Fire and Climate Interactions in the Boreal Zone

FireFire andand ClimateClimate InteractionsInteractions inin thethe BorealBoreal ZoneZone

Amber J. Soja with donations from Nadezda M. Tchebakova, Brian Stocks, Susan G. Conard, Anatoly I. Sukhinin, Douglas J. McRae, Paul W. Stackhouse Jr., Pavel Ya. Groisman, Tatiana V. Loboda, and Ivan A. Csiszar Photo courtesy of Land Cover Land Use Change, 2007 Brian Stocks WhyWhy areare RussianRussian borealboreal ecosystemsecosystems important?important? The largest temperature increases are predicted to occur in northern hemisphere upper latitudes (Cubash et al., 2001).

The greatest temperature increases are currently found in the cold, dry air masses over central Siberia (Balling et al., 1998).

Boreal forests hold the largest reservoir of terrestrial carbon (Apps et al., 1993; Zoltai and Martikainen, 1996; Alexeyev and Birdey, 1998).

Two-thirds of the boreal forest are located in Russia (25% of the worlds forest). ClimateClimate ChangeChange RealitiesRealities andand ProjectionsProjections

¾ GCMs project 1.4 – 5.80 C increase in global mean temperature by 2100 ¾ Observed increases across Observations above – summer temperature west-central Canada and Siberia over past 40 years ¾ Greatest increases will be at high latitudes, over land and winter/spring ¾ Projected increases in extreme weather and extreme fire weather

Predicted changes 2080-2100 Mean seasonal temperature changes 1965 to 2004 . winter Temperatures are increasing, particularly in the Northern Hemisphere winter and spring, which leads to longer growing spring seasons, increased potential evapotranspiration and extreme fire weather. [Groisman et al., 2007; Jones and summer Moberg, 2003, updated] Analysis of Inter-model Consistency in Regional Warming Atmospheric-Ocean General Circulation Models (AOGCM) are in agreement that warming in Northern Eurasia could be in excess of 40% above the global average (at least 7 of 9 models) (IPCC, 2001). BorealBoreal forestsforests andand firefire

Boreal forest communities are particularly responsive to climate change. (Stocks and Street, 1982; Clark, 1988; Payette and Gagnon, 1985; Flannigan and Harrington, 1988; Sirois, 1992)

Synoptic weather patterns are the major regional factor controlling the occurrence of fire. (Stocks and Street, 1982; Pastor and Mladenoff, 1992) Wildfire is an integral component of boreal landscapes.

Dominant driver of ecological processes in boreal ecosystems (Lutz, 1956; Rowe and Scotter, 1973; Van Cleve and Viereck, 1981; MacLean et al., 1983; Bonan and Shugart, 1989).

Wildfire maintains age structure, species composition, and the floristic diversity of boreal forests by resolving the beginning and end of successional processes (Zackrisson, 1977; Van Wagner, 1978; Heinselman, 1981; Antonovski et al., 1992).

In a landscape-scale steady-state system, wildfire acts as a disturbing agent that maintains the current stability, diversity and mosaic structure of boreal ecosystems (Loucks, 1970; Viereck and Schandelmeier, 1980; Romme, 1982; Shugart et al., 1991). WhatWhat isis expected?expected?

Under current climate change scenarios…

Boreal forest fire frequency and area burned are expected to increase (25 – 50%) (Overpeck et al., 1990; Flannigan and Van Wagner 1990, 1991; Wotton and Flannigan 1993; Stocks et al. 1998, 2000; Flannigan et al. 2001).

Increased ignition from lightning (20-40%) (Fosberg et al. 1990, 1996; Price and Rind 1994) WhatWhat isis expected?expected? Under current climate change scenarios…

Increased fire season length (up to 51 days, average 30 days) (Streeet 1989; Wotton and Flannigan 1993; Stocks et al. 1998)

Increased fire weather severity (46% average) (Street 1989; Flannigan and Van Wagner 1991, Stocks et al. 1998)

Fire weather severity is predicted to be greater in Russia than in North America (Stocks and Lynham 1996) I. Climate holds the ultimate key to altering boreal ecosystems (temperature and precipitation).

