Key Points in Linking Dynamic Ecosystem Models with Permafrost and Hydrology Models A. David McGuire (UAF), Eugenie Euskirchen (UAF), and Shuhua Yi (UAF) Arctic System Model Workshop, August 6 and 7, 2007

Interactions of Northern High Latitude Terrestrial Regions with the Earth’s Climate System Regional ClimateGlobal Climate Northern High Latitude Terrestrial Regions Impacts Water and energy exchange Exchange of carbon-based greenhouse gases (CO 2 and CH 4 ) Delivery of freshwater to Arctic Ocean

From McGuire, Chapin, Walsh, and Wirth Integrated regional changes in arctic climate feedbacks: Implications for the global climate system. Annual Review of Environment and Resources 31: Physiology Climate warming Structure Land Use composition, vegetation shifts Disturbance CO 2, SH  Permafrost warming, thawing Physical feedbacks Biotic control Mediating processes       Snow cover 1, 2, 3, 4 5, 6, 7 8, 9 10, 11 12, 13 A B C enzymes, stomates fire, insects logging, drainage, reindeer herding D E I II IV III V  fast (seconds to months) intermediate (months to years) slow (years to decades) Response time Mechanisms:  : albedo GH: ground heat flux SH: sensible heat flux CO 2, CH 4 : atmospheric concentration Physiological feedbacks: (1) higher decomposition CO2  (2) reduced transpiration SH  (3) drought stress: CO2  (4) PF melting: CH4  (5) longer production period: CO2  (6) NPP response to N min: CO2  (7) NPP response to T: CO2  Structural feedbacks: (8) shrub expansion:   (9) treeline advance:  , CO2  (10) forest degradation   but CO2, SH  (11) light to dark taiga:   but CO2, SH  (12) more deciduous forest:  , SH  (13) fire / treeline retreat:   Physical feedbacks: (14) increased, then reduced heat sink GH ,SH  (15) watershed drainage SH  (16) earlier snowmelt  

Terrestrial Research Focus Areas at IARC Physical Feedbacks Involving Permafrost Responses Feedbacks Involving Carbon and Water Responses Feedbacks Involving Snow Responses Feedbacks Involving Responses of Vegetation Composition and Structure

Friedlingstein et al. 2006; IPCC SRES 2000 Coupled Climate-Carbon Cycle Model Intercomparison (C 4 MIP) ppm Biospheric Carbon-Climate Feedback - All soils treated as mineral soils - No C-hydrology dynamics in peatlands - No C-thawing dynamics in permafrost - No Nitrogen-Phosphorus limitations - Most models don’t have fire - Most don’t have vegetation dynamics Up to +1.5°C Atm. CO 2 difference (ppm)

Feedbacks Involving Carbon and Water Cycle Responses Some Key Issues: - vulnerability to fire and permafrost thaw - delivery of carbon from high latitude terrestrial ecosystems to marine environments - dynamic simulation of wetlands

Vulnerability to CO 2 and CH 4 release Zhuang et al Geophysical Research Letters.  Permafrost thawing (MIT IGSM Scenarios)  Fire disturbance increase (~1% yr -1 )

Soil Thermal Module (STM) Hydrological Module (HM) Terrestrial Ecosystem Model (TEM) Methane Consumption and Emission Module (MCEM) Soil Temperature Profile Active Layer Depth Water Table and Soil Moisture Profile Labile carbon Vegetation Characteristics –1 SourceSink (g CH 4 m -2 year -1 )

Vegetation type;Snow pack; Soil moisture Soil temperature Terrestrial Ecosystem Model (TEM) couples biogeochemistry and soil thermal dynamics

Snow Thawing front Moss Peat Mineral Temperature update Moisture update Moss growth Fire disturbance

Decadal patterns of simulated soil temperature in top 10 cm of of mineral soil in black spruce forests of interior Alaska for Different topographic positions (Yi, McGuire, and Kasischke). Field observations and modeling have shown that permafrost in black spruce stands on different topographic positions have been warming since the mid-1960s, which means that over this time period, deeper duff layers in black spruce forests have become warmer and drier.

34 cm 28 cm 25 cm 0 cm 12 cm Control of depth to permafrost and soil temperature by the forest floor in Black spruce/Feathermoss Communities C.T. Dyrness 1982 USDA, Forest Service, Pacific Northwest Forest and Range Experiment Station, Research Note: PNW-396 Site: Washington Creek Fire Ecology Experimental Area, north of Fairbanks

Effects of Org Thickness on active layer depth (S. Yi) 6 cm : moss 14 cm : peat 0 cm : moss 14 cm : peat 0 cm : moss 9 cm : peat 0 cm : moss 0 cm : peat DFCC site Thawing frontFreezing front

Kougarok burn site (k2) Biome: Tussock Tundra Lat: o N Lon: o W Elev: 110 m Aspect: south Slope: 3 o Fire History: 1971, 2002

K2 soil profiles Before Fire Upper organic layer –Thick : 4 cm –Porosity : 90 Lower organic layer –Thick: 10 cm –Porosity : 80 Mineral –Sand :20, Silt: 58, Clay :22 After Fire Upper organic layer –Thick : 0 cm –Porosity : 90 Lower organic layer –Thick: 5 cm –Porosity : 80 Mineral –Sand :20, Silt: 58, Clay :22 Run from 1901 to The initial soil structure uses the one before fire. At July 2002, top two organic layers are removed, and only 5 cm organic layer is left. No other changes have been made at fire event.

