Satellite observations of high northern latitude vegetation

OPEN ACCESS

Satellite observations of high northern latitudevegetation productivity changes between 1982 and2008 ecological variability and regionaldifferencesTo cite this article Pieter S A Beck and Scott J Goetz 2011 Environ Res Lett 6 045501

View the article online for updates and enhancements

Related contentEcosystem responses to recent climatechange and fire disturbance at northernhighlatitudes observations and modelresults contrasting northern Eurasia andNorthAmericaS J Goetz M C Mack K R Gurney et al

Spatial variation in vegetation productivitytrends fire disturbance and soil carbonacross arctic-boreal permafrostecosystemsMichael M Loranty Wil Lieberman-CribbinLogan T Berner et al

Canadian boreal forest greening andbrowning trends an analysis ofbiogeographic patterns and the relativeroles of disturbance versus climate driversDamien Sulla-Menashe Curtis EWoodcock and Mark A Friedl

Recent citationsClimatic Change and Human ActivitiesLink to Vegetation Dynamics in the AralSea Basin Using NDVITimur Berdimbetov et al

Disturbance suppresses the abovegroundcarbon sink in North American borealforestsJonathan A Wang et al

Arctic tundra shrubification a review ofmechanisms and impacts on ecosystemcarbon balanceZelalem A Mekonnen et al

This content was downloaded from IP address 811903581 on 30082021 at 1843

IOP PUBLISHING ENVIRONMENTAL RESEARCH LETTERS

Environ Res Lett 7 (2012) 029501 (3pp) doi1010881748-932672029501

Corrigendum Satellite observations ofhigh northern latitude vegetationproductivity changes between 1982 and2008 ecological variability and regionaldifferences2011 Environ Res Lett 6 045501

Pieter S A Beck and Scott J Goetz

The Woods Hole Research Center 149 Woods Hole Road Falmouth MA 02540 USA

E-mail pbeckwhrcorg

Received 26 March 2012Accepted for publication 26 March 2012Published 18 April 2012Online at stacksioporgERL7029501

Due to an error in our implementation of the Vogelsangtest (Vogelsang 1998) the criterion used to determinewhere trends in remotely sensed gross productivity (Prs)were deemed deterministic at α = 005 was overly le-nient This resulted in an overestimation of the totalgeographical area exhibiting deterministic Prs trends at

Figure 2 Trends in remotely sensed gross productivity (Prs) between 1982 and 2008 Gray shading indicates the trend wasnon-deterministic based on a Vogelsang test (α = 005 in the top panel and α = 01 in the bottom panel) Areas in white were excludedfrom the analysis as described in the text

α = 005 This error affected figures 2 3 4 and 5 butdoes not modify any of the conclusions drawn from theanalysis

Corrected versions of the figures are included belowIn addition but unrelated titles of the panels in figure 3

were inadvertently omitted They are included here

11748-932612029501+03$3300 ccopy 2012 IOP Publishing Ltd Printed in the UK

Environ Res Lett 7 (2012) 029501 Corrigendum

Figure 3 Areal fraction of the non-anthropogenic vegetated landscape displaying statistically significant deterministic changes in remotelysensed gross productivity (Prs with bold lines representing α = 005 and thinner lines representing α = 01) when consideringprogressively longer time series since 1982 Tundra and boreal biomes were outlined using FAO (2001) (a) Tundra in North America(b) Tundra in Eurasia (c) Boreal forest in North America and (d) boreal forest in Eurasia

Figure 4 MODIS tree cover in areas with and without statistically significant deterministic trends (α = 005 in the top panel and α = 01in the bottom panel) in Prs between 1982 and 2008 in North American and Siberian tundra and boreal areas (as defined by a globaleco-floristic zone map (see text for sources) KruskalndashWallis tests showed statistically significant differences (p lt 0001) in tree coverbetween areas with positive and negative Prs trends in both North American and Eurasia

