The investigated region and grid points of a numerical model are shown in Fig. 1. Obtained long-term (119-year period) air temperature time series for the NW Russia are shown in Fig. 2.

Fig. 1. Map of the region of investigation and location the ECHAM-4 model grid points at the studied region.

Considerable differences are observed also in the spectral structure of changes in the global and regional air temperature. Investigated region is notable for a peculiar climate resulting both from some specific features of the atmospheric processes in the Atlantic Ocean, Arctic Ocean, and Siberia, as well as from the Large European Lakes (LEL) and White Sea effects on the drainage basins. The high percentage of coverage of the territory by surface waters, forests and wetlands influences the climate formation here. Total annual river runoff for the total territory of Karelia was re-calculated using measured runoff data (Shnitnikov 1966; Kuusisto 1992).

For the analysis the water balance equation was used,

where R is river runoff from the territory, P is precipitation, E is total evaporation, DW is change in storage into the basin, and e is total error of calculation of the water balance elements.

Fig. 2. The regional annual air temperature variations and 15-year moving average values for the period 1880-1999.

For relatively long time intervals, it is possible to assume DW = 0 and e = 0. Total annual river runoff R for the region was recalculated with using measured runoff data as mean weight values:

where indices designate annual runoff (in mm) for the main rivers Kovda (1), Kem (2), Nizhniy Vyg (3), Shuja (4), Vodla (5) and Suna (6), correspondingly. Total precipitation P for the studied territory was calculated using data from instrumental observations at four weather stations located in the region. Their comparison with the total precipitation in Finland obtained by R. heino (1994) shows the high correlation between these two series (r = 0.78). Total evaporation E for the territory of Karelia was calculated with use of the well-known Oldekop's formula:

where E_o is potential rate of evaporation (or evaporativity), in mm. The 15-year moving averages of E_o were calculated according to the regional functional dependence, obtained in Northern Water Problems institute:

where T is air temperature, mean for the territory of North-West of Russia, ^oC (15-year moving averages). The analysis of the water balance elements for the studied area was carried out in accordance with the formula

Time series of the basic water balance elements of the territory of North-West of Russia for the period 1880-1999 (15-year moving averages and their linear trends) are shown in Fig. 3.

A comparative analysis of observed data and parameters was calculated with use of the Max-Planck Institute for Meteorology model (ECHAM-4) for the period 1960-1999. Possible climate and water balance changes in the studied region were evaluated for the period 2000-2050 (two scenarios of CO₂ changes: with/without aerosol).

Fig. 3 Time series mean as a whole for the territory of North-West of Russia precipitation (1), total evaporation (2) and river runoff (3). All elements are the 15-years moving average values.

Climate and water balance elements variability and water level fluctuations for the period of instrumental observations

Regional changes in temperature anomalies seem, in the first approximation, to be directly proportional to the increase of the mean global air temperature. Against the background of global warming, regional climate changes in the regions of exhibited variations are related to the variability of parameters on both interannual scales. The analysis of general tendencies in the long-term meteorological and hydrological time series in north-western Russia from the 1880's up to now shows the presence of positive linear trends for annual air temperature (T), precipitation (P), total evaporation (E) and river runoff (R) for all large (more than few thousand square kilometers) drainage basins. High values of linear correlation between basic climate and water balance parameters were obtained. From this estimate it can be concluded that there are high causal relationships between basic parameters of the water balance of the territory and, therefore, changes in any of those will lead to changes in others.

A comparative analysis shows that on the north of the studied area warming becomes apparently more visible. For instance, during the 1880-1999 period in Murmansk district (Kola Peninsula) (69^oN) a linear trend of annual air temperature is equal to +0.8^oC/100 years, whereas in Petrozavodsk-Karelia (62^oN) for the same period it is only to +0.17^oC/100 years. The analysis of seasonal air temperature trends for weather stations located in North-Western Russia shows that there are few zones for each season with positive and negative linear tendency (Fig. 4).

Only spring temperatures have positive trends all over the studied area with the largest values up to +3.5^oC/50 years. For other seasons, zones of positive trends are located mainly in the southern part of region close the largest lakes of Europe - Ladoga and Onega. The autocorrelation and spectral analysis of the time series demonstrates presence of low and high frequency quasi-periodical oscillations with different time scales. Spectra of the time series contain the most visible quasi-periodical components with duration approximately 20, 6-8, and 2-3 years. As it was shown by filatov (1997) the inter-annual fluctuations of the water level in the largest lakes of Europe (Ladoga, Onega and Saima) also contain the similar quasi-periodical components with similar time scales. The detailed analysis of seasonal temperatures recorded in the period of instrumental observations indicates that although the autumn and winter seasons in the second half of the 20th century became colder, positive trends in the spring temperatures resulted in the general increase of mean annual air temperatures. Obtained from 128-year-long series segments, non-stationary spectra S (w ,t) demonstrate that the share of variance occurring in the high frequency range did not change throughout the period of observations, whereas the relatively short-term (10- and 30-year) fluctuations prove amenable for energy re-distribution in time.

Fig. 4. Distribution of zones with positive (+) and negative (-) trends of the seasonal air temperatures in the territory of NWR for the period 1951-1999 (1 stands for winter, 2 for spring, 3 for summer, and 4 for autumn). All numeric values are given in ^oC/50 years.

A rather short length of the hydrometeorological time series and features of spatial distribution of weather stations allow making the analysis only for processes and phenomena restricted by relatively short time-space scales. In findings of investigations of tree-ring time series obtained on more than 450 dendro-chronological samples of the Eastern Fennoscandia including taiga sub-zones of the Kola Peninsula, Karelia, Vologda and Arkhangelsk regions, the significant cyclic 30-40 and 13-year quasi-periodical oscillations were calculated. Considerable climate changes in the region in those years are manifested also through a shorter period of snow cover of the catchment areas and a longer ice-free period on the lakes. The average duration of the ice-free period for Lake Ladoga area varies from 103 to 181 days, mainly due to the effect of latitude and local conditions. The characteristics of the ice cover on lakes and climatic parameters are in the rather complicated correlation, e.g., ice thickness, and the time of ice break-up depending on a wealth of factors. In the case of Lake Ladoga and Lake Onega, the ice-free period coincides with changes of the air temperature, i.e. there is a certain correlation between the water temperature upward tendency and the ice-covered area decrease.

In spring, ice melting proceeds along the south-north axis Water in shallower areas warms up and cools down faster, hence freezing and ice break-up occur earlier in these areas Regarding the ice cover on Lake Onega, from 1880 to the present, the number of ice-free days has increased from 217 to 225.

At the same time, data presently available are not sufficient to identify the climate-induced changes in lake ecosystems. However, they prove that the main cause of lacustrine ecosystem changes is affected by the anthropogenic factors (Current status 1987, Filatov 1997).

References can be found here.