetation.

Thermometers are laid on soil surface so that their reservoirs and external cover are half into the soil, but not covered by it. If snow cover is present, thermometers should be installed similarly but on the snow surface. This method was used from 1947. Before that year, the methods of surface temperature measurement were repeatedly changed. Before 1940 the observations were carried out on a surface with natural cover, and from 1941 to 1946 mainly at 2 cm above the soil/snow surface, on a plot with a bare surface. Such changes of methods caused significant differences in the observational series. Therefore time series of observations are usually only published and used beginning with 1947 data.

Soil surface temperatures were commonly recorded with a precision of 0.1 degree C, but because the measurement method does not provide such high precision, beginning in 1959 the data are rounded off to whole degrees.

1.3.2. Temperature of upper soil layers using bent-stem thermometers (in the warm season)

The temperature of the upper soil layers is measured by mercury bent-stem thermometers (Savinov's thermometers) which are installed at depths of 0.05, 0.10, 0.15 and 0.20 m on tilled (cultivated) areas in 4 x 6 m plots that are free of vegetation. Minimum and maximum thermometers are also installed. At some stations during the earlier years, bent-stem thermometers were also installed at 0.30 m. These thermometers are installed in the spring after snow melting and preparation (digging) of the plot, and are removed in the autumn when the temperature has fallen to 0 degree C at the 5 cm depth. This method became generally accepted from the 1940s. Before that time the instructions did not contain specific instructions as to the type of surface and which observations were carried out using various procedures, but mainly gave instructions for taking measurements with natural cover. Such observations are now carried out at a limited number of stations.

1.3.3. Deep soil temperature from insertion thermometers

Observations are carried out using soil insertion thermometers, placed into ebonite tubes and installed on a plot with natural cover grass in summer, snow in winter. At present the standard depths adopted are principally 0.20, 0.40, 0.80, 1.60, 3.20 m, with additional depths at 0.60, 1.20 and 2.40 m. However, at some stations in earlier years other depths were used.

Earlier, thermometers at some stations were installed in tubes made from other materials (wooden, metallic and so on). But the results of measurements in metallic tubes can be distorted and can show a distorted picture of temperature distribution at depth because of their high heat conductivity. At a few stations, at the same time that observations were taken under a natural surface, observations were also made under a bare surface.

1.3.4. The depth of soil freezing/thawing

The depth of freezing was determined by the occurance of layers of ice, primarily by the method of drilling, or by the occurance of ice crystals and cementation of soil samples taken from different depths. Since the 1950s at most stations Danilin's cryopedemeter has been used to determine the depth of freezing.

2. Short description of sources and content of information about soil temperature for various years:

2.1. The main sources of information

1. Climate reference book of the USSR: Meteorological data for individual years. Issue 1-27, - Part VII. Soil temperature, fogs, rain, snow-storm, hail. - L.: Hydrometeoizdat, 1960-64.

   This reference book contains monthly data about soil temperature at depths (according to insertion and bent-stem thermometers) since beginning of observations up to 1950. Edited in 27 volumes for different territories (regions) that existed at the time of the regional administration of the Hydromeorological Service.

2. Climate of the USSR: Meteorological data for separate years. Issue 1-34. - Part VIII. Soil temperature. - 1970-78.

   This reference book contains the continuation of soil temperature at depths for 1951-1955, and also the temperature of surface from the beginning of observations up to 1965 (for the Yakutsk region up to 1970). The results of measurements of the depth of soil freezing are also published for separate regions.

3. Meteorological monthly. Issue 1-36. - Part II. - 1966-90. No 1 -13.

   Published during the 1960s-80s. (Number 13 contains summary data for the year). Monthly volumes can be considered as the continuation of the previous reference books, although the structure of publishing the data changed somewhat. 1990 was the last year this series was published.

2.2. Comments on the content of published data:

The temperature regime of soil and snow surface is characterized in reference books by the following values, calculated for measured data:

• Average monthly and yearly temperature of the soil surface;
• Average temperature minimum on soil surface;
• Average temperature maximum on soil surface;
• Absolute temperature minimum on soil surface;
• Absolute temperature maximum on soil surface;
• Dates of the first and last frost on the soil surface and duration of the frost-free period.

To characterize the soil temperature regime at depths the following values were used:

• Average monthly temperature of upper soil layers according to the bent-stem thermometers;
• Average monthly and yearly temperature of soil according to the insertion thermometers;
• Number of days with frost in soil;
• Date of the first and the last frost in soil;
• Depth of the 0 degree C temperature isotherm (calculated by means of interpolation between neighboring depths using daily data).

In addition, the results of direct measurements of freezing depth in separate regions were published.

In the reference books a number of specific statistics are included:

• Number of days with frost on soil surface;
• Minimum and maximum temperatures for a month and a year at each depth.

However, the dates of the first and the last frost in soil, the depth of penetration of 0 degree C isotherm and freezing depth were not published.