II. Climate has the potential to affect boreal fire regimes by:

(1) altering species composition;

(2) altering fire ignitions from lightening and;

(3) altering weather conditions conducive to fire. FutureFuture FireFire DangerDanger LevelsLevels ¾ All GCM and RCM projections show increase in strength and areal extent of extreme fire danger ¾ Increase in fire season length ¾ Increased weather variability and extremes at regional scale RoleRole ofof FutureFuture ClimateClimate ChangeChange

Historical Fire Weather GCM: 2X CO2

Courtesy Brian Stocks Fires across Siberia - Large scale synoptic patterns - Seasonal patterns Soja et al., IJRS, 2004 Soja et al., IJRS, 2004 65 - -70 65 60 5 - 60 5 - 55 0 50 - 5 5 Degrees north 4 - 45

0 Average counts from 1999 and 2000

9 1

0 8

7 /

1 7

6 /

0 6

6 /

- 6

(month/day)

2 10-day range 6

5 /

- 5

3 5

2 5

1 4

- 4

0.00 0.50 0.40 0.30 0.20 0.10 0 /

4 fire counts fire Proportion of of Proportion Soja et al., IJRS, 2004 70,000 2001 60,000 2002 50,000 2003 40,000 2004 30,000

20,000

10,000 number of fire detections of fire number 0 Jan18000 Feb Mar Apr May June July Aug Sept Oct Nov Dec Loboda et al., month 2007 15000

12000 t 9000 Seasonal 6000

fire coun Patterns 3000 of Fire in 0 1 3 5 6 7 9 0 1 2 /2 Boreal 4/ 5/ 7/0 - - 5/0 - - 6/1 - - 7 01 3 15 06 19 4/ 4/2 5/ 6/ 6/28 7/ 8/10 - 8/2 2000 Siberia 1999 Soja et al., IJRS, 2004 month/day 25 AreaArea burnedburned inin SiberiaSiberia Avialesookrana 20 R2 = 0.1507 Ground based on Avialesookrana ground reports Sukachev Satellite h 15 1992 Sate llite

10 1987 Sate llite

Area burned (M 1998 Sate llite 5

Linear 1980 1985 1990 1995 2000 (Avialesookrana 0 Ground ) Area Burned Annually (M ha) Area Burned

Years In Siberia, 7 of the last 9 years have resulted in extreme fire seasons, and extreme fire years have also been more frequent in both Alaska and Canada. AreaArea8 burnedburned inin borealboreal NorthNorth AmericaAmerica 7 Alaska Canada 6 Alaska R 2 = 0.03 5 Canada R 2 = 0.08 North America R 2 = 0.17 4

1 Area Burned Annually (M ha) Area Burned

- 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005

Soja et al., 2007 FireFire RegimesRegimes VaryVary WidelyWidely

Fires burning in Sakha (Yakutia), August 2002 These fires burned from late April through early October Over 5 M ha burned in Sakha, which accounts for almost half of the Siberian 2002 total carbon emissions, perhaps 10% of the total global carbon emissions from forest and grassland burning (Andreae and Merlet, 2001 Soja et al., JGR 2004). Extreme fire event that burned in Siberia, Peak in Aerosol Index July 26, 2006. Landsat (30 m resolution) data showing the fire scars in Evenkia, Siberia.