Figure 5 Proportion of area in boreal Alaska showing decreases or increases in remotely sensed gross productivity (Prs) or nodeterministic trend in productivity (α = 005 in the top panel and α = 01 in the bottom panel) over the period 1982ndash2008 along a gradientfrom evergreen to deciduous tree dominance as mapped by (Beck et al 2011a) The deciduous fraction ranges from 0 which representspurely evergreen stands to 100 which represents purely deciduous stands

References

Vogelsang T J 1998 Trend function hypothesis testing in thepresence of serial correlation Econometrica 66 123ndash48

CorrigendumSatellite observations of high northern latitude vegetation productivity changes between 1982 and 2008ecological variability and regional differencesPieter S A Beck and Scott J Goetz 2011 Environ Res Lett 6 045501

Received 2 November 2011Published 23 November 2011

In the first paragraph of the section lsquo2 Data sets and methodsrsquo the Normalized Difference Vegetation Index (NDVI) data setused was incorrectly referred to as GIMMS-NDVI version 3G with a 0084 spatial resolution This should be corrected toGIMMS-NDVI version G with a 007 spatial resolution

Accordingly the acknowledgement should state lsquoWe would like to thank Jim Tucker and Jorge Pinzon for providing theGIMMS version G datarsquo instead of lsquoWe would like to thank Jorge Pinzon for providing the GIMMS 3G datarsquo

1748-932611049501+01$3300 1 copy 2011 IOP Publishing Ltd Printed in the UK

Satellite observations of high northernlatitude vegetation productivity changesbetween 1982 and 2008 ecologicalvariability and regional differencesPieter S A Beck and Scott J Goetz

E-mail pbeckwhrcorg

Received 31 July 2011Accepted for publication 14 September 2011Published 11 October 2011Online at stacksioporgERL6045501

AbstractTo assess ongoing changes in high latitude vegetation productivity we compared spatiotemporalpatterns in remotely sensed vegetation productivity in the tundra and boreal zones of NorthAmerica and Eurasia We compared the long-term GIMMS (Global Inventory Modeling andMapping Studies) NDVI (Normalized Difference Vegetation Index) to the more recent andadvanced MODIS (Moderate Resolution Imaging Spectroradiometer) NDVI data set andmapped circumpolar trends in a gross productivity metric derived from the former We thenanalyzed how temporal changes in productivity differed along an evergreenndashdeciduous gradientin boreal Alaska along a shrub cover gradient in Arctic Alaska and during succession after firein boreal North America and northern Eurasia We find that the earlier reported contrastbetween trends of increasing tundra and decreasing boreal forest productivity has amplified inrecent years particularly in North America Decreases in boreal forest productivity are mostprominent in areas of denser tree cover and particularly in Alaska evergreen forest stands Onthe North Slope of Alaska however increases in tundra productivity do not appear restricted toareas of higher shrub cover which suggests enhanced productivity across functional vegetationtypes Differences in the recovery of post-disturbance vegetation productivity between NorthAmerica and Eurasia are described using burn chronosequences and the potential factorsdriving regional differences are discussed

Keywords NDVI GIMMS MODIS Arctic tundra boreal forest fire shrubs greeningbrowning

1 Introduction

Ongoing and projected acceleration of climate change at highlatitudes is impacting a broad range of ecosystem processeswith multi-faceted implications for the regional carbon balance(Hinzman et al 2005 Soja et al 2007 Groisman and Soja2009) Understanding these changes is important becausethey will feed back to climate directly through their effectson atmospheric CO2 concentrations and also indirectly byaltering terrestrial energy budgets and hydrologic cycles

(Chapin et al 2008 Wookey et al 2009) Changes inenvironmental conditions associated with warming such as alengthening of the growing season have been directly linkedto changes in vegetation productivity and composition (Angertet al 2005 Goetz et al 2005) For example changes in shrubcover have been documented in situ in recent decades in tundraareas (Tape et al 2006) Increased shrub cover on the tundramay alter nutrient cycling (Shaver and Chapin 1991 Sturmet al 2005b) hydrology (Sturm et al 2001) trophic interactions(Joly et al 2007 Tape et al 2010) and permafrost dynamics

