Representing permafrost affected ecosystems in the CLM: An example of incorporating empirical ideas into the CLM Hanna Lee Climate and Global Dynamics Division National Center for Atmospheric Research
Knowledge from the field/lab Model development Global estimation of permafrost C-climate feedback Reducing uncertainty 2
Biogeochemical consequences - Permafrost C Permafrost 1672Pg Carbon stored in permafrost 1/2 of global soil C stock x 2 more than C in the atmosphere Thawed permafrost will release C-based greenhouse gases at a faster rate! Schuur et al. 2008 BioScience Tarnocai et al. 2009 GBC 3
Potential Arctic-climate feedback Global warming + Positive feedback Arctic warming CO 2 uptake by Plant growth Permafrost thaw Soil N Decomposition CO 2 CH 4 4
Potential Arctic-climate feedback - Negative feedback CO 2 uptake by Plant growth Global warming Arctic warming CO 2 CH 4 Permafrost thaw Soil N Decomposition 5
Physical consequences - Thermokarst formation Land surface subsidence created by ice rich permafrost thaw Changes in local hydrology: Aerobic vs. Anaerobic -> C cycling 6
Climate effects from permafrost C release Laboratory incubation Aerobic Anaerobic Deep permafrost C release under aerobic and anaerobic conditions : Faster C release under aerobic conditions Aerobic Modified from Lee et al. 2012 GCB 7
Climate effects from permafrost C release Laboratory incubation Aerobic Anaerobic Anaerobic Deep permafrost C release under aerobic and anaerobic conditions : Comparable in atmospheric forcing with CH 4 effect Modified from Lee et al. 2012 GCB 8
Research Question 1. How does permafrost thaw influence ecosystem carbon balance under warmer world? 9
Interior Alaska tundra site Denali National Park 2003 aerial photo 10
Interior Alaska tundra site Deep permafrost T increase Natural gradient of permafrost thaw and thermokarst development Osterkamp & Romanovsky 1999 PPP Three sites: Minimal Thaw: Typical tussock tundra before thawing Moderate Thaw: 15-20 yrs of permafrost thaw Extensive Thaw: over 50 yrs of thaw and deep thermokarst Denali National Park 2003 aerial photo 11
Aboveground carbon balance CO 2 uptake: GPP Balance: NEE CO 2 release: R eco Autochambers 12
Aboveground Net Ecosystem Exchange of carbon Balance between carbon uptake and release Net Ecosystem Exchange (gc m -2 ) 40 20 0-20 -40-60 -80 C sink: More uptake C source: More release Minimal Moderate Extensive Over 3 years: Minimal neutral Moderate = sink ( GPP, R eco ) Extensive = source ( GPP, R eco ) Modified from Schuur, Vogel, Crummer, Lee, Sickman, Osterkamp 2009 Nature 459: 556-559. Vogel, Schuur, Trucco, Lee 2009 J Geophys Res 114, G4, doi:10.1029/2008jg000901. 13
NEE Net Carbon Exchange Projections + sink 0 1965 2005 2099 source - 1950 Thawing started Year Carbon neutral when thawing started 14
NEE Net Carbon Exchange Projections + sink 0 Early stage C uptake +25 g m -2 1965 2005 2099 source - 1950 Year Carbon uptake of ~25 g m -2 in the early stage of thawing 15
NEE Net Carbon Exchange Projections + sink 0 Later stage C release 1965 2005 2099 source - 1950-32 g m -2 Year Carbon release of ~32 g m -2 in the later stage of thawing 16
NEE Net Carbon Exchange Projections + sink 0 1965 2005 2099 source - 1950 Year -52 to -69 g m -2 Expected equilibrium Carbon loss by 2099: 4.4-6.0 kg m -2 (9.4-12.9%) 17
Permafrost Carbon Loss in global carbon context Using the three sites as representatives of permafrost thaw Permafrost Zone Soil C Gelisol Soil Order (3m) Permafrost C Loss Current Land Use Change 818 Pg x 9.4-12.9% 77-106 Pg (0.8-1.1 Pg/yr) (1.5±0.5 Pg/yr) Schuur, Vogel, Crummer, Lee, Sickman, Osterkamp 2009 Nature 459: 556-559. 18
Research Question 2. What is the climate feedback from permafrost C? 19
Improvements in Community Land Model 4.5 Diffuse solar Emitted longwave Downwelling longwave Latent heat flux Sensible heat flux Reflected solar Absorbed solar Aerosol deposition SCF Soil (sand, clay, organic) Bedrock Surface fluxes Soil heat flux Precipitation Momentum flux Wind speed 0 u a Evaporation Dust Transient layer that acts like permafrost Sublimation Melt Soil Aquifer recharge Unconfined aquifer Hydrology Transpiration Throughfall Water table Evaporation Infiltration Surface runoff Wetlands Permafrost layer Saturated fraction Sub-surface runoff Phenology Fire Biogeochemical cycles Photosynthesis CO BVOCs 2 /CH 4 dynamics Soil C/N Vegetation C/N Litterfall N mineralization Heterotrophic respiration Root litter N uptake Autotrophic respiration N dep N fixation Denitrification N leaching Glacier Lake Urban Wetland River discharge Wood harvest Runoff River Routing Land Use Change Competition Vegetation Dynamics Disturbance Growth Lawrence et al., Journal Advances Modeling Earth Systems, 2011 20
Physical consequences - Thermokarst formation Land surface subsidence created by ice rich permafrost thaw Changes in local hydrology: Aerobic vs. Anaerobic -> C cycling 21
Excess ice and permafrost parameterization Agriculture & Agri-Food Canada 22
Excess ice and permafrost parameterization Soil layers and Excess ice Control Excess ice Melting Surface subsidence Ice and soil mixture at layers 8-10 Control: Regular CLM soil layers Used ground ice data from NSIDC 23
Continuous: 90-100% Discontinuous: 50-90% Sporadic: 10-50% Isolated: 0-10% High: > 20% Medium: 10-20% Low: 0-10% Brown, J., O.J. Ferrians, Jr., J.A. Heginbottom, and E.S. Melnikov.. 2002. Circum-Arctic Map of Permafrost and Ground-Ice Conditions. Version 2. [indicate subset used]. Boulder, Colorado USA: National Snow and Ice Data Center.
Soil temperature patterns: Annual variability Soil warming slowed ~1 C by the end of the century 25
Soil moisture patterns Soil moisture increases Soil water storage increases Most of excess ice melt water goes to soil water storage 26
Surface subsidence simulations (Recent/RCP8.5) Land surface subsidence as a function of excess ice melting Grid cell mean 27
Improved model and climate feedback Global warming Arctic warming CO 2 uptake by Plant growth Expanded wetlands Permafrost thaw Soil N CH 4 Decomposition CO 2 Lakes drain, soils dry Enhanced predictions of Arctic-climate feedback with improved models 28