—Plenary Paper— Thermal State and Fate of Permafrost in Russia: First Results of IPY

V.E. Romanovsky, A.L. Kholodov, S.S. Marchenko Geophysical Institute, University of Alaska Fairbanks N.G. Oberman MIRECO Mining Company, Syktivkar, Russia D.S. Drozdov, G.V. Malkova, N.G. Moskalenko, A.A. Vasiliev Institute of Earth Cryosphere, Tyumen, Russia D.O. Sergeev Institute of Environmental Geoscience, Moscow, Russia M.N. Zheleznyak Melnikov Permafrost Institute, Yakutsk, Russia

Abstract To characterize the thermal state of permafrost, the International Permafrost Association launched its International Polar Year Project #50, Thermal State of Permafrost (TSP). Ground temperatures are measured in existing and new boreholes within the global permafrost domain over a fixed time period in order to develop a snapshot of permafrost temperatures in both time and space. This data set will serve as a baseline against which to measure changes of near- surface permafrost temperatures and permafrost boundaries, to validate climate model scenarios, and for temperature reanalysis. The first results of the project based on data obtained from Russia are presented. Most of the observatories show a substantial warming during the last 20 to 30 years. The magnitude of warming varied with location, but was typically from 0.5 to 2°C at the depth of zero annual amplitude. Thawing of Little Ice Age permafrost is ongoing at many locations. There are some indications that Late Holocene permafrost has begun to thaw at some undisturbed locations in northeastern Europe and in northwest Siberia. Projections of future changes in permafrost suggest that by the end of the 21st century, Late Holocene permafrost in Russia may be actively thawing at all locations and some Late Pleistocene permafrost could start to thaw as well.

Keywords: dynamics of permafrost; long-term thaw; modeling; temperature regime.

Introduction scenarios, and for temperature reanalysis. Permafrost has received much attention recently because More than half of Russia is occupied by permafrost, surface temperatures are rising in most permafrost areas of constituting a significant portion of the entire Northern the Earth, which may lead to permafrost thaw. Thawing of Hemisphere permafrost area. Hence, without comprehensive permafrost has been observed at the southern limits of the understanding of permafrost dynamics in Russia it will be permafrost zone; thawing can lead to changes in ecosystems, very difficult to draw any general conclusions about the in water and carbon cycles, and in infrastructure performance. state and fate of permafrost in the Northern Hemisphere. If the current trends in climate continue, warming of Permafrost research in Russia has a long, rich history. Many permafrost will eventually lead to widespread permafrost historically active institutions are still active in permafrost thawing in the colder permafrost zones. There is, however, research today, though there is a strong need to develop uncertainty concerning where this thawing will occur first, an integrated network of permafrost research stations to the rate of thaw, and the consequences for arctic, subarctic, improve the efficiency and sustainability of these efforts. and global natural systems. Hence, it is critically important to The Russian-US TSP project funded by the US National organize and sustain continuous observations of the thermal Science Foundation (NSF) was established to initiate state of permafrost in various locations and for various the process of collaborating and integrating US (mainly natural settings within the entire Earth permafrost domain. To Alaskan) and Russian permafrost observing stations into characterize the thermal state of permafrost, the International an International Network of Permafrost Observatories Permafrost Association launched its International Polar Year (INPO) within the framework of the International Polar Year Project #50, Thermal State of Permafrost (TSP). Ground (IPY). Several institutions from the universities and the US temperatures are measured in existing and new boreholes Geological Survey and more than ten Russian institutions within the global permafrost domain over a fixed time period and organizations are involved in this project. This project in order to develop a snapshot of permafrost temperatures is open to new participants, both individual and institutional. in both time and space (Brown & Christiansen 2006). The The first results of this project based on both currently resulting data set will serve as a baseline against which to measured and historical data from several permafrost regions measure changes of near-surface permafrost temperatures in Russia (Fig. 1) are presented in this paper. and permafrost boundaries, to validate climate model Examination of past trends in permafrost conditions and

1511 1512 Ni n t h In t e r n a t i o n a l Co n f e r e nc e o n Pe r m a f r o s t