1748-932611045501+10$3300 copy 2011 IOP Publishing Ltd Printed in the UK1

Environ Res Lett 6 (2011) 045501 P S A Beck and S J Goetz

(Sturm et al 2005b) If widespread a shift toward a greaterdominance of shrubs might affect the carbon balance (Shaveret al 1992 Mack et al 2004) and radiation budget (Chapin et al2005 Sturm et al 2005a Loranty et al 2011) of the tundrabiome amongst others through shifts in fire activity (Higueraet al 2008)

In the boreal forest biome fire is currently the dominantdisturbance agent and exerts strong controls on the carbonbalance (Bond-Lamberty et al 2007 Kasischke et al 2010Beck et al 2011a Turetsky et al 2011) Fire is also associatedwith broad shifts in vegetation composition and has beenlinked with historical evergreen tree establishment in Alaska(Lynch et al 2003 Johnstone et al 2010a) Fire frequencyand severity have increased with recent climate warming inboreal North America (Gillett et al 2004 Kasischke andTuretsky 2006) and Siberia (Groisman et al 2007 Soja et al2007) Alaskan boreal forests contain evergreen coniferswhich dominate older stands in particular as well as deciduousbroadleaf trees which can dominate specific topographicalpositions and areas recovering from fire disturbance (Chapinet al 2006 Mack et al 2008) Moreover areas exposed tosevere burning show increased deciduous cover for decadesafter fire (Johnstone et al 2010b Beck et al 2011a Barrettet al 2011) In addition to altering the regional carbonbalance a transition from evergreen-dominated to deciduousforest or vice versa would substantially alter albedo andevapotranspiration (Bala et al 2007 Bonan 2008 Shumanet al 2011) as well as profoundly alter ecological communities(Werner et al 2006 Macdonald and Fenniak 2007 Wannebo-Nilsen et al 2010) These examples illustrate the need fora comprehensive examination of the magnitude and directionof changes in primary productivity across the northern highlatitudes as a result of altered ecosystem processes associatedwith climate warming This is particularly true in the caseof the biophysical implications of climate change such asland-atmosphere feedbacks associated with shifts in energyand carbon balance (McGuire et al 2009 Wookey et al2009 Euskirchen et al 2010) Satellite observations haveprovided unique insights into global primary productivitypatterns and changes therein Remotely sensed proxies ofproductivity (Prs) revealed widespread increases in aboveground primary production associated with climate warming athigh latitudes in the mid 1990s (Myneni et al 1997) A decadelater divergent responses in Prs were discovered betweenboreal and tundra biomes across North America (Goetz et al2005) and the circumpolar region (Bunn and Goetz 2006Piao et al 2011) Increases in Prs in high latitudes havebeen attributed to a release of cold temperature constraintson photosynthesis (Angert et al 2005) whereas decreasesin Prs or so-called lsquobrowningrsquo trends have been linked todrought stress associated with higher vapor pressure deficitsduring summer (Bunn et al 2005 Lotsch et al 2005 Verbyla2008) Field measurements since the 1990s support theseinterpretations showing increases in shrub growth in Alaskantundra (Tape et al 2006) and historically low growth rates inblack and white spruce trees in Alaskan boreal forests (Barberet al 2000 Beck et al 2011c)

