NEACC

  • Increase font size
  • Default font size
  • Decrease font size
Research Research Climate change projections for Russia and Central Asia States

Climate change projections for Russia and Central Asia States

Print

CLIMATE CHANGE PROJECTIONS AND IMPACTS IN RUSSIAN FEDERATION AND CENTRAL ASIA STATES
Lead Author: Vladimir Kattsov
Contributing Authors: Veronika Govorkova, Valentin Meleshko, Tatyana Pavlova, Igor ShkolnikContents

Executive summary
1. Introduction
2. Constructing climate change scenarios for RF and CAS
2.1 Regions under consideration and specific features of their climates
2.2 State-of-the-art climate models
2.3 Climatological baseline and time slices for the 21st century
2.4 Sources of projection uncertaintiesf
3. RF/CAS climate change projections for the 21st century
3.1 Surface air temperaturef
3.2 Precipitation
3.3 Runoff
3.4 Sea level pressure and wind
3.5 Cryosphere
4. Concluding remarks
References

Executive summary

Increased levels of atmospheric greenhouse gases (GHG) will have a larger effect on climate in Northern Eurasia, particularly in its arctic and subarctic regions, than in most of other regions of the Earth. To estimate a possible future climate change over the territories of Russian Federation (RF) and five Central Asian States (CAS: Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan, and Uzbekistan) ensembles of up to 16 comprehensive global (coupled atmosphere-ocean) climate models are used in this Assessment. Additional estimates are obtained using a regional climate model customized to two major domains in the Northern Eurasia. Future concentrations of GHG and aerosol emissions can be estimated assuming demographic, socio-economic and technological changes through the 21st century. Within IPCC, a set of emission scenarios has been prepared; in this Assessment so called A2, A1B, and B1 scenarios are considered. These are, correspondingly in the upper, middle and lower parts of the range of scenarios provided by the IPCC.

Projections with the global models provide a physically consistent quantitative picture of climate change through the 21st century. The projected changes in RF and CAS in many cases continue the tendencies already observed, while increase of their rates as well as inter-scenario differences are accelerating by the end of the 21st century. Relative to the base line climate (1980-1999), the area averaged annual mean warming for the RF and CAS by 2011-2030 is 1.0-1.2ºC. By the end of the 21st century, the increase of the area averaged annual mean temperature varies from 3.0±1.0ºC (B1) to 5.5±1.2ºC (A2) for RF, and from 2.6±0.7ºC (B1) to 4.7±0.9ºC (A2) for CAS. Area averaged winter warming is stronger in RF (from 3.8±1.3ºC (B1) to 7.2±1.5ºC (A2)), than in CAS (from 2.7±1.0ºC (B1) to 4.6±1.0ºC (A2)). In summer, on the contrary, CAS territory warms stronger (from 2.8±0.6ºC (B1) to 5.0±1.0ºC (A2)), than RF (2.3±0.9ºC (B1) to 4.2±1.3ºC (A2)).

The warming is accompanied by decreases in yearly number of frost days (with surface air temperature below 0ºC), in duration of extremely low winter temperatures, and in annual ranges of extreme temperatures. E.g., by mid-21st century in Central and East Siberia the decrease of the number of the frost days is 10-15 days, in European RF, Kazakhstan, Turkmenistan and Uzbekistan – 15-30 days, in Kyrgyzstan and Tajikistan, as well as Baltic region, – 30-35 days. An increase is projected in duration of summer temperature high extremes. The most severe heat waves are projected to occur in West Siberia and CAS.

Over the territories of both RF and CAS, the increase of winter precipitation is a robust feature of projections for all scenarios. In summer, however, the increase is projected only in the northern and eastern parts of RF, while South-West of RF and CAS demonstrate a decrease in precipitation and thus an increase in drought risk. By the end of the 21st century, over the territory of RF, the increase of the area averaged annual mean precipitation varies from 11.3±3.1% (B1) to 17.7±3.7% (A2). For CAS, projected annual mean changes usually do not exceed the intermodel standard deviation, with a few exclusions for the territory of Kazakhstan. In many regions along with decreasing mean precipitation an increase in very intensive precipitation is projected. However, credibility of precipitation projections is estimated to be lower than that of the temperature.

Runoff is projected to increase over the catchments of Siberian rivers, while over the southern watersheds runoff will decrease both due to precipitation decrease and evaporation increase in the warm season. Over the European RF the terrestrial snow cover is projected to decrease, while in Siberia, where the solid precipitation dominates, the snow mass accumulated during the cold season will be increasing. As a result, winter runoff is expected to increase in European RF. In Asian Russia, the combination of the increase of the snow mass accumulated during the winter and acceleration of its melting in spring results in increasing risks of flooding.

Degradation of permafrost in the warming climate will manifest itself in increasing seasonally thawing depths, and shifting northward the boundary between seasonal thawing and seasonal freezing of the grounds. Arctic sea ice is projected to be shrinking through the 21st century, with a faster disappearing old, multi-year ice. In A2 scenario, in late 21st century, some models project entire disappearance of sea ice in the Northern hemisphere by the end of summer.

RF/CAS territory is characterized by complex and still insufficiently understood climate processes and feedbacks, contributing to the challenge, which the region poses from the viewpoint of climate modelling. Local and regional climate features, such as enhanced precipitation or winds close to steep mountains, are not well represented in global climate models due to their limited horizontal resolution. Physically based methods for local and regional climate simulation rely on high resolution models run over limited time slices. Such methods can be used to interpret global simulations on finer scales and capture areas with intensified precipitation, extreme wind events etc. Unfortunately, only very limited high-resolution regional model results for the RF/CAS territory have been available for this Assessment.

Natural variability in the region under consideration is large and could mask or amplify a change due to human activities. This effect could be larger or smaller depending on the region, the climate parameter (temperature, precipitation, snow cover, etc.) and the time and space scales. To assess the relative importance of natural variability versus a prescribed climate forcing massive ensembles of differently formulated climate models run from an ensemble of initial conditions should be used. This technique requires substantial computing resources. Massive ensembles containing on the order of hundred simulations would give a better estimate of climate change probability distributions, including changes in the frequency of winter storms, temperature extremes, etc.

 

1. Introduction

To assess climate change impacts on societies, ecosystems, man-made infrastructure etc. possible changes in physical climate parameters must first be estimated. The physical climate change estimates must in turn be calculated from changes in external factors that can affect the physical climate. Examples of such factors are atmospheric composition, in particular atmospheric concentrations of greenhouse gases (GHG) and aerosols; land surface changes due to for instance man-made de-forestation; etc. Physically based climate models are used to obtain climate scenarios – plausible representations of the future climate that are consistent with assumptions about future emissions of GHG and other pollutants (emission scenarios) and with current understanding of the effects of increased atmospheric concentrations of these components on the climate. Correspondingly, by a climate change scenario, the difference is understood between a climate scenario and the current climate. Being dependent on sets of prior assumptions on future human activities, demographic and technological change, and their impact on the composition of the atmosphere, the climate (change) scenarios are not predictions, but just internally consistent descriptions of possible future climates.