Figure 1. Location of selected Russian TSP research areas discussed Figure 2. Present-day distribution of permafrost of different ages in in this paper. Russia (after Lisitsyna & Romanovskii 1998). distribution (especially during the last glacial-interglacial meters of permafrost was ice-rich, such as in the Pechora River cycle) can also facilitate better understanding of the possible basin and in the northern and central parts of West Siberia, rates and pathways of permafrost degradation in the permafrost had not disappeared completely; it is still present future. The primary reasons for this are: 1) many present- at greater depths (200 m and deeper). Permafrost on land day features in permafrost distribution both vertically and was generally stable and did not experience any widespread laterally were formed during the last 100,000 years, and 2) thaw during the Holocene Optimum within the northern we can expect that with persistent future climate warming, part of Central Siberia and within the entire continuous the first permafrost to thaw will be the youngest Little Ice Age permafrost zone in East Siberia and in the Russian Far East permafrost, followed by Mid and Late Holocene permafrost, (Fig. 2). However, numerous thermokarst lakes have rapidly and last to thaw will be the Late Pleistocene permafrost. developed during this period causing localized thawing of Thawing of the Little Ice Age permafrost is ongoing at many permafrost under lakes that were sufficiently deep. locations. There are some indications that Late Holocene Climate cooling during the Middle and Late Holocene permafrost has started to thaw at some specific undisturbed resulted in a reappearance of permafrost in many areas of locations in northeastern Europe, in northwestern Siberia, the present-day discontinuous permafrost zones (Fig. 2). In and in Alaska. In this paper we will briefly describe our some areas, permafrost aggradation was accompanied by knowledge of permafrost development in Russia during the an accumulation of new sediments resulting in so-called last glacial-interglacial cycle and provide currently available syngenetic permafrost formation. More commonly, this information about recent long-term permafrost thawing in new Holocene permafrost was formed by refreezing of this region. already existing sediments and bedrock (termed epigenetic permafrost). In the areas where the Late Pleistocene Permafrost History in Russia During the Last permafrost was still in existence at some depth, two- layered permafrost was formed. The number of newly Glacial-Interglacial Cycle formed thermokarst lakes significantly declined within the Permafrost distribution changed during the last glacial- continuous permafrost area. interglacial cycle in response to changes in climate. During Generally, Holocene climate has been much more stable the last glacial maximum (ca. 20ky BP), permafrost underlay than during the Late Pleistocene. However, several relatively more land area than today. Significant portions of non- warm and cold several-centuries-long intervals can be traced glaciated territory of Europe, northern Eurasia, and North in the Middle and Late Holocene (Velichko & Nechaev America were affected by permafrost. With the termination 2005). During these intervals, new, fairly shallow, short- of the last glacial epoch during the transition from glacial to lived permafrost appeared and disappeared several times in interglacial climate, permafrost started to thaw rapidly both some specific landscape types found within the sporadic and from the top and from the bottom at the southernmost limits discontinuous permafrost zones near the southern boundary of its Late Pleistocene maximum distribution. With climate of present-day permafrost distribution. The last and warming in progress, more and more permafrost in this area probably the coldest of such intervals was the Little Ice Age, became involved in rapid degradation. As a result, by the which dominated most of the Northern Hemisphere climate time of the Holocene Optimum (5-9 ky BP), permafrost between ca. 1600 and 1850. During this period, shallow had completely disappeared from most of the territory of permafrost (15 to 25 m, e.g. Romanovsky et al. 1992) was deglaciated Europe, from northern Kazakhstan, and from established within the sediments that were predominantly a significant portion of West Siberia in northern Eurasia unfrozen during most of the Holocene. Present-day warming (Yershov 1998). In areas where the upper several hundred initiated the Little Ice Age permafrost thawing that has been Ro m a n o v s k y e t a l . 