Large wildfire-induced aerosol event from Siberia. Seasonal changes in the 2006 fire hazard index (red line), with rainfall (blue bars) for the 7000 36 6340 32 Boguchany 6000 Осадки, мм 6199 Показатель ПВ-1 28 5000 24 weather 4000 20 -1

ПВ 3000 16 station . 12 2000 8 1000 4 0 0 3.8.06 8.8.06 4.7.06 9.7.06 4.6.06 9.6.06 13.8.06 14.7.06 19.7.06 24.7.06 29.7.06 14.6.06 19.6.06 24.6.06 29.6.06 10.5.06 15.5.06 20.5.06 25.5.06 30.5.06 Дата On July 26, 2006, the Aerosol Index (AI) peaked in Siberia due to an extreme wildfire event, which traveled northwest over the North Pole. During this time, CALIPSO captured the vertical height and structure of the smoke plume. - CALIPSO combines an active lidar instrument with passive infrared and visible imagers to probe the vertical structure and properties of thin clouds and aerosols. - 30 meter vertical resolution, 100 m swath. at is Wh ? - Data extends from sea level to 30 km. IPSO CAL - Launched on April 28, 2006 with CloudSat. http://www-calipso.larc.nasa.gov/ Coincidence in Siberian wildfire, MODIS Aerosol Optical Depth (AOD) and the CALIPSO path.

CALIPSO flight track (green & red tracks shown below) AM path About 12:00 to 13:00 pm local time PM path Begin: 2006/07/26 05:22:09.5520 UTC End: 2006/07/26 06:14:29.6124

CALIPSO data Expanded view, next slide FirstFirst ViewsViews ofof CALIPSOCALIPSO datadata

CALIPSO captures both the maximum altitude of the smoke plume at about 6 km and the vertical smoke structure, as it evolves over time. Initial albedo analysis from Sakha, Scar Scar Siberia 3 shows 2 decreased albedo for about 4 to 5 years, then it begins to increase.

Gray-scale AVHRR image of band 2. This July, 17 2002 image shows dark fire scars (black) and active fires with smoke steaming from them (whitish). Remarkably, increased reflectance from the 1986 fire scars is still apparent, 16 years after the fire events. PredictedPredicted cchangehangess inin 20902090 precipitationprecipitation forfor SiberiaSiberia

(Hadley(Hadley Center,Center, HadCMHadCM3GGa1)GGa1)

Precipitation (-4/+25%)

January temperature (+4/+9oC)

July temperature (+4/+6oC)

Tchbakova an Parfenova, 2000; Parfenova and Tchbakova, 2001 ClimaticClimatic anomaliesanomalies byby 20002000 inin thethe northnorth (upper)(upper) andand inin thethe southsouth (lower)(lower) ofof CentralCentral SiberiaSiberia.. T, 0C, January T, 0C, July Precipitation, mm

74 74 74 Хатанга Хатанга 72 72 72 Хатанга 40 6 10 70 70 70 30 Игарка 5 68 Игарка 8 68 68 4 20 Игарка 6 66 66 3 66 10 Тура Тура 4 64 64 2 64 0 Тура 2 62 62 1 62 -10 0 0 60 -20 60 Енисейск Богучаны 60 Енисейск Богучаны -2 -1 -30 58 58 58 Енисейск Богучаны 56 56 56 86 88 90 92 94 96 98 100 102 104 86 88 90 92 94 96 98 100 102 104 86 88 90 92 94 96 98 100 102 104 Красноярск Красноярск Красноярск 56 56 56

55 55 40 Шира 55 Шира 10 6 Шира 30 5 54 Абакан 8 54 Абакан 54 Абакан 20 4 6 53 53 3 53 10 4 2 0 52 Кызыл 52 Кызыл 2 1 52 Кызыл -10 0 0 -20 51 51 51 Мугур_Аксы Эрзин -2 Мугур_Аксы Эрзин -1 Мугур_Аксы Эрзин -30 50 50 50 89 90 91 92 93 94 95 96 89 90 91 92 93 94 95 96 89 90 91 92 93 94 95 96 56