If compared to other methods of constructing climate (change) scenarios, only comprehensive global coupled atmosphere-ocean general circulation models (AOGCMs), possibly in combination with dynamical or statistical downscaling methods, have the potential to provide geographically and physically consistent estimates of regional climate change due to increased GHG. AOGCM projections are available for a large number of climate variables, a variety of temporal scales, and in regular grid points all over the world, which should be sufficient for many impact studies. The representativeness of AOGCM-based scenarios is increased by employing an ensemble of different models.

The most current generation of AOGCM simulations and projections has recently been completed for the Intergovernmental Panel on Climate Change (IPCC) in order to provide input to the IPCC’s Fourth Assessment Report (AR4). This has been a major international effort within the framework of WMO’s World Climate Research Programme “Coupled Model Intercomparison Project”, phase 3 (CMIP3) [Meehl et al., 2007]. Numerous diagnostic subprojects aimed at analyses of a great variety of 20th and 21st century climate evolution aspects fed the IPCC AR4, and continue to feed region-focused climate assessments, like this one, devoted to projected climate change over the territory of Russian Federation and five Central Asia States. Some of the results presented in this Assessment were obtained within research projects of the Russian Federal Service for Hydrometeorology and Environmental Monitoring (Roshydromet), Russian Federal Targeted Research Programmes “R&D of the Russian science-technological complex for 2007-2012” and “World Ocean”, projects supported by the Russian Foundation for Basic Research and by the US National Science Foundation via the International Arctic Research Center of the University of Alaska Fairbanks.

Modeling groups are acknowledged for making their simulations available for analysis, the Program for Climate Model Diagnosis and Intercomparison (PCMDI) for collecting and archiving the CMIP3 model output, and the WCRP's Working Group on Coupled Modelling (WGCM) for organizing the model data analysis activity. The WCRP CMIP3 multi-model dataset is supported by the Office of Science, US Department of Energy.

 

2. Constructing climate change scenarios for RF and CAS

2.1 Regions under consideration and specific features of their climates The territory considered in this Assessment presents a tremendous variety of climatic zones – from the Arctic in the north to subtropics in the south, and from low lands to the north of the Caspian Sea to the high mountains in East Siberia, Kyrgyzstan and Tajikistan. Such a variety requires separate consideration when it comes to climate change projection analysis.

National and internal administrative boundaries within the entire region are also essential for this Assessment, particularly when it comes to adaptation options and strategies at national level. For the purposes of this report the entire territory of Russian Federation (RF) is divided into three major regions: European RF, West Siberia, and East Siberia with the Russian Far East. The boundaries between the regions coincide with those of some RF administrative regions. Kazakhstan, Kyrgyzstan, Uzbekistan, Tajikistan and Turkmenistan are considered in this Assessment either individually, or, in many cases, as a single region – Central Asia States (CAS).

There also exists a system of natural boundaries such as the boundaries of major river watersheds which has proven to be helpful when assessing future changes in the terrestrial hydrological regime.

 

RF

Different types of climate are observed over the huge territory of RF: arctic, subarctic, moderate and even subtropic. On the one hand, surface air temperatures below 0ºC are observed over the greater part of the territory, at least during half a year. On the other hand, droughts, dust storms and wild fires are typical for the south of RF.

RF is the world’s coldest country with its annual mean area averaged temperature of -4.1ºC. In winter, when solar radiation income is small, the climate in RF is determined by atmospheric circulation. In East Siberia, a persistent winter high causes formation of a cold pole with mean winter temperature around -42ºC, mean minima of -55ºC, the absolute minimum -67ºC, and 280 frost (<0ºC) days per year. The European RF is influenced by the warm Atlantic air, so that mean winter temperature varies from -2 – -4ºC in the south-west to -10 – -12ºC in the central part, achieving positive values in the Black Sea region. Mean summer temperatures vary from 4-5ºC over the arctic coast to 20-22ºC in Kalmyk and Astrakhan regions, where mean maxima achieve 36-38ºC, and absolute maxima are about 45ºC. The maximum continuous duration of the temperature above 25ºC is up to 4 days.

A specific feature of the climate in RF is large daily and annual amplitudes, especially in the deep continental Asian part of the country. In spring, daily amplitude can be as large as 15-17ºC in East Siberia and Transbaikal region (and up to 19ºC at Verkhoyansk and 22ºC at Oymyakon). Annual amplitudes vary from 8-10ºC at the west coast of the Barents Sea to 63ºC in East Siberia (Verkhoyansk ridge). The range of extreme temperatures here achieves 90ºC.

Over the past 100 years (1907-2006), the annual mean area averaged temperature over RF territory increased by 1.29ºfС with an acceleration over last three decades (0.43ºfС/decade). The greatest warming over the last three decades was observed in the RF North-West, southern West Siberia, and the North Caucasus. In Chukotka, a cooling was observed in winter, and warming in summer.

The most important feature of precipitation regime over the territory of RF is the large amount of solid precipitation. Annual sums of precipitation vary from 3200 mm at the Black Sea coast of the Caucasus to 150 mm at arctic islands and the South-East Altai. Winter monthly sums are less geographically variable making 20-40 mm, with the pronounced increase up to 60-100 mm at the Black Sea coast of the Caucasus and at Kamchatka peninsula. Summer monthly sums are much more spatially variable with less than 30 mm along the arctic coast of Asian RF and the Caspian Sea. In the South-East RF, and in mountainous regions of Caucasus, Altai, and Sayan, summer monthly sums are up to 100-140 mm. Daily precipitation maxima vary from 40-60 mm in the Arctic to 300-360 mm in the south-east and at the Black Sea coast of the Caucasus.

Centennial trends of annual precipitation are insignificant. Over the last three decades, some increase of precipitation over the RF territory was observed (7.2 mm/decade), with maxima in West and Central Siberia. The most pronounced precipitation increase over the RF territory was observed in spring (16.8 mm/decade).

The wind regime over the territory of RF is dominated by seasonality of baric centers over Eurasia and North America, as well as by the orography. Weak winds over a vast region in the continental North-East Asia are a result of the Siberian High persistence. Strong winds are observed in the coastal regions of the Russian Far East (up to 34 m/s) and in the Arctic (38 m/s). Due to specific orographic features, high wind speeds (up to 34-36 m/s) are also typical for the southern part of European RF, along the border with Kazakhstan, in Novosibirsk and Kemerovo regions.

From 50 to 75% of the RF rivers’ annual discharge is formed by thawing snow. Spring flooding is most typical for East Siberian rivers. Almost all rivers in RF are covered by ice in winter. Over the last three decades, the annual discharge of major Russian rivers, including the Volga and the great Siberian rivers, increased. Floods became more frequent in the RF South-East and the North Caucasus – due to heavy rains; as well as in East Siberia – due to increases of snow mass and acceleration of its melting in combination with ice jams in river deltas.

Long existence of terrestrial snow cover is another characteristic feature of the climate over the greater part of RF. The highest mean snow cover (90 cm) is observed westward from the Ural Mountains. In the central part of European RF the mean height of the snow cover is about 50 cm. In the south (Krasnodar and Stavropol regions), the height of snow cover decreases down to 20 cm. In the Caucasus the amount of snow increases, e.g. at the 2500 m height in Dagestan the thickness of snow cover achieves 1.5 meter. In the southern part of Kamchatka peninsula the height of the snow cover can exceed 2 m. In the second half of the 20th century, duration of snow cover existence increased over RF territory, excluding western regions of its European part.