1513 documented for several regions in North America (Jorgensen During the 2007 field season, as a part of the Russian TSP et al. 2001, Payette et al. 2004). project, three boreholes were reconstructed by drilling new boreholes next to the old ones for continuity of temperature A Short Description of Selected Research measurements. The first results of these activities will be Areas and Methods of Measurements presented in this paper. Most of the boreholes at the Russian permafrost research A large number of borehole temperature measurements at stations were equipped with permanently installed thermistor different depths were obtained for the Eurasian permafrost strings and temperatures were measured periodically. In some regions starting in the 1960s. Only a small fraction of boreholes thermistor strings were inserted into the boreholes these data became available for analysis during the initial only for the short period during which measurements implementation stage of the Russian TSP project. Comparing were performed (but long enough to equilibrate thermistor retrospective data with results of modern observation allows temperature with ambient borehole temperature). The estimation of the trends in thermal state of permafrost during accuracy of the measurements using calibrated thermistors the last few decades. During the first year of this project was typically at or better than 0.1°C. In the Vorkuta region (2007), some data from the European North of Russia, north temperatures were monitored using mercury thermometers of the western Siberia and Yakutia regions were collected. with a scale factor from 0.05 to 0.1°C. The thermometers New temperature monitoring instruments were installed in were placed in cases filled with an inert material such as, more than 100 already existing and newly drilled boreholes for example, grease. The frequency of measurements at various locations within the Russian permafrost domain. varied at different locations from once per year to monthly This instrumentation allows automatic continuous collection measurements. Starting in 2006, boreholes are being of temperature data with sub-daily time resolution. equipped with HOBO U-12-008 temperature data loggers The longest permafrost temperature time series from and TMC-HD temperature sensors (www.onsetcomp.com/ Russia in our records are from northeast European Russia products/data-loggers/u12-008). Ice bath testing of these and from northwest Siberia. Permafrost temperatures at sensors and loggers always shows an accuracy of 0.1°C or various depths have been measured since the early 1970s better. Time resolution of these measurements is typically at in the Nadym and Urengoy research areas, since the late four-hour intervals. 1970s in the Vorkuta research area, and since the early 1980s The diversity of past measuring techniques could near the Mys Bolvansky meteorological station located near lead to uncertainty when comparing data obtained using the shore of the Barents Sea (Fig. 1). At each location the these different sensors. Special field experiments were specially designed temperature-monitoring boreholes were performed during the 2007 field season to address this established in different natural landscape settings within concern. Temperatures were measured simultaneously these research areas. with mercury thermometers and data loggers in a borehole Permafrost temperatures in more than 200 boreholes were within the Vorkuta research area (Oberman 2008). In the measured in the mid-1980s in the Transbaykal Chara research Urengoy research area, data logger and thermistor string area (Romanovskiy et al. 1991). This area is characterized measurements were performed simultaneously in the same by extremely high variability in landscape and permafrost borehole. Readings in both cases differed on average by conditions. Temperature measurements were re-established 0.05°C. These experiments assure the comparability of all in the 2000s in a very limited number of selected boreholes measurement techniques at an overall accuracy of 0.1°C. as a part of the Russian TSP activities. The high temporal resolution of data obtained by the newly Temperatures in permafrost boreholes have been measured installed sensors and data loggers also demonstrates that the in the Yakutsk research area (Fig. 1) since the 1960s; however, depths of zero annual amplitude at the Urengoy, Nadym, most collected data are not readily available. In this paper we Vorkuta, and Mys Bolvansky research areas are relatively present permafrost temperature data for 1990–2006 collected shallow, not exceeding 7 to 8 m. by scientists from the Melnikov Permafrost Institute at their long-term monitoring station in Yakutsk. A permafrost Long-Term Changes in Permafrost temperature time series similar in length was obtained from Temperatures the Tiksi research station as a collaborative effort between the Melnikov Permafrost Institute and Hokkaido University, At the Urengoy research area in northwest Siberia (Fig. Japan (Prof M. Fukuda). 1) permafrost temperature at the depth of zero seasonal Since the 1970s, the researchers from the Institute amplitude increased during 1974–2007 in all landscape of Physical-Chemical and Biological Problems of Soil units (Fig. 3A). An increase of up to 2°C was measured at Science (Russian Academy of Science, RAS) have obtained colder permafrost sites (e.g., borehole UR1503, Fig. 3A). occasional temperature measurements within the network of Up to 1°C warming for the same period was observed in boreholes on the Kolyma lowland (Fig. 1). However, many warmer permafrost. A 2°C warming was observed in warm of these boreholes are abandoned now and only a few are still permafrost in a borehole that was situated in deciduous available for further observations. Recently, temperatures in forest (not shown). A similar increase was characteristic four boreholes established in the 1990s were re-measured. of marshes with standing water at the surface (borehole 1514 Ni n t h In t e r n a t i o n a l Co n f e r e nc e o n Pe r m a f r o s t