55 Climate change in the 8 54 Sayan Mountains: e 4 d tu 53 ti a observations,

L 0 52 -4 1980-2000 (left) and 51 -8 predicted (right) 50 89 90 91 92 93 94 95 96 Lo ngitude by a Hadley Centre Annual precipitation (%) scenario 56 56 (HadCM3GGa1) for 55 55 6 6 2090. 54 54 4 4 ude ude t t i i 53 53 t

La 2

Lat 2 52 52 Winter temperatures 0 0 51 51 -2 -2 have already 50 50 89 90 91 92 93 94 95 96 89 90 91 92 93 94 95 96 exceeded 2090 model Lon gitude Lo ngitu de estimates, while January temperature (oC) 56 56 summer temperatures

55 6 55 6 have not. Patterns of 5 54 54 5 4 4 precipitation are e d ude t i 53 3 tu 53 3 ti a Lat 2 L 2 currently difficult to 52 1 52 1 0 predict, particularly at 51 0 -1 51 -1 the GCM scale. 50 50 89 90 91 92 93 94 95 96 89 90 91 92 93 94 95 96 Lon gitude Long itude July temperature (oC) Soja, Tchebakova et al., 2007 Vegetation distribution in Siberia: (A) current and (B) future (2100) based on Hadley scenario (HadCM3GGal) (IPCC, 1996). Water (0), tundra (1), forest–tundra (2), N. dark taiga (3) and N. light taiga (4), middle dark taiga (5) M. light taiga (6), S. dark taiga (7) and S. light taiga (8), forest–steppe (9), steppe (10), semidesert (11), broadleaved (12), temperate forest– steppe (13) and temperate steppe (14).

Tchebakova and Parfenova, 2000, 2001 ModeledModeled biomebiome areaarea changechange (2090)(2090) BiomeBiome CurrentCurrent Area,Area, AreaArea change,change, 104 sq. km 104 sq. km (%) Tundra 110.8 7.5 (- 93%) Forest-Tundra 91.3 12.2 (- 87%) Northern dark taiga 1.9 1.1 (- 42%) Northern light taiga 191.1 48.6 (- 75%) Middle dark taiga 52.7 10.2 (- 81%) Middle light taiga 237.0 66.2 (- 72%) Southern dark taiga 108.2 127.3 (+18%) Southern light taiga 98.6 87.0 (- 12%) Forest-steppe 62.6 328.1 (+5-fold) Steppe 139.2 176.5 (+27%) Temperate biomes 11.8 328 (+30-fold) “Hot spots” of forest cover changes by 2090

Green – new forest habitats, Orange – new steppe habitats, white – no change in forest vegetation Outlook

North America ¾ Increasing fire activity and severity ¾ Increasing vulnerabilities ¾ Diminishing marginal returns on more expenditures

Russia ¾ Fire management program without funding ¾ Widespread forest exploitation of forests exacerbating fire problems ¾ No funding for forest protection despite fact that natural resources (mining/gas and oil) are a major driver of the Russian economy ¾ Increasing vulnerabilities ¾ Shift to regional fire control – will it 1 year after happen? Will it be funded? severe crown fire ¾ Major uncertainties going forward Conclusions (1 of 2)

Fire weather and fire danger has already increased across boreal North America and Siberia.

Boreal forests and fire are largely under the control of weather and climate, and they feedback to the climate system, in terms of fire emissions and changes in the Solar Radiation Budget (albedo).

There is currently evidence of climate- and fire- induced change across Siberia, particularly at treelines (northern, southern, and the upland and lowland mountainous treelines) Conclusions (2 of 2)

Wildfire is a catalyst that serves two basic purposes in boreal forest: 1)a mechanism to maintain stability and diversity in equilibrium with the climate and; 2) a mechanism by which forests move more rapidly towards equilibrium with climate.

“An altered fire regime may be more important than the direct effects of climate change in forcing or facilitating species distribution changes, migration, substitution, and extinction.” Weber and Flannigan (1997) Thank-you!