Multiyear frozen grounds are another specific feature of RF climate, especially over vast areas in West and East Siberia. Altogether about 70% of the RF territory is covered by permafrost. Seasonally frozen grounds are observed down to 40-45ºN. In last decades the temperature of the upper layer of multi-year frozen grounds increased (0.1-0.68ºС/decade) with an increase of seasonal thawing depths in some regions.

The greater part of Russian marine Arctic is characterized by sea ice cover which decreases by the end of summer allowing for operation of so called Northern sea route which has a great potential for Russian economy. According to satellite data, the annual minimum area of sea ice in the Northern hemisphere (September) was decreasing by 72 thousand km2 per decade with an acceleration in the most recent years: the absolute minimum of the total sea ice area in the Northern Hemisphere (4.13 million km2) was observed in September 2007.

 

CAS

The openness of the CAS territory from the north and north-west is favourable for penetrating of cold arctic and continental air masses in winter. Winters in CAS are cold: in the northern regions, mean January temperature is -12 – -14ºC, and only in the south it reaches 0-2ºC. Summer at the plains and near the foothills is hot and dry, with the mean temperature contrasts less than in winterf: from 25-26ºC in the north up to above 30ºC in the southern deserts.

Over the flat part of the CAS, small quantity of precipitation falls in the end of winter and in spring, when the polar front approaches. The annual precipitation sum is 200 mm in the Kazakh small hills and is only 75-100 mm in the Karakum and Kyzylkum deserts. Moistening of mountainous regions is highly variable depending on the orography; the maximum sum of precipitation can reach 1600 mm/year.

Essential factors influencing the regional climate are the mountain massifs in the south, which create a natural barrier on the way of the moist air from the Atlantic. They catch a greater part of the moisture and condense it over glaciers. It is noteworthy that the area of mountain glaciers is decreasing with an acceleration unprecedented over past 12000 years [Solomatina, 1992].

Two large lakes are situated in CAS: the drying up Aral Sea with the drying up Syr Darya and Amu Darya rivers draining into it, and the half salty Balkhash Lake.

Observed 20th century changes in the surface air temperature and precipitation over CAS are discussed in [Gruza and Ran'kova, 2003]. During the period of 1951-2000, annual mean temperature over Middle Asia increased by 1.0-2.5ºC. In October-March the air temperature increased stronger (1.0-3.0ºC) than in April-September (0.5-1.5ºC).

The annual and cold season sums of precipitation decreased almost everywhere, but only by 2.5-5.0%. During the warm season, the sum of precipitation decreased over the northern part of the territory (by 5-10%) and increased in the south (by 2-5%).

 

2.2 State-of-the-art climate models

AOGCMs are widely acknowledged as the most sophisticated tools available for global climate simulations, and, particularly, for projecting future climate states. AOGCMs are made up of a number of component models (of the atmosphere, ocean, cryosphere, land surface, and biosphere), which are interactively coupled via exchange of data across the interfaces between them. AOGCMs are permanently developing climate projection tools.

Choosing an AOGCM output to be used in an impact study is not a trivial task, having in mind the variety of models, which continues to increase. While models do not necessarily improve with time, later versions (often, with higher resolution) are usually preferred to earlier ones. An important criterion of selecting an AOGCM to be used in constructing regional climate scenarios is its validity evaluated through analyses of the AOGCM performance in simulating the present-day and past climates (particularly, the 20th century climate) presuming the corresponding level of credibility of the future climate projections. The validation is made by comparing the model output against observed climate and with other models in the region of interest, and also over larger scales, to determine the ability of the model to simulate large-scale circulation patterns. Well established systematic comparisons of this type are provided by international Model Intercomparison Projects (MIPs), particularly the above mentioned CMIP3.

The state-of-the-art climate models described in this Assessment (hereafter “CMIP3 models”) relate to the early 2000s and is that evaluated by IPCC AR4 [Randall et al., 2007], particularly in comparison with the previous model generation. The dynamical cores of the CMIP3 models including both numerics and spatial resolution have improved. More processes have been incorporated, and parameterizations of physical processes have become more comprehensive. These improvements, while not always straightforwardly identifiable in the output, have resulted in improvements of AOGCM simulation of many aspects of the Earth climate system. Among evidences of these improvements is the fact that most of the current AOGCMs no longer use flux adjustments to reduce the climate drift. Nevertheless, AOGCMs still show significant biases and inter-model scatters both in simulating observed and projecting future climates, particularly at the regional scale. This is only partly a result of limitations of computing power (the highest-resolution models do not obviously outperform others), but also a result of insufficient scientific understanding.

The 16 AOGCMs whose outputs for the 21st century were analyzed in this

Assessment are listed in Tab.1. This is a subset of the 23 CMIP3 models participated in CMIP3. Initially, the whole set of AOGCMs whose outputs were available from the CMIP3 archive were analyzed. Later on, this set was short-listed down to 16 AOGCMs, mainly for the reasons of the models’ ability to reproduce current climate of the North Eurasia (see for details [Govorkova et al., 2008; Kattsov et al., 2007a,b; Meleshko et al., 2007, 2008a; Pavlova et al., 2007; Randall et al., 2007; Roshydromet, 2008]).

Some scenario simulations for the 21st century were not available for some CMIP3 models or for some variables, thus different sub-ensembles of the models are used in obtaining the 21st century estimates discussed in this Assessment.

Table 1. A subset of CMIP3 models used in this Assessment. (For details and references see [Randall et al., 2007].)

While the resolution of AOGCMs used in projections of future climate is rapidly improving, it is still insufficient to capture the fine-scale structure of climatic variables in many regions of the world that is necessary for impact assessment studies. Hence, a number of techniques exist to enhance the resolution of AOGCMs’ outputs. One of these techniques is the use of high resolution regional (or limited area) climate models (RCMs) restricted to a domain with simple lateral boundaries, at which they are driven by outputs from GCMs or wider RCMs. On time scales of a few years to decades and beyond, RCMs have shown their strength in comparison with coarser resolution global models, as they are capable of capturing fine scale details of climatic processes – such as the presence of complex topography and small scale weather features like tropical cyclones and even polar lows – much more realistically than global models. Of course, RCM projections are limited by the skill of the global model projections used in the lateral boundary conditions. In this connection, a potential problem in the mismatch between scales in the driving coarse-resolution model and the high-resolution regional model has often been mentioned.

RCMs have been used for a wide variety of research worldwide, generating a sizeable research community. However, only very few groups have focused on RF/CAS so far, although several new initiatives have recently been taken. Atmospheric regional modelling systems developed at Voeikov Main Geophysical Observatory (MGO RCM) [Shkolnik et al., 2006, 2007] are applied to the two partially overlapping regions: centred in the European RF and in Siberia. MGO RCM projections combined for the two domains are considered in this Assessment.

 

2.3 Climatological baseline and time slices for the 21st century

Climatological baseline is a period of years representing the current climate, the latter being understood as a statistical description in terms of the mean and variability over the period. To satisfy widely adopted IPCC criteria, a baseline period should:
- be representative of the present-day or recent average climate in the region considered;
- be of sufficient duration to encompass a range of climatic variations;
- cover a period for which data on all major climatological variables are abundant, adequately distributed over space and readily available;
- include data of sufficiently high quality for use in evaluating impacts;
- be consistent or readily comparable with baseline climatologies used in other impact assessments.