Figure 4. Time series of mean annual permafrost temperatures at 30 m depth at the Yakutsk (YK; circles) and Tiksi (TK; squares) sites.

that permafrost temperature has not changed at the depth of zero annual amplitude at these sites (boreholes ND14, ND 12, and ND1, Fig. 3B). High temporal resolution data obtained with new data loggers show that all annual variations in temperature have occurred in the upper 2 meters of soil, indicating that permafrost has already begun to thaw internally at these sites. Relative cooling occurred in the Vorkuta region in the late 1970s, mid to late 1980s, and in the late 1990s (Fig. 3C). The most significant warming occurred between the late 1980s and late 1990s. The total warming since 1980 was almost 2°C at the Vorkuta site (Oberman 2008). At the Bolvansky station in northwest Russia the warming trend in air temperature for the last 25 years is 0.04°C/yr; observed trends in mean annual permafrost temperatures vary from 0.003 to 0.02°C/yr in various natural landscapes (Malkova 2008). For the last 10 years, an increase in climatic variability and alternation of extremely cold and extremely warm years has been observed. These changes initially led to a considerable increase in permafrost temperature, followed in 2007 by a small decrease in temperature in most boreholes (Fig. 3D). Figure 3. Decadal trends in permafrost temperatures at selected Continuous 15-year permafrost temperature time series sites in Russia. were obtained in Tiksi and Yakutsk by researchers from the Melnikov Permafrost Institute in collaboration with UR59, Fig. 3A) Generally, the most significant changes scientists from Hokkaido University in Japan. Permafrost in permafrost were measured in the forested and shrubby temperatures at 30 m depth show a slight positive trend for areas; often the formation of taliks up to 10 m thick were the 1990–2005 time period (Fig. 4). In contrast, permafrost observed. In undisturbed tundra, permafrost is still generally temperatures apparently did not change significantly in the stable (Drozdov et al. 2008). It was also observed that most Kolyma research area (Fig. 1) in the eastern Siberian Arctic warming occurred between 1974 and 1997. At the majority during the last 10 to 20 years (Kraev et al. 2008). However, of locations permafrost temperatures did not change, or further study is needed to test this conclusion, which was even cooled between 1997 and 2005. A slight warming has based on recent temperature measurements in or near the occurred since then at sites characterized by temperatures historical boreholes that were recently re-occupied as a part colder than -0.5°C (Fig. 3A). of the TSP program. In the Nadym research area (Fig. 3B) the most significant Permafrost temperatures obtained from the Transbaikal permafrost warming occurred before 1990, by about 1°C research area (Fig. 1) generally increased during the last at a colder site and by up to 0.5°C at the warmer sites 20 years. Borehole #6 was established in the mid-1980s by (Moskalenko 2008). At the warmest site (borehole ND23) researchers from Moscow State University and is located permafrost temperature was -0.1 to -0.2°C and did not in the upper belt of the Udokan Range; temperatures re- change for the entire measurement period (1975–2007). measured at this site show a 0.9°C increase between 1987 Since the temperature reached the same values at the rest of and 2005 at 20 m depth (Fig. 5). Since 2006, temperatures in the warm sites in the late-1980s or early-1990s, it appears this borehole have been recorded at 4-hour time intervals. Ro m a n o v s k y e t a l . 1515

of about 35 m, the permafrost base has been slowly rising. Temperature, centigrade -6 -5 -4 -3 -2 -1 0 Comparing small-scale maps based on 1950–1960 data with 0 maps based on 1970–1995 data shows a shift of the southern 2 limit of permafrost by several tens of kilometers northwards 4 (Oberman 2001). This also indicates that permafrost is 6 mostly degrading in the southernmost part of the region. 8 Permafrost is also degrading in the Nadym and Urengoy 10 research areas. Temperature records from five of a total of

