1. Estimation of possible active layer depth changes in North-East of Russia using climate projections and deterministic-stochastic approach Liudmila Lebedeva 1,3, Olga Semenova 2,3 1 St.Petersburg State University 2 State Hydrological Institute, St. Petersburg, Russia 3 Hydrograph Model Research Group www.hydrograph-model.ru

2. Goal and objectives Application and testing of the Deterministic-Stochastic Modelling system Assessment of possible changes of active layer depth properties on the base of IPCC climate change scenarios Objectives The goal of the research is to develop a tool for assessment of possible climate change impacts in permafrost environment of the North-East Russia using deterministic-stochastic modelling approach Requirements Process-oriented deterministic model with physically observable parameters, minimum of calibration and ability to port calibrated parameters to similar environment Downscaled climate change projections in probabilistic mode

3. Research stages To simulate soil thawing and freezing processes at different landscapes using observed meteorological data as input To refine the deterministic model parameters on historical data To generate continuous series of daily meteorological data (30 years span) according the A1F1 and B1 IPCC climate change scenarios using the stochastic model To simulate water and energy fluxes in the permafrost sites with randomly generated series of meteorological elements as forcing data using the deterministic model with physical parameters To assess and compare possible changes in active layer properties within variable conditions and according different climate change scenarios

4. Hydrograph Model Deterministic distributed model of runoff formation processes Heat and water dynamics simulations in soil profile Use of observable physical properties of landscapes as model parameters Minimum of manual calibration Forcing data: precipitation, temperature, relative humidity Output results: runoff, soil and snow state variables, full water balance R

5. Applications of the Hydrograph model in permafrost environments (runoff simulations) Granger watershed, 8 km 2 (Yukon basin, Canada – zone of discontinuous permafrost) 19992000 Vitim at Bodaybo, 186000 km 2 (Eastern Siberia, continuous permafrost) Yuzhny creek, KWBS 0.27 km 2

6. Stochastic Model “Weather” (SMW) Simulation of daily precipitation, temperature and relative humidity Simulation of annual and intra-seasonal variations Spatial and temporal correlation of meteorological elements Initial parameters are estimated from observed series of meteorological data Parameters may be modified according to climate change projections

7. Applications of SMW (observed and simulated series) Calc Obs Monthly distribution of precipitation, Bodaybo station Monthly distribution of air temperature, Vostochnaya station Daily values of precipitation, Suntar-Hayata station Annual sums of precipitation, Chara station

8. Research strategy Stochastic Model of Weather Deterministic hydrological model Physically observable parameters Parameters of observed daily meteorological series Climate change projections Simulated ensembles of meteorological data according to IPCC climate change projections Deterministic simulation of processes using stochastic data Numerical evaluation of hydrological changes

9. Sketch of the KWBS Study area Kolyma water-balance station (KWBS) Kolyma water-balance station (KWBS) – small research watershed (22 km 2 ) in the upper Kolyma river. Watershed boundaries Meteorological Station Rain gauge Recording rain gauge Pit gauge Snow survey line Cryopedometer Evaporation plot Pan evaporation plot Snow evaporation plot Water balance plot Mean annual temperature – -11,6 0 C Precipitation – 314 mm/year Open wood, bare rocks Continuous permafrost High-mountain relief Representative for the North-East of Russia

10. Active layer depth in different landscapes Site 1 Upper part of the slope: Site 2 Lower part of the slope: clay slate rock debris rock debris absence of vegetation absence of vegetation peaty ground peaty ground swamp larch forest swamp larch forest

11. Soil physical properties The main parameters for simulation soil thawing and freezing processes in the Hydrograph model are physical soil properties Porosity, % Field capacity, % Heat capacity, J/m 3 *K Heat conductivity, W/m*K Peat 805019200.8 Clay slate 50407502.3 Crushed stone 55308101.7 Crumbling rock 55137902

12. Deterministic modelling of active layer depth Site 1: 1100 m South-facing slope Absence of vegetation Rock debris Active layer depth up to 1.7 m m Site 2: 850 m North-facing slope Sphagnum, shrubs Soil profile – peat, clay loam, clay slate Active layer depth up to 0.7 m Observed and calculated active layer depth in two landscapes, KWBS

13. IPCC emission scenarios Implications of emission scenarios for global Tº by 2100 relative to 1990 (chosen scenarios and the model marked as red) Atmospheric-Ocean General Circulation Models ScenarioGlobal ΔT( 0 C) A1F14.5 A1B2.9 A1T2.5 A23.8 B12.0 B22.7 ModelCountryΔT glob CCSR/NIESJapan4.4 CGCM2Canada3.5 CSIRO Mk2Australia3.4 ECHAM4/OPYC3Germany3.3 GFDL R30U.S.A.3.1 HadCM3United Kingdom3.2 NCAR DOE PCMU.S.A.2.4

14. ECHAM4/OPYC3 model projection according to A1F1 and B1 scenarios for 2010-2039

15. Soil thawing – depth projections Maximum active layer depth: projected according to B1 and A1F1 scenarios and historically observed in different landscape Mean active layer depth: projected according to B1 and A1F1 scenarios and historically observed in different landscapes

16. Duration of thawing period (B1, A1F1 and historical) Bare rocks Swamp forest

17. Results Both mean and maximum annual active layer depths are projected to increase by 2039. Mean soil thawing is expected to be 40 and 50 cm deeper than historically observed reaching 195 and 205 cm in the bare rock site according to B1 and A1F1 scenario Mean soil thawing is expected to be 70 cm deeper than historically observed reaching 130 cm in the swamp forest site for both scenarios The starting date of soil thawing in bare rocks is projected to be almost one month earlier due to strong effect of south-facing slope and solar radiation income The starting dates of soil thawing in swamp forest landscape are projected to be only one week earlier in comparison with historical data Total duration of thawing period is projected to extend by 1 – 1.5 months for bare rocks and about 1 month for forest landscape

18. General conclusions The deterministic hydrological model Hydrograph is able to simulate adequately the processes of soil thawing and freezing in permafrost environment Observable physical properties of landscapes are used as the model parameters; the model requires minimum of calibration To assess the possible effect of climate change on active layer depth the processes-based deterministic models are required. The Hydrograph model may be considered to be one of those models The stochastic model of Weather was used here to downscale climate change projections for specific sites and generate numerous continuous series of meteorological data with assigned parameters. In general, it can be replaced with some more advanced models and was used here as an example of the approach

19. Acknowledgements This study was conducted within the research grant provided by the Russian-German Otto-Schmidt Laboratory for Polar and Marine research in 2010 The conference attendance was made possible with the support of APECS which is highly appreciated

20. Thank you for attention! More results on applications of the Hydrograph model in assessment of climate change impacts on runoff in permafrost environment… Poster 124 Evaluation of climate change impact on river runoff in Eastern Siberia by Semenova et al.