Until recently, the most widely used baseline period has been the WMO-defined “classical” 30-year period. Usually it is 1961-1990 (particularly, in the first three IPCC Assessment Reports). In some cases, an earlier period 1951-1980 was also used. The 20-year period 1981-2000 has been selected as the ACIA [Kattsov and Kallen, 2005] climatological baseline. The IPCC AR4 mainly used the new climatological baseline – 1980-1999, which was not available for the previous three IPCC Assessments.
In this Assessment, in line with the IPCC AR4, the 20-year baseline period 1980-1999 is used.

For the 21st century time slices centered at 2020, 2050 and 2090 are used. For AOGCMs, these are usually the 20-year periods: 2011-2030, 2041-2060, and 2080-2099, correspondingly.

For RCM projections considered in this Assessment, shorter – 10-year – time slices centered at 2050 and 2090 are only available.

 

2.4 Sources of projection uncertainties

Different assumptions about the future social and economic development introduce one of the major uncertainties in the climate change scenarios, e.g., emission scenarios presented in the IPCC Special Report on Emissions Scenarios (SRES) [Nakicenovic et al., 2000]. Emission scenarios are plausible representations of the future development of emissions of radiatively active substances (GHG, aerosols), based on a coherent and internally consistent set of assumptions about demographic, socio-economic, and technological changes and their key relationships in the future. Therefore climate projections should be distinguished from climate predictions – much stronger statements about climate behaviour in the future, which implicitly presumes the prediction of the social and economic development. The SRES emissions scenarios were built around four narrative storylines that describe the evolution of the world in the 21st century. Altogether, 40 different emission scenarios were constructed. Six of these (A1B, A1T, A1FI, A2, B1 and B2) were chosen by the IPCC as illustrative marker scenarios. The SRES scenarios include no additional climate initiatives, which means that no scenarios are included that explicitly assume the implementation of the United Nations Framework Convention on the Climate Change or the emission targets of the Kyoto Protocol. No probabilities are assigned to the various SRES scenarios. Three IPCC SRES scenarios for the 21st century were considered in this Assessment: A2, A1B, and B1. By the end of the 21st century, the A2 scenario has the greatest increase of greenhouse gas concentrations (to about 825 ppm CO2) and suggests the strongest global warming, while the B1 is the least extreme of the three (about 550 ppm CO2 by 2100) (see for details [Nakicenovic et al., 2000]).

Emission scenarios are converted into concentration scenarios which are used as input into climate models to compute climate projections. There is also uncertainty associated with this conversion. Additional uncertainty is related to the calculation of the radiative forcing associated with given concentrations, which is done implicitly within AOGCMs but is problematic in particular for aerosols.

Due to different representations of processes and feedbacks in the climate system (see, e.g., [Bony al., 2006]), or by missing some of them, AOGCMs differ in their sensitivity to the same radiative forcing. Sometimes, the difference in AOGCM sensitivity manifests itself only in regional climate change patterns, the magnitudes of global mean changes remaining close to each other. At long leads into the future, the forcing uncertainty (differences between emission scenarios) and model uncertainty are of approximately equal importance.

Insufficient resolution of AOGCMs prevents their outputs from being directly used in impact assessments. In most cases, a climate scenario for a certain region requires a combination of the simulated variables and observed data. Combination of AOGCM outputs and observations may be done by different methods. Additionally, observational data often don’t capture the full range of decadal-scale natural variability. Finally, gridding the observational data – in order to create baseline climatologies – introduces errors. Employing RCMs to enhance spatial and temporal resolution of AOGCM outputs is associated with further uncertainties arising from individual features of the RCMs.

In addition to anthropogenic forcing, climate changes in the real world are affected by largely unpredictable natural variability. Part of the natural variability is thought to be due to variations in solar and volcanic activity, but a substantial part of it is unforced, resulting from the internal dynamics of the climate system. Climate models also simulate unforced natural variability, so that, when the same model is run with the same forcing scenario but different initial conditions, there are non-negligible differences in the results, in particular in regional details that are affected by internal variability much more than global means. When different models are run using the same forcing scenario, part but not all of the differences in their results arise from different realizations of natural variability.

 

3. RF/CAS climate change scenarios for the 21st century

3.1 Surface air temperature

Through the 21st century, surface air temperature change is projected to increase over the entire territory of RF/CAS in all scenarios, with the maximum warming in winter. In the cold season the mean warming is increasing from south to north and reaches its maxima in the Arctic (Fig.1). On the contrary, in summer, the high-latitude warming is the weakest, because the temperature over the Arctic Ocean is kept at the sea-ice melting point. The continental part of Northern Eurasia shows the strongest warming.

Within next several decades, the annual mean changes of the temperature do not differ significantly neither between the regions, nor between the scenarios, and remain well within the inter-model scatter measured as the standard deviation (Tab.2). The area averaged annual warming for the RF and CAS by 2011-2030 is 1.0-1.2ºC. With further increase of the temperature, in the second half of the 21st century, both the difference between the scenarios (Fig.2) and the inter-model scatter start to increase. By the end of the 21st century the increase of the area averaged annual mean temperature varies from 3.0 ±1.0ºC (B1) to 5.5±1.2ºC (A2) for RF, and from 2.6 ±0.7ºC (B1) to 4.7±0.9ºC (A2) for CAS.

By the end of the 21st century, area averaged winter warming is stronger in RF (from 3.8 ±1.3ºC (B1) to 7.2±1.5ºC (A2)), than in CAS (from 2.7 ±1.0ºC (B1) to 4.6±1.0ºC (A2)) (Tab.3). In summer, on the contrary, CAS warms stronger (from 2.8 ±0.6ºC (B1) to 5.0±1.0ºC (A2)), than RF (2.3 ±0.9ºC (B1) to 4.2±1.3ºC (A2)).

Future changes of the temperature extremes were estimated using the MGO RCM [Shkolnik et al., 2006; 2007] (Figs. 3,8,9) and an ensemble of 9 CMIP3 models (Figs.4-7). By mid-21st century temperature yearly minima will increase over the entire territory of RF and CAS. The strongest increase (by 4-6ºC) is projected in the South and North-West of European RF. In central European RF, the Urals, and in East Siberia the increase of yearly minima by mid-21st century is 2-4ºC. Unlike the increase in the annual mean temperature, the projected increase of yearly minima in Siberia and Russian Far North are relatively small.

Analyses of variability of daily mean and daily minimum temperatures using climate models show [Hegerl et al., 2004; Shkolnik et al., 2006; Kharin et al., 2007], that in the middle and high latitudes the contribution of lower percentiles (close to yearly minima) into the increase of winter temperatures is the greatest – much greater than those of higher percentiles. The asymmetry of the probability distributions of winter temperatures is increasing, while the inter-quartile range and extremity of the winter thermal regimes are decreasing accordingly.

Table 2. Annual mean surface air temperature changes (ºС) by early, mid- and late 21st century for B1 (14 models), A1B (15 models), and A2 (16 models) scenarios. Standard deviations given as lower indices provide a measure of inter-model scatter.

Table 3a. Seasonal surface air temperature changes (ºС) by early, mid- and late 21st century for B1 scenario (14 models). Standard deviations given as lower indices provide a measure of inter-model scatter.