Depth, m seven boreholes in the Nadym area show that cold winter 12 Borehole ?6 temperatures do not penetrate deeper than 2 m into the ground 14 elev. 1712 m a.s.l. and that permafrost became thermally disconnected from 16 seasonal variations in air temperature. This also indicates 18 that the constituent ice is already actively thawing in the 20 09 August 1987 13 August 2005 upper permafrost although the permafrost table (based on the formal 0°C definition of permafrost) is still located just -1.5 below the active layer. In Urengoy, permafrost is thawing -2 in the forested and shrubby areas, developing closed taliks -2.5 TSP stage, (Drozdov et al. 2008). -3 2006-2007 Permafrost degradation in natural undisturbed conditions -3.5 not associated with surface water bodies has not been reported -4 from the other research areas discussed in this paper. There -4.5

Temperature, centigrade are numerous occasions of long-term permafrost thawing in -5 Central Yakutian areas around the city of Yakutsk, but all -5.5 are directly related to natural (forest fire) or anthropogenic -6 (agricultural activities, construction sites) disturbances -6.5 (Fedorov 1996, Fedorov & Konstantinov 2003).

02.09.2006 02.10.2006 02.11.2006 02.12.2006 02.01.2007 02.02.2007 02.03.2007 02.04.2007 02.05.2007 02.06.2007 02.07.2007 Permafrost Temperature Reanalysis and 5 m 10 m 15 m 20 m Date Modeling of Past and Future Changes in Figure 5. Upper graph: Temperature profiles from the Transbaikal Permafrost Temperatures research area obtained in 1987 and 2005. Lower graph: Continuous temperature records at four different depths. Two levels of permafrost modeling are implemented in our research: a “permafrost temperature reanalysis” and Evidence of Long-Term Thawing of spatially-distributed physically based permafrost modeling. Permafrost The first level of modeling is the “permafrost temperature reanalysis” approach (Romanovsky et al. 2002). At this Recently observed warming in permafrost temperatures level, a sophisticated numerical model (Sergueev et al. have resulted in thawing of natural, undisturbed permafrost 2003, Marchenko et al. 2008), which takes into account in areas close to the southern boundary of the permafrost the temperature-dependent latent heat effects, is used to zone. Most observed thawing of long-term permafrost reproduce active layer and permafrost temperature field has occurred in the Vorkuta and Nadym research areas. At dynamics at the chosen sites. The input data are prescribed several locations in the Vorkuta area, long-term thawing of specifically for each site and include a detailed description permafrost has led to the development of new closed taliks of soil thermal properties and moisture for each distinct (Oberman 2008). At one of these locations, the permafrost layer, surface vegetation, snow cover depth and density, table lowered to 8.6 m in 30 years. It lowered even more, and air temperature. In this modeling approach variations to almost 16 m, in an area where a newly developed in air temperature and snow cover thickness and properties closed talik coalesced with an already existing lateral talik. are the driving forces of permafrost temperature dynamics. Permafrost thawing during the last 30 years also resulted in The second level of permafrost modeling involves the the deepening of previously existing closed taliks. The total application of a spatially distributed physically based model increase in depth of the closed taliks developed here ranged that was recently developed in the University of Alaska from 0.6 to 6.7 m depending on the geographical location, Fairbanks (UAF) Geophysical Institute Permafrost Lab genetic type of a particular talik, ice content and lithological (GIPL, Sazonova & Romanovsky 2003). characteristics of the bearing sediments, hydrological, Permafrost temperature reanalysis was successfully hydrogeological, and other factors. applied to several locations within the Russian permafrost As a result of recent climatic warming, permafrost domain (Marchenko & Romanovsky 2007, Romanovsky patches 10 to 15 m thick thawed out completely in this area et al. 2007). The model has been calibrated (Nicolsky et al. (Oberman 2008). In deposits perennially frozen to a depth 2007) for each specific site using several years of measured 1516 Ni n t h In t e r n a t i o n a l Co n f e r e nc e o n Pe r m a f r o s t