2.2.1. Temperature characteristics of the soil surface

We remind the reader that the temperature of soil surface is the temperature of soil surface in the absence of snow cover, but is the temperature of the snow surface when snow is present.

Data for most regions are given since 1947, that is from the beginning of observations, where the temperature is on a surface bare of vegetation. In cases where the data for preceding years are published (as for Yakutia), one should take note of their heterogeneity.

The following statistics are given:

• Average monthly and yearly temperatures of soil surface: This is the average soil temperatures for each month and year, obtained from daily observations by the instantaneous thermometer.

• Average minimum temperature of soil surface: This is calculated from daily average minimum temperatures of soil surface, for month and for a year, obtained from observations with a minimum thermometer.

• Average maximum temperature of soil surface: This is calculated from daily average maximum temperatures of soil surface, for a month and for a year, obtained from observations with a maximum thermometer.

• Absolute minimum temperature of soil surface: This characterises the lowest temperature of the surface for a month (year) as observed with a minimum thermometer.

• Absolute maximum temperature of soil surface: This characterises the highest temperature of surface for a month (year) as observed with a maximum thermometer.

• The dates of the first and the last frost on soil surface and duration of frost-free period: These are the dates of the first frost in the autumn and the last frost in the spring. A day with frost on the soil surface is a day with a minimum temperature <=0 degree C. The duration of the frost-free period is obtained by calculating the a number of days between the dates of last spring and first autumn frost. The frost-free period is considered absent if the duration of the interval between frost is not more than 30 days. The dates of the first and the last frosts in this case are not defined. In some years frosts can be observed in the middle of the summer, moreover before and after them the duration of the frost free period may not exceed 30 days. Such years are considered to be either years with two frost-free periods, or years with a frost-free period interrupted by frost.

• Number of days with frost on the soil surface: This is given in the Monthly Reports (for 1960s-80s). In earlier reference volumes this was not published.

2.2.2. Temperature of upper soil layers (according to bent-stem thermometers, in warm season of year)

Published data, as a rule (if it is not stipulated separately), are given for plots without vegetative cover ("black fallow"). Values given are:

• Average monthly temperature of upper soil layers on bent-stem thermometers: This is the average for monthly soil temperatures at given depth, obtained from daily observations according to Savinov's bent-stem thermometers.

• Maximum and minimum temperatures for a month: This is maximum and minimum temperatures, observed at given depth for a month (that is absolute). These values are given only in the Monthly Reports of the 1960s-80s.

Note that, because of peculiarities of the methods of observations in the warm season, average yearly temperatures, as well as their statistics, involve the 0( C temperature transition, which in individual cases cannot be defined.

2.2.3. Temperature statistics of soil at depths (from insertion thermometers)

Published data, as a rule (if it is not stipulated separately) are given for plots with a natural surface. The following values are given:

• Average monthly and yearly soil temperature according to insertion thermometers: This is the average monthly temperatures of soil at a given depth, obtained from daily observations using insertion thermometers.

• Maximum and minimum temperatures for a month: These are the maximum and minimum temperatures, observed at given depth for a month (that is absolute). They are given only in Monthly Reports for the 1960s-80s.

• A number of days with frost in soil: A day with frost in soil is a day in which at a given depth a temperature (0(C was observed.

• Dates of the first and the last frost in soil: These are the dates of the first (in the autumn) and the last (in the spring) frost in soil at different depths.

• The maximum depth of the 0 degree C isotherm: The deepest penetration of 0(C temperature for a month is calculated using daily temperatures, by means of linear interpolation between two neighbouring depths, at one of which the temperature is positive, and on the other negative. The depth is conditionally considered equal to 0, if 0(C temperature does not reach the depth nearest to the surface thermometer. Generally speaking, the 0 degree zero isotherm divides the regions of positive and negative temperatures. The depth of penetration of zero temperatures into the soil is either the penetration of positive temperatures "when soil thaws" or negative temperatures "when soil freezes" (inverted commas are used because the depth of penetration of temperature 0 degree C into soil does not generally coincide with the depth of freezing). The first is characterised when the layer of seasonal freezing-thawing is underlain by permafrost, the second if it is underlain by a non-freezing layer. Two tables of the depth of 0(C temperature penetration, "when freezing" and "when thawing", are given for the stations at which the depth of the seasonal freezing-thawing layer is greater than the depth of installation of the thermometers. The given value was not published in the Monthly Reports, but it can be derived in a similar manner (by interpolation) using extremes of monthly temperatures: minimum "when freezing", maximum "when thawing".

2.2.4. Soil freezing depth

The results of soil freezing depth measurements are not published for all regions. Various contradictory opinions are expressed about the homogeneity of the series when the method of freezing depth measure has changed, and also about advantages and deficiencies of each of these methods, that in some regions the results of observations are published using only one of these methods. These are published monthly, but for some stations they can be found in other editions.