Table 3b. Seasonal surface air temperature changes (ºС) by early, mid- and late 21st century for A1B scenario (15 models). Standard deviations given as lower indices provide a measure of inter-model scatter.

Table 3c. Seasonal surface air temperature changes (ºС) by early, mid- and late 21st century for A2 scenario (16 models). Standard deviations given as lower indices provide a measure of inter-model scatter.



Figure 1. Surface air temperature changes (ºС) by early-(top), mid-(middle) and late (bottom) 21st century in winter (left) and summer (right) for A2 scenario.

Figure 2. Surface air temperature change difference (ºfС) between the “strong” A2 and “weak” B1 scenarios by early-(top), mid-(middle) and late (bottom) 21st century in winter (left) and summer (right).

Figure 3. Change of daily mean temperature interquartile range in winter (ºC) by mid 21st century (A2 scenario) as simulated by the MGO RCM for the two domains: European RF and Siberia. The changes shown indicate at decrease of temperature variability and therefore decrease in cold extremes. A coherence between changes in snow cover duration (Fig.16) and temperature variability imply the most pronounced changes in variability occur in the regions where snow completely retreats causing temperature increase. Areas beyond the modeling domains are left in white.

The increase of the yearly highest daily maxima of surface air temperature over the entire territory of RF and CAS is projected to be weaker than that of the yearly minima. By mid-21st century, over the greater part of the territory the former does not exceed 3ºС.

By mid-21st century the yearly amplitude of the extreme temperatures decreases over the entire territory (Fig.4).

Figure 4. Changes (ºC) of the annual extreme temperature range (ETR, difference between highest and lowest temperatures in a year) as simulated by mid 21st century relative to the late 20th century. The stippling indicates areas where at least 2/3 of the models agree on the sign of the change. The negative values indicate at decrease of the ETR largely due to significant increase of the lowest temperatures in winter. The changes are most pronounced over the European RF, West Siberia, and CAS where satisfactory intermodel agreement on sign of changes is found.

By mid-21st century the duration of extremely low winter temperatures decreases by 6-8 days in Russian North-West and Far North. A pronounced decrease of the number of yearly cold extremes is expected along the Pacific coast, accompanied with the decrease of number of days with temperatures below 0ºC (Fig.5). In Central and East Siberia the decrease of the number of the frost days is the smallest – by 10-15 days, in European RF, Kazakhstan, Turkmenistan and Uzbekistan – 15-30 days. Kyrgyzstan and Tajikistan, as well as Baltic region, demonstrate most dramatic decrease of the frost days – by 30-35, which may have important implications for the mountain glaciers in the region.
f
Figure 5. The changes of annual number of frost days (AFD, number of days in a year with Tmin<0ºC) as simulated by mid 21st century relative to the late 20th century. The decrease of the AFD is most pronounced over the Baltic countries and mountainous areas in CAS. This indicates at greater snow/glaciers melting rate in the upcoming decades and increasing risk of avalanches/mud-torrents events.

An increase is projected of the number of days with daily maximum summer temperatures above 90th percentile in the baseline climate. By mid-21st century, the number of days with extremely high temperatures will increase in Russian Far North (5-10 days) and in the Black sea region (10-20 days). In the Northern Caucasus the duration of extremely hot temperature will increase by 20 and more days per year. At the same time the number of days with daily maxima above the 90th percentile is projected to increase over greater part of Siberia – by 2-4 days, Russian North-West, Central RF and the Northern Caucasus – by 3-5 days.

The above projections show reasonable agreement with the tendencies observed since the second half of the 20th century [Frich et al., 2002; Alexander et al., 2006].

Figure 6. Relative change (%) of annual mean (left) and maximum (right) durations of sequential days with daily Tmax exceeding 25ºC threshold by the mid 21st century as compared against respective exceedances in the late 20th century. Shown are changes for exceedances greater than 2 days. All durations tend to increase throughout all regions, however, the extremely long periods exhibit larger increase as compared to their annual mean values. The differences between changes in durations in (left) and (right) are most significant in midlatitudes where mean and maximum durations are shorter than 10 and 25 days, respectively.

Figure 7. Shift of the last frost date to the beginning of the year (left) and shift of the first frost date to the end of the year (right). Units are days.

Figure 8. As in Fig.3 but for changes of 5th (left) and 95th (right) percentiles (ºC) of daily mean temperature distributions in summer. The 5th percentile is centered at the lowest daily temperatures interval in summer, the 95th is centered at the extremely high temperatures. The changes reveal unequal magnitudes across percentiles: the extremely high temperatures will increase at a faster rate as compared against rest of the temperature distributions including very low maximum daily temperatures. The feature is most pronounced over CAS and Transbaikal region. This implies the summer temperature distributions will widen in a warmer climate causing more warm extremes. The changes by the mid-21st century are generally within the range of 40-50% of the changes shown.

Figure 9. As in Fig.3 but for heat wave durations (sequences of days with daily Tmax exceeding local 90th percentile of summer Tmax distributions) by the end of 21st century. The most severe heat waves are expected to occur in the West Siberia and CAS while heat waves in the East Siberia undergo minor changes. Despite of pronounced increase of the both mean and extreme temperatures the heat wave durations in Transbaikal region show only moderate increase due to increased daily temperature variability in the area.

 

3.2 Precipitation

The northern RF, especially its arctic regions, are the areas where the percentage increase in projected precipitation is largest (Fig.10). Like with the surface air temperature, differences between the three scenarios are comparatively small and do not exceed the inter-model scatter in the first half of the 21st century (Tab.4). Towards the end of the 21st century the differences between the scenarios increase (Fig.11). The general increase in high-latitude precipitation with global warming is a robust and qualitatively well-understood result from climate change experiments. With increasing temperature, the atmospheric circulation’s transport of moisture from lower to higher latitudes increases, leading to an increase in precipitation in the polar areas where the local evaporation is relatively small.

By the end of the 21st century, over the territory of RF, the increase of the area averaged annual mean precipitation varies from 11.3±3.1% (B1) to 17.7±3.7% (A2). For CAS, projected annual mean changes usually do not exceed the intermodel standard deviation, with a few exclusions for the territory of Kazakhstan.

While precipitation averaged over the entire territory of RF is projected to increase in all seasons (Tab.5), especially in winter, such behavior is not the case for some of its regions in summer, especially in the South-West. In such regions model projections vary from disagreement in sign to a robust decrease in total precipitation. Some regions demonstrate a non-monotonous evolution of precipitation starting from its increase in the beginning of the 21st century, and decreasing afterwards. For example, in the Ob basin, about a half of the models in the A2 simulations project a decrease in summer precipitation by the end of the 21st century. CAS demonstrate clear tendency to decrease in summer precipitation through and especially towards the end of the 21st century.

On the contrary, the intensity of heavy precipitation exceeding 90th summer percentile increases almost everywhere including the Caucasus, North Kazakhstan, Ukraine and South of RF, but not in at least three of the five CAS (Fig.12). The frequency of such extreme precipitation cases tends to decrease along with total amount of precipitating water, however, at significantly higher rate. This implies the daily precipitation accumulations will be more rare, but more intense. However, credibility of precipitation projections is estimated to be lower than that of the temperature.