Figure 7. Calculated permafrost temperature time series at several depths for a specific location at the Yakutsk meteorological station. Figure 6. Observed (thick gray line) and calculated (thin black line) ground temperatures at the Yakutsk meteorological station at several depths. 1930s. A colder period in the 1940s and especially in the 1960s interrupted this warming trend. Only by the year 2000, permafrost and active layer temperature data and climatic permafrost temperatures at the 20 m depth had returned to data from the closest meteorological station for the same the pre-cooling level. However, permafrost temperature at time interval. After validation on measured data that were greater depths (e.g., 50 m in Fig. 7) continued to increase, not involved in the calibration process, the calibrated model although at a slower rate. can then be applied to the entire period of meteorological The spatially distributed permafrost model GIPL1 that records at each station, producing a time series of permafrost was developed at the Permafrost Lab of the Geophysical temperature changes at various depths. Figure 6 shows the Institute, UAF (Sazonova and Romanovsky 2003, Sazonova results of calibration of the permafrost model using climate at al. 2004) was applied for the entire permafrost domain and soil temperature data from the meteorological station in of northern Eurasia (Fig. 8). For present-day climatic Yakutsk. The differences between measured and modeled conditions, the CRU2 data set with 0.5° x 0.5° latitude/ temperatures are typically less than 0.5°C and rarely exceed longitude resolution (Mitchel & Jones 2005) was used. The 1°C. Deviation of modeled from measured temperatures future climate scenario was derived from the MIT 2D climate decreases with depth (Fig. 6). model output for the 21st century (Sokolov & Stone 1998). The calibrated model was then used to calculate the Due to this model’s spatial resolution (0.5 x 0.5 latitude/ permafrost temperature dynamics during the transition longitude) it is practically impossible to reflect the period from the Little Ice Age up to the present (Fig. 7). discontinuous character of permafrost in the southern Climate forcing includes the air temperature and snow cover permafrost zones. That is why we choose the ground surface dynamics for the period 1833–2003. Air temperature was and soil properties for each cell that will produce the coldest reconstructed by using observed temperature records from possible mean annual ground temperatures within the cell. 1834–1853 and 1887–2003 and a spectral analysis technique This choice means that all results produced by this model (Shender et al. 1999). It was also assumed that there were no reflect permafrost temperatures only in the coldest landscape long-term trends in the snow cover characteristics. types within this area. It also means that if our results show The initial (1833) temperature profile was derived from thawing permafrost somewhere within the domain, we should permafrost temperatures observed by Prof. Middendorf in interpret this to mean that permafrost is thawing in practically 1844–1846 in the Shergin’s mine in Yakutsk (Sumgin et al. all locations within the area. It also means that in the stable 1939). All measurements were made in narrow horizontal permafrost area identified in Figure 8, partial thawing of holes in the walls of the mine. The length of each hole was permafrost may occur at some specific locations. about 2 m. Independent boreholes near Shergin’s mine According to this model, by the end of the 21st century confirmed the characteristic permafrost temperature. Another permafrost that is presently discontinuous with temperatures independent method to test the choice of initial conditions is between 0 and -2.5° C will have crossed the threshold comparing calculated with measured temperatures for the and will be actively thawing. In Russia, the most severe time intervals when both are available. Such a comparison permafrost degradation is projected for northwest Siberia shows a satisfactory agreement (e.g., compare Figs. 4 and and the European North. This model also shows that by the 7). mid-21st century most of the Late Holocene permafrost will The results of calculations shown in Figure 7 indicate be actively thawing everywhere except for the south of East that the most rapid permafrost warming in the Yakutsk area Siberia and the Far East of Russia. occurred during the second half of the 19th century. Mean By the end of 21st century, practically all Late Holocene annual temperatures at the shallow depths (5 and 10 m) were permafrost will be thawing and some Late Pleistocene already at the present-day level by the 1880s and 1890s, permafrost will begin to thaw in the European North and in while deeper temperatures continued to increase up to the Siberia (compare Figs. 2 and 8). Ro m a n o v s k y e t a l . 1517