Soil Temperature Simulation fire X-axis: doy Y-axis: temperature (degc)

Soil Moisture Simulation --surface X-axis: doy Y-axis: soil wetness (%) fire

Soil Moisture Simulation --shallow layer X-axis: doy Y-axis: soil wetness (%) fire

Soil Moisture Simulation --deep layer X-axis: doy Y-axis: soil wetness (%) fire

Implementation of fire disturbance Thawing front Moss Peat Mineral Slope Aspect Elevation Soil temperature Moisture Active layer depth Other issues affecting burn depth Burn depth

Implementation of moss growth and organic matter conversion Vegetation biomass Moss biomassMoss thickness livedead fibric mesic humic mineral abovebelow

Observations and model predictions at the Alaska-Canada scale, (R 2 = 0.82 (p<0.0001) for period )

Vegetation Soil Organic Matter Soil Inorganic Carbon CO 2(g) abvR CO 2(aq) HCO 3 - CO 3 -2 rootR RHRH CO 2 (g) Alkalinity CO 2(aq) Shaded area = Modified TEM soilR DOC Stream Export CO 2 (g) Chemical Weathering POC GPP erodePOC leachDOC harvest leachCO2leachALK evadeCO2 fire Delivery of Carbon to Marine Environments

RegionLeachDOC (Tg C yr -1 ) Raymond et al. (Tg C yr -1 ) Ob’ Yenisei Lena Mackenzie Yukon Arctic Rivers Pan-Arctic Rivers * Comparison of TEM Estimated DOC Leaching Rates during the 1990s to Measured DOC Export from Arctic Rivers *includes rivers draining directly into the Arctic Ocean, the Arctic Archipeligo, Hudson Bay, and the Bering strait D. Kicklighter, J. Melillo, and A.D. McGuire

Depth to water table (DTW) (m) of 1990’s July Dynamic simulation of wetlands in the Yukon River Drainage Basin using a TOPMODEL approach M. Stieglitz, D. Kicklighter, J. Melillo, and A.D. McGuire

Feedbacks Involving Snow Responses Retrospective Studies of Carbon and Energy Feedbacks Vulnerability of Climate System to Changes in Snow

-Examine patterns in snowmelt, snow return, and the duration of the snow free season as they impact atmospheric heating -Perform analyses for the 1910 –1940 and time periods over the arctic-boreal land area above 50º N at a half-degree latitude by longitude spatial resolution E. Euskirchen and A.D. McGuire

< >0.1 Days per year shorterDays per year longer Change in the duration of snow covered ground (anomaly): Between , the number of days of snow covered ground decreased by an estimated 2.5 days per decade across the pan-Arctic From Euskirchen et al. in press.

W m -2 decade -1 Cooling Heating Across the pan-Arctic, an overall reduction in the duration of snow covered ground by ~2.5 days per decade resulted in atmospheric heating of ~1.0 W m -2 per decade. Changes in atmospheric heating due to changes in the snow season, From Euskirchen et al. in press. Heating magnified in period Spring more important than autumn Tundra important (high albedo contrast)

Feedbacks Involving Responses of Vegetation Composition and Structure

Energy budget feedbacks to regional summer climate Feedbacks from vegetation change –Tussock to shrub transition: 3.9 W/m 2 –Tussock to forest transition: 5.0 W/m 2 2% change in solar constant: 4.6 W/m 2 –(glacial to interglacial change) Doubling atmospheric CO 2 : 4.4 W/m 2 Chapin and McFadden

Soil thermal model coupled to TEM DVM - TEM MVP – TEM includes leaf, wood, and root components Vegetation type;Snow pack; Soil moisture Soil temperature Multiple vegetation pools Dynamic vegetation model Soil Temps. at Different Depths Upper Boundary Conditions Snow Cover Moss & litter Frozen Ground Thawed Ground Frozen Ground Lower Boundary Conditions Heat Conduction Organic Soil Mineral Soil Prescribed Temperature Prescribed Temperature Snow Depth Moss Depth Organic Soil Depth Mineral Soil Depth Moving phase plane Heat balance surface Lower boundary Heat Conduction E. Euskirchen and A.D. McGuire

Warming of 12°C (SRES A2 Scenario) Warming of 6°C (SRES B2 Scenario) Warming of 2°C (SRES B1 Scenario) Mean (± standard deviation) percent change in plant net primary productivity between Dynamic Vegetation Model coupled to the Terrestrial Ecosystem Model Changes in plant productivity between 2003 – 2100 in northern Alaska: Large variation among the plant functional types in the shrub tundra, represented with the error bars. Boreal forest Shrub tundra Sedge tundra E. Euskirchen and A.D. McGuire

Estimated Cumulative Area Burned for Interior Alaska A2 Hadley B2 Hadley CRU A2 PCM B2 PCM Area Burned (km^2)

deciduous white spruce black spruce A2 Hadley (Most Area Burned) Single Replicate

Estimated Change in Summer Energy Budget A2 Hadley CRU A2 PCM B2 PCM B2 Hadley

Liu et al., 2005 Changes in surface albedo in response to fire Grey line = Recent burn Black line = Control

Coupling of DVM-TEM with CCSM3.0 Coupling of DVM/TEM and frozen soil/permafrost module within CCSM Mölders, Euskirchen, and McGuire