 

3.3 Runoff

Through the 21st century, annual mean precipitation less evaporation (P-E) over the Arctic Ocean terrestrial watersheds (including the catchments of the great Siberian rivers), and accordingly the runoff, increase in all scenarios [Kattsov et al., 2007a]. The strongest relative increase in the river discharge is projected for the Lena (33% by the end of the 21st century in the A2 scenario); the weakest relative change is shown by the Ob with its 14% for A2 (compared e.g. to the Lena’s 20% for the “weak” B1 scenario). The projected increase of the total freshwater discharge into the Arctic Ocean by the end of the 21st century ranges from about 14% in B1 to more than 25% in A2.

Southern regions of RF, especially in European RF, and greater part of CAS show negative trends in P-E (Fig.13). An essential uncertainty, particularly over CAS territory is introduced by the inter-model scatter of the projections. However, for the greater part of RF at least 2/3 of the models agree in the sign (positive) of the projected changes.

In the warming climate the local winter runoff shows tendency to increasing, especially in the West RF, while the most pronounced changes are projected for the spring – almost everywhere (Fig.14).

Table 4. Annual precipitation changes (%) by early, mid- and late 21st century for B1 (14 models), A1B (15 models), and A2 (16 models) scenarios. Standard deviations given as lower indices provide a measure of inter-model scatter. Estimates are highlighted (black and bold) for which the change exceeds the standard deviation.

Table 5a. Seasonal precipitation changes (%) by early, mid- and late 21st century for B1 scenario (14 models). Estimates are highlighted (black and bold) for which the change exceeds the inter-model standard deviation.

Table 5b. Seasonal precipitation changes (%) by early, mid- and late 21st century for A1B scenario (15 models). Estimates are highlighted (black and bold) for which the change exceeds the inter-model standard deviation.

Table 5c. Seasonal precipitation changes (%) by early, mid- and late 21st century for A2 scenario (16 models). Estimates are highlighted (black and bold) for which the change exceeds the inter-model standard deviation.



Figure 10. Precipitation changes (%) by early-(top), mid-(middle) and late (bottom) 21st century in winter (left) and summer (right) for A2 emission scenario. The stippling indicates areas where at least 2/3 of the models agree on the sign of the change.

Figure 11. Precipitation change difference (%) between the “strong” A2 and “weak” B1 scenarios by early-(top), mid-(middle) and late (bottom) 21st century in winter (left) and summer (right).

Figure 12. Change in intensity of daily precipitation exceeding 90th summer percentile threshold. The stippling indicates areas where at least 2/3 of the models agree on the sign of the change.

The vast mountainous area in the South of CAS territory is the main regional source of water resources. Spatial resolution of the state-of-the-art AOGCMs and difficulties in representing physical processes in the regions with complex orography prevent from proper quantification of the future changes in water resources in the region. However, even qualitative estimates (the sign) of hydrological regime component changes are important. Below future water budget changes are considered over the mountainous region extending from South-West (62ºE) to North-East (87ºE). The entire area of the region is 1.13 million km2 (119 cells of the 1°x1° latitude-longitude grid).

Table 6. Annual mean area averaged changes (A2 scenario) in the surface air temperature (TAS, ºC), total precipitation (PR, %), solid precipitation (PRSN, %), evaporation (E, %), runoff (P-E, %) over the CAS mountainous region through the 21st century as simulated by the 16 AOGCM ensemble. Standard deviations given as lower indices provide a measure of inter-model scatter. Estimates are highlighted (black and bold) for which the change exceeds the standard deviation.

While total precipitation and evaporation do not show significant changes through the 21st century (Tab.6), there are some indications of precipitation decrease and evaporation increase. As a result the decrease of annual runoff stands out above the projection standard deviation by the end of the 21st century. According to model calculations, in the baseline period (1980-1999) snowfall in the region comprises 50% of the total precipitation. Through the 21st century solid precipitation decreases, allowing to expect less accumulation of snow over the cold period and its faster melting out in the warming climate.

Precipitation seasonality changes are most pronounced in the winter increase and summer decrease. The combined effect of the winter precipitation increase and snow melt intensification is the increase of winter runoff (Tab.7). On the other hand, the rest of the year runoff shows tendencies to decrease.

The above estimates for the CAS mountainous region should be used with caution having the large inter-model scatter of the projections.

Table 7. Changes (%, A2 scenario) in seasonality of the local runoff over the CAS mountainous region through the 21st century as simulated by the 10 AOGCM ensemble. Standard deviations given as lower indices provide a measure of inter-model scatter. Estimates are highlighted (black and bold) for which the change exceeds the standard deviation.

Soil water content is spatially highly variable climate parameter, whose simulation is determined not only by model-dependent physical parameterizations of processes in the atmosphere and the active layer of soil, but also by prescribed characteristics of land surface and subsurface which are not unified in the state-of-the-art AOGCMs. While this prevents from proper quantification of the projected changes in the soil water content in the warming climate, it can be concluded that in the 21st century, southern regions of RF, as well as CAS will experience an increasing risk of droughts. Drought conditions show a tendency of earlier formation and longer duration.

Aspecific integrator of climate process over a territory of about 3.5x106 km2 is the Caspian Sea whose level has demonstrated significant interannual variability over the past century. The variability is dominated by the two factors: the influx of the water from the terrestrial watershed of the sea, including the catchments of the Volga and the Ural Rivers, and visible evaporation from the sea surface which is the difference between the actual evaporation and precipitation. Those AOGCMs whose spatial resolution allows to explicitly reproduce heat and moisture exchange processes at the surface of the Caspian Sea can be used in projecting future changes of the sea level (Fig.15). Both the annual mean precipitation over the Volga catchment and evaporation from the surface of the Caspian Sea are projected to increase through the 21st century.

However, results of different studies diverge significantly on the projections of the Caspian Sea level, and the question on how the global warming will influence its variations remains open.

Figure 13. Annual mean precipitation less evaporation (P-E) changes (%) by early (top), mid- (middle) and late (bottom) 21st century for A2 scenario. The stippling indicates areas where at least 2/3 of the models agree on the sign of the change.

Figure 14. Seasonality of local runoff (%) (1) in the baseline period (1980-1999) and (2) by mid-21st century (scenario A2): the Ob (upper panel) and the Lena (lower panel) as related to the mean runoff in the base line period. Ranges show the inter-model standard deviations (13 models). [Roshydromet, 2008]

Figure 15. Evolution of annual mean water budget components over the Volga river catchnment and the Caspian Sea surface as simulated by 7 CMIP3 AOGCMs and smoothed with a 10-year running averaging (A2 scenario). Shown are: over the Volga catchment (a) precipitation (mm/day); (b) integral runoff (km3/year); over the Caspian sfea (c) visible evaporation (mm/day); (d) sea level change (cm/year). Shaded are inter-model scatters. Thick curves are ensemble means. Two horizontal lines mark the interval which, for the unchanged statistical structure of the baseline climate, must theoretically contain 95% of the decadal mean values. Thus, if the curve is outside this interval, the changes in the ensemble mean are statistically significant at the 5% level. The following CMIP3 AOGCMs are used in the calculations (in the brackets, the number of the Caspian sea grid cells is indicated for each model): CSIRO-Mk3.0 (15), ECHAM5/MPI-OM (11), MIROC3.2 (medres) (8), CNRM-CM3 (6), UKMO-HadCM3 (4), CGCM3.1(T47) (3), INM-CM3.0 (2). [Roshydromet, 2008]

3.4 Sea level pressure and wind An analysis of projections of sea level pressure interannual variability [Meleshko et al., 2008b] shows that it will increase in winter in the region eastward from the southern Urals and northern Kazakhstan. At the same time, the interannual variability of sea level pressure in the northern part of Central and East Siberia will decrease. As for variations of daily mean sea level pressure and wind speeds, the inter-model scatter prevents from quantifying changes of the probability distribution functions – anywhere over the territory under consideration.