Conclusions • Most of the permafrost observatories in Russia show substantial warming of permafrost during the last 20 to 30 years. The magnitude of warming varied with location, but was typically from 0.5 to 2°C at the depth of zero annual amplitude. • This warming occurred predominantly between the 1970s and 1990s. There was no significant observed warming in permafrost temperatures in the 2000s in most of the research areas; some sites even show a slight cooling during the late-1990s and early-2000s. Warming has resumed during the last one to two years at some locations. • Considerably less or no warming was observed during the same period in the north of East Siberia. • Permafrost is already thawing in specific landscape settings within the southern part of the permafrost domain in the European North and in northwest Siberia. Formation of new closed taliks and an increase in depth of pre-existing taliks has been observed in this area during the last 20 to 30 years. • Permafrost temperature reanalysis provides a valuable tool to study past changes in permafrost temperature, which helps to place recent changes into a long-term perspective. • An implemented spatially-distributed permafrost model shows that if warming in air temperatures continues to occur, as predicted by most climate models, widespread thaw of Late Holocene permafrost may be in progress by the mid-21st century. If warming continues, some Late Pleistocene permafrost will begin to thaw by the end of the 21st century.

Acknowledgments This research was funded by the Polar Earth Science Program, Office of Polar Programs, U.S. National Science Foundation (ARC-0632400, ARC-0520578), by the Russian Foundation for Basic Research, by the Russian Federal Agency for Mineral Resources Management (№AM-18/35 pr, 30.01.2006), by a NASA Water and Energy Cycle grant, by the State of Alaska, and in cooperation with several earlier and on-going Russian activities. We would like to thank Drs. Jerry Brown, Kenneth Hinkel, and Candace O’Connor for their valuable suggestions, which improved the quality of this paper.

References Brown, J. & Christiansen, H.H. 2006. Report of the International Permafrost Association. Permafrost and Periglacial Processes 17: 377-379. Figure 8. Calculated distribution of mean annual ground Drozdov, D.S., Malkova, G.V. & Melnikov, V.P. 2008. Recent temperatures at the bottom of the active layer in Northern Eurasia advances in Russian geocryological research: a con- averaged for three time intervals: 1990-2000 (upper graph), 2040- tribution to the international polar year. Proceedings 2050 (middle graph), and 2090-2100 (lower graph). The area with of the Ninth International Conference on Permafrost, widespread permafrost thawing from the top down is shown in Fairbanks, Alaska, June 29-July 3 (this proceedings). pink. 1518 Ni n t h In t e r n a t i o n a l Co n f e r e nc e o n Pe r m a f r o s t