3.5 Cryosphere Over the territory of European RF the snow cover (and snow mass) and periods of its existence will decrease (Fig.16). In Siberia, on the contrary, snow mass accumulation during the cold period will be increasing, especially in the East. In spring the greater mass of snow will be melting faster than in the baseline climate (Tab.8), thus increasing the risk of flooding, particularly in the Lena river catchment (Fig.17).

Figure 16. Decrease of snow cover duration (days) by mid 21st century (A2 scenario) as simulated by the MGO RCM for the two domains: European RF and Siberia. [Roshydromet, 2008].

Table 8. Annual mean snow melt relative to the base line climate (%) by early, mid- and late 21st century for A2 scenario (10 models). Standard deviations given as lower indices provide a measure of inter-model scatter.



Figure 17. Evolution of anomalies (relative to 1910-1959) in March (left) and May (right) of the integral snow mass (kg) through the 20th and 21st centuries (A2 scenario) as simulated/projected by an ensemble of CMIP3 models over the catchments of (a) the Ob, (b) the Yenissey, (c) the Lena). The horizontal lines indicate the 5% significance level for the changes. The figure demonstrates the qualitative differences in the snow melting out between the catchments: unlike the Ob, both the Yenissey and the Lena show an accelerating faster spring melting of the greater snow mass through the 21st century which implies an increase of flooding risk. This is already confirmed by observations particularly for the Lena with its notorious flooding events in the past.

Degradation of permafrost in the warming climate will manifest itself in increasing seasonally thawing depths, and northward shifting of the boundary between seasonal thawing and seasonal freezing of the grounds (Fig.18).

Figure 18. Multi-year frozen ground changes by mid-21st century projected by the MGO permafrost model [Malevsky-Malevich et al., 2001] driven by with outputs from the CMIP3 AOGCMs as upper boundary conditions: regions: (1) seasonal thawing, (2) seasonal freezing, and (3) transition from the regime of seasonal thawing to that of seasonal freezing in the upper 3-meter layer. Contours show an increase of thawing depths (cm) relative to 1980-1999; (4) the simulated current boundary of permafrost defined as the position of zero-degree isotherm at the 3-meter depth; (5) an approximate observed current position of the permafrost boundary. [Roshydromet, 2008].

Projections of the Arctic Ocean ice cover show accelerating decreases in its area and mass (particularly, seasonal minima) through the 21st century. The largest reductions by the end of the 21st century occur in the case of scenarios A1B and A2. The annual mean ice area reduction in the 21st century is on average about 3.5% per decade and less than 2.5% per decade in scenario B1 [Zhang and Walsh, 2006]. The multiyear ice coverage in the Arctic decreases more rapidly than the seasonal ice area. In most model simulations, the extent and duration of existence of summer ice-free areas, including the Russian Arctic, become increasingly larger throughout the 21st century. In some models, the Arctic Ocean ice cover becomes entirely seasonal by the late 21st century.

Fig. 19 [Kattsov et al., 2007b] shows the evolution of the maximum and minimum sea ice area in the seasonal cycle in the Northern hemisphere during the 20th and 21st centuries (A2 scenario). As can be seen, forced sea ice area changes in the Northern Hemisphere occurred between 1980 and 2000, with a small lead of the decrease in the minimum (September) sea ice area over a decrease in annual maximum (March). Further, the lag between September and March in the ensemble-mean sea ice area reduction rate increases up to the late 21st century. This occurs against the background of the increasing scatter among the models. The scatter increases more rapidly in September than in March. By the late 21st century, the scatter among September ice areas in the Northern Hemisphere begins to decrease because of the entirely ice-free Arctic Ocean in several models.

The Northern Hemisphere sea-ice mass reduction in the AOGCM ensemble becomes statistically significant at the 5% level in early 1980s. Further mass reduction is much more rapid in March than in September, which indirectly confirms the tendency of thick multiyear ice to become seasonal [Kattsov et al., 2007b].

Fig.20 shows the March and September sea-ice distributions in the Northern Hemisphere in the simulations with 12 AOGCMs, for which data were available for A2 scenario. The differences among March sea-ice distributions through the 21st century are not significant, as are the differences of each of the time slice from the corresponding distribution in the base line period 1980–1999. However, there is an entirely different situation for September. By the late 21st century, the Arctic Ocean becomes ice-free in several AOGCMs. There is also a strong reduction, from scenario B1 to A1B and next to A2, in the central Arctic, which is consistent with the findings that much of the multiyear ice in the Northern Hemisphere becomes seasonal by the late 21st century.

 

4. Concluding remarks

State-of-the-art projections have different credibility for different climate characteristics – e.g., higher for temperature and lower for precipitation, or higher for means and lower for extremes. Changes in the distribution of climatic events are at least as interesting as changes in the mean when the impacts of climate change are estimated. In this Assessment the results of up to 16 AOGCMs have been used to quantify future changes in the climate over the territories of RF and CAS. In order to decrease the uncertainties of the projections (including those due to natural variability, model sensitivity to prescribed forcings and due to forcings themselves), a much larger sample of simulations is needed. Estimates of extreme events and their frequency of occurrence also require massive ensemble simulations. In addition, it would be advantageous to increase model resolution to better capture physical processes and to better describe sharp spatial gradients, which are often in the regions where extreme events occur. Both types of improvements require large additional computing resources.

Further research work is needed to find a reasonable balance between ensemble size, model resolution and complexity of physical process descriptions. Increasing accuracy of the ensemble projections of probability distribution functions for numerous climatic variables also depends on the progress in developing metrics for objective discrimination of the models included in the ensembles.

Figure 19. Ensemble-mean evolution of the Northern hemisphere sea ice area anomalies (%, relative to 1910-1950) in March (upper panel) and September (lower panel) through the 20th and 21st centuries (A2 scenario). The horizontal red lines indicate the interval beyond which the change becomes significant at 5% level. The shaded area shows scatter of model estimates. [Kattsov et al., 2007b]

Figure 20. Sea ice distribution in the Northern Hemisphere simulated by 12 AOGCMs for March (upper panel) and September (lower panel): 1980-1999 (left); 2041-2060 (middle); and 2080-2099 (right) – A2 scenario. For each 2.5°x2.5° longitude-latitude grid cell, the figure indicates the number of models that simulate at least 15% of the area covered by sea ice. Grid cells in which the ice is present in all 12 models are dark gray.

The observed (1980-1999) 15% concentration boundaries (red line) are based on the Hadley Centre Sea Ice and Sea Surface Temperature data set (HadISST [Rayner et al., 2003]).

Saint Petersburg, February 2008 

Acknowlegement

This work was originally prepared for the World Bank study "Adaptation to Climate Change in Eastern Europe and Central Asia".

 

References

Alexander L.V., X. Zhang, T. C. Peterson, J. Caesar, B. Gleason, A. M. G. Klein Tank, M. Haylock, D. Collins, B. Trewin, F. Rahimzadeh, A. Tagipour, K. Rupa Kumar, J. Revadekar, G. Griffiths, L. Vincent, D. B. Stephenson, J. Burn, E. Aguilar, M. Brunet, M. Taylor, M. New, P. Zhai, M. Rusticucci, and J. L. Vazquez-Aguirre, 2006: Global observed changes in daily climate extremes of temperature and precipitation. J.Geophys.Res., 111, D05109, doi:10.1029/2005JD006290.
Arzel, O., T. Fichefet, and H.Goose, 2006: Sea ice evolutionover the 20th and 21st centuries as simulated by current AOGCMs. Ocean modeling, 12, 401-415.
Bony, S., R. Colman, V. Kattsov, R. Allan, C. Bretherton, J.-L. Dufrense, A. Hall, S. Hallegatte, M. Holland, W. Ingram, D. Randall, B. Sodfen, G. Tselioudis, M. Webb, 2006: How Well do we Understand and Evaluate Climate Change Feedback Processes? J.Climate, 19, 3445-3482.
Frich, P., L.V. Alexander, P. Della-Marta, B. Gleason, M. Haylock, A.M.G. Klein-Tank, and T.Peterson, 2002: Observed coherent changes in climatic extremes during the second half of the twentieth century. Clim. Res., 19, 193-212.
Govorkova, V.A., V.M. Kattsov, V.P. Meleshko, T.V. Pavlova, I.M. Shkolnik, 2008: Climate of Russia in the 21st century. Pt.2: Evaluation of CMIP3 atmosphere-ocean general circulation models validity for projecting climate change over the territory of Russia. Russian Meteorology and Hydrology. (submitted)
Gruza, G.V., and E.Ya. Ran'kova, 2003: Climate Oscillations and Changes over Russia. Izvestiya Russian Academy of Sciences. Atmospheric and Oceanic Physics, 39, № 2, 166-185
Hegerl G., F.W.Zwiers, P.A.Stott, and V.V.Kharin, 2004: Detectability of anthropogenic changes in annual temperature and precipitation extremes. J. Climate, 17, 3683-3700. Kattsov, V., and E. Kallen, 2005: Future climate change: modeling and scenarios for the Arctic. – In Arctic Climate Impact Assessment (ACIA). Cambridge University Pess, Cambridge, UK, 99-150.
Kattsov, V.M., J.E. Walsh, W.L. Chapman, V.A. Govorkova, T.V. Pavlova, and X. Zhang, 2007a: Simulation and Projection of Arctic Freshwater Budget Components by the IPCC AR4 Global Climate Models. J. Hydrometeorology, 8, 571-589.
Kattsov, V.M., G.A. Alekseev, T.V. Pavlova, P.V. Sporyshev, R.V. Bekryaev, V.A. Govorkova, 2007b: Modeling the evolution of the World Ocean ice cover in the 20th and 21st centuries. Izvestia of Russian Academy of Sciences: Physics of Atmosphere and Ocean, 43, 165–181.
Kharin V.V., F.W. Zwiers, X. Zhang, G. Hegerl, 2007: Changes in temperature and precipitation extremes in the IPCC ensemble of global coupled model simulations. J.Climate, DOI: 10.1175/JCL14066.1.
Malevsky-Malevich, S.P., E.K.Molkentin, E.D. Nadyozhina, O.B. Shklyarevich, 2001: Numerical Simulation of Permafrost Parameters Distribution. Cold. Reg. Science and Tech., №32, 1-11.
Meehl, G.A., C. Covey, T. Delworth, M. Latif, B. McAvaney, J.F.B. Mitchell, R.J. Stouffer and K.E. Taylor, 2007: The WCRP CMIP3 Multimodel Dataset: A New Era in Climate Change Research, BAMS, 88, 1383-1394, DOI: 10.1175/BAMS-88-9-1383 Meleshko, V.P., V.A. Govorkova, T.V. Pavlova, 2007: Impact of the 2ºC global warming on hydrological cycle and permafrost in the Northern Eurasia. Trudy GGO (MGO Proceedings), 556, 29-40.
Meleshko, V.P., V.M. Kattsov, V.M. Mirvis, V.A. Govorkova, T.V. Pavlova, 2008a: Climate of Russia in the 21st century. Pt.1: New evidences of antropogenic effect on climate and new opportunities for its changes evaluation over the territory of Russia. Russian Meteorology and Hydrology. (in press)
Meleshko, V.P., V.M. Kattsov, V.A. Govorkova, P.V. Sporyshev, I.M.Shkolnik, B.Ye. Shneerov, 2008b: Climate of Russia in the 21st century. Pt.3: Projecting future climate changes over the territory of Russia by an ensemble of CMIP3 atmosphere-ocean general circulation models. Russian Meteorology and Hydrology. (submitted)
Nakicenovic N., J. Alcamo, G. Davis, B. de Vries, J Fenhann, S. Gaffin, K. Gregory, A.Grubler, T.Y. Jung, T. Kram, E.L. La Rovere, L. Michaelis, S. Mori, T. Morita, W.Pepper, H. Pitcher, L. Price, K. Raihi, A. Roehrl, H.-H. Rogner, A. Sankovski, M.Schlesinger, P.Shukla, S. Smith, R. Swart, S. van Rooijen, N. Victor, Z. Dadi, 2000: IPCC Special Report on Emission Scenarios. Cambridge University Press, United Kingdom and New York, NY, USA.
Pavlova, T.V., V.M. Kattsov, Ye.D. Nadyozhina, P.V. Sporyshev, V.A. Govorkova, 2007: Terrestrial cryosphere evolution through the 20th and 21st centuries as simulated with the new generation of global climate models. Earth Cryosphere, 11, No.2, 3-13.
Randall, D.A., R.A. Wood, S. Bony, R. Colman, T. Fichefet, J. Fyfe, V. Kattsov, A. Pitman, J. Shukla, J. Srinivasan, R.J. Stouffer, A. Sumi and K.E. Taylor, 2007: Climate Models and Their Evaluation. In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.
Rayner N.A., D.E.Parker, E.B.Horton, C.K.Folland, L.V.Alexander, and D.P.Rowell, E.C.Kent, A.Kaplan, 2003: Global analysis of sea surface temperature, sea ice, and night marine air temperature since the late ninteeth century. J.Geophys.Res., 108 (D14), doi:10,1029/2002JD002670.
Roshydromet, 2008: First Assessment Report on Climate Changes and Their Impacts over the Territory of Russian Federation (in press). Shkolnik, I.M., V.P. Meleshko, V.M. Kattsov, 2007: The MGO regional climate model for climate studies over Siberia. Meteor. Gidrologia (Russian Meteorology and Hydrology), No.6, 5-18.
Shkolnik I.M., V.P. Meleshko, V.M. Kattsov, 2006: Possible climate change over the European part of Russia and adjacent territories by the end of the 21st century: a computation with the MGO regional model. Meteor. Gidrologia (Russian Meteorology and Hydrology), No.3, 5-16.
Solomatina O.M., 1992: Mountain glaciers at late glacial period. Nature. №5 (in Russian)
Zhang, X., and J.E. Walsh, 2006: Toward a seasonally ice-covered Arctic Ocean: Scenarios from the APCC AR4 model simulations. J. Climate, 19, 1730-1747.