Federov, A.N. 1996. Effects of recent climate change Oberman, N. 2008. Contemporary permafrost degradation on permafrost landscapes in central Sakha. Polar of the European north of Russia. Proceedings of Geography 20: 99-108. the Ninth International Conference on Permafrost, Federov, A. & Konstantinov, P. 2003. Observations of Fairbanks, Alaska, June 29-July 3 (this proceedings). surface dynamics with thermokarst initiation, Yukechi Payette, S., Delwaide, A., Caccianiga, M. & Beauchemin, site, Central Yakutia. Proceedings of the Eighth M. 2004. Accelerated thawing of subarctic peatland International Permafrost Conference, Switzerland, permafrost over the last 50 years. Geophysical Research July 21-25, pp. 239-243. Letters 31: L18208, doi:10.1029/2004GL020358. Jorgenson, M.T., Racine, C.H., Walters, J.C. & Osterkamp, Romanovskiy, N.N., Zaitsev, V.N., Volchenkov, S.Yu., Zagryaz- T.E. 2001. Permafrost degradation and ecological kin, D.D. & Sergeyev, D.O. 1991. Alpine permafrost tem- changes associated with a warming climate in central perature zonality, northern Trans-Baikal region, USSR. Alaska. Climatic Change, 48(4): 551-571. Permafrost and Periglacial Processes 2: 187-195. Kraev, G.A., Abramov, Bykhovets, S., Fyodorov-Davydov, Romanovsky, V.E., Garagula, L.S. & Seregina, N.V. 1992. D., Kholodov, A., Lupachev, A., Mamykin, V., Freezing and thawing of soils under the influence of Ostroumov, V., Sorokovikov, V., Gilichinsky, D., 300- and 90-year periods of temperature fluctuation. Zimova, G. & Zimov, S. 2008. Thermal state of Proceedings of the International Conference on the permafrost in the eastern Arctic. Proceedings of Role of Polar Regions in Global Change, University the Ninth International Conference on Permafrost, of Alaska Fairbanks 2: 543-548. Fairbanks, Alaska, June 29-July 3 (this proceedings). Romanovsky, V.E., Burgess, M., Smith, S, Yoshikawa, K. Lisitsyna, O.M. & Romanovskii, N.N. 1998. Dynamics of & Brown, J. 2002. Permafrost temperature records: permafrost in Northern Eurasia during the last 20,000 Indicator of climate change. EOS 83(50): 589-594. years. Proceedings of the Seventh International Romanovsky, V.E., Sazonova, T.S., Balobaev, V.T., Shender Permafrost Conference, Yellowknife, Canada, June N.I. & Sergueev, D.O. 2007. Past and recent changes 23-27, pp. 675-681. in permafrost and air temperatures in Eastern Siberia. Malkova, G.V. 2008. Last twenty-five years of changes Global and Planetary Change 56: 399-413. in permafrost temperature of the European north Sazonova, T.S. & Romanovsky, V.E. 2003. A model for of Russia. Proceedings of the Ninth International regional-scale estimation of temporal and spatial Conference on Permafrost, Fairbanks, Alaska, June variability of the active layer thickness and mean 29-July 3 (this proceedings). annual ground temperatures. Permafrost and Marchenko, S.S. & Romanovsky, V.E. 2007. Modeling the Periglacial Processes 14(2): 125-139. effect of organic layer and water content on permafrost Sazonova, T.S., Romanovsky, V.E., Walsh, J.E. & Sergueev, dynamics in the Northern Hemisphere. Eos Trans. D.O. 2004. Permafrost dynamics in 20th and AGU, 88(52): Fall Meet. Suppl., POSTER, Abstract 21st centuries along the East-Siberian Transect. GC23A-0985. Journal of Geophysical Research 109: D01108, Marchenko, S.S, Romanovsky, V.E. & Tipenko, G.S. 2008. doi:10.1029/2003JD003680. Numerical modeling of spatial permafrost dynamics Sergueev, D., Tipenko, G., Romanovsky, V. & Romanovskii, in Alaska. Proceedings of the Ninth International N. 2003. Mountain permafrost thickness evolution Conference on Permafrost, Fairbanks, Alaska, June under influence of long-term climate fluctuations 29-July 3 (this proceedings). (results of numerical simulation). In: Permafrost, Mitchell, T.D. & Jones, P.D. 2005. An improved method Phillips, M., S. Springman, and L.U. Arenson (eds.), of constructing a database of monthly climate Swets & Zeitlinger, Lisse, 1017-1021. observations and associated high-resolution grids. Shender, N.I., Romanovsky, V.E. & Tetelbaum, A.S. 1999. International Journal of Climatology 25(6): 693- A forecast of the natural variability of climate in Ya- 712. kutsk and Fairbanks (in Russian). Science and Educa- Moskalenko, N.G. 2008. Vegetation and permafrost changes tion 2: 24-29. in West Siberia northern taiga. Proceedings of the Sokolov, A.P. & Stone, P.H. 1998. A flexible climate model Ninth International Conference on Permafrost, for use in integrated assessments. Climate Dynamics Fairbanks, Alaska, June 29-July 3 (this proceedings). 14: 291-303. Nicolsky, D.J., Romanovsky, V.E. & Tipenko, G.S. 2007. Sumgin, M.I., Kachurin, S.P., Tolstikhin, N.I. & Tumel, Using in-situ temperature measurements to estimate V.F. 1939. General Geocryology (in Russian), Soviet saturated soil thermal properties by solving a sequence Academy of Science Publishing, 340 pp. of optimization problems. The Cryosphere 1: 41-58. Velichko A.A. & Nechaev, V.P. (eds.). 2005. Cenozoic Oberman, N.G. 2001. Intra-century dynamics of the climatic and environmental changes in Russia, edited permafrost zone in the European Northeast of Russia. by H.E. Wright Jr., T.A. Blyakharchuk, A.A. Velichko Proceedings of the 2nd Conference of Geocryologists & O. Borisova. Special Paper 382. The Geological of Russia, Moscow, Russia, June 6-8, 2001 2: 212- Society of America, 226 pp. 217. Yershov, E. 1998. General Geocryology. Cambridge: Cambridge University Press, 580 pp.