A refined interpretation of mapped changes in Prs andultimately the attribution of trends and patterns to particular

ecological process are needed (McGuire et al 2009) The roleof different vegetation functional types differentially drivingthe changes in Prs observed at coarse spatial resolution islargely unknown other than the broad biome differencespreviously noted Such attribution is essential to determinewhether there are links between local observations such astundra shrub expansion and Arctic-wide remotely sensedobservations Similarly attribution of trends in Prs to particularfunctional types is needed to forecast changes in more complexecological processes associated with changes in productivityAt the same time the attribution of observed trends todifferent climatic and non- or indirect climatic drivers such asdisturbance is needed to estimate biophysical implications ofchanges in Prs including energy budgets and carbon storagein vegetation and soils (Chapin et al 2000 Randerson et al2006) In general however assessing and linking Prs dynamicswith the process that might be driving them is hampered by thediscrepancy in spatial resolution between in situ measurementsand long-term systematic satellite data records which aregenerally on the order of hundreds of meters (Beck et al 2007)

Here we assessed differences in Prs changes since 1982between regions and vegetation functional groups Wecompared the high latitudes of Eurasia where forests aredominated by larch (Larix sibirica and L gmelinii) adeciduous conifer and North America where forests aredominated by evergreen conifer species In Alaskarsquos borealforest we analyzed trends in Prs along a gradient of deciduousto evergreen forest cover On Alaskarsquos North Slope ie northof the Brooks Range and the largest expanse of tundra inthe USA we summarized changes in summer NDVI alonga gradient of shrub cover to determine to which degreevegetation functional types exhibit differential responses toclimate warming Finally we compared regional differencesin Prs responses to fire in boreal regions of North America andSiberia to further dissect regional differences in productivitytrends

2 Data sets and methods

Advanced very high resolution radiometer (AVHRR) mea-surements provide the longest record of continuous globalsatellite measurements sensitive to live green vegetation Tocreate a consistent data set from these measurements for globalchange research the global inventory modeling and mappingstudies (GIMMS) creates global maps of normalized differencevegetation index (NDVI) at 0084 spatial resolution and withtwice-monthly frequency (GIMMS-NDVI version 3G Tuckeret al 2005) Because the GIMMS record starts in July 1981and is well documented it is most frequently used to assesslong-term change in global and regional terrestrial vegetationproductivity

To validate the interannual NDVI signal in the GIMMSdata set trends in summer NDVI (JulyndashAugust) over highlatitudes of North America were compared to NDVI measure-ments from the moderate resolution imaging spectroradiometer(MODIS) aboard the Terra satellite launched in 1999 Thesemonthly MODIS NDVI data (MOD13A3 Huete et al 2002)have a 1000 m spatial resolution and are of higher radiometric

and geometric quality than the GIMMS data They werefiltered to retain only data of the highest quality based on thequality assessment flags provided with the MOD13A3 dataFor the comparison temporal linear trends in average yearlysummer NDVI for the MODIS-GIMMS overlap period (2001ndash8) were calculated along with their associated uncertainty usingordinary least squares regression

We estimated trends in Prs across northern high latitudesover progressively longer periods of 21ndash27 years since 1982and calculated the aerial fraction of positively and negativelytrending areas in the 7 nested time series Temporal patterns inthese fractions were then compared in North America versusEurasia and in tundra versus boreal biomes to reveal anyregional differences in recent productivity shifts Tundrawas delimited as the lsquopolarrsquo class and the boreal biome asthe lsquoboreal coniferous forestrsquo lsquoboreal tundra woodlandrsquo andlsquoboreal mountain systemrsquo classes mapped by the Food andAgriculture Organization of the United Nations (FAO 2001)

Prior to estimating trends in Prs we applied a spatialfilter to limit the analysis to areas dominated by non-anthropogenic vegetation cover as mapped in the InternationalGeospherendashBiosphere Programme (IGBP) classification of00042 (sim1 km) MODIS reflectance data (MOD12Q1 Friedlet al 2002) The spatial filter is designed to exclude areasdominated by man-made land cover but it allows up to60 of non-vegetated cover within a GIMMS grid cell asnon-vegetated areas have invariant NDVI In practice thefilters excluded GIMMS grid cells where more than 40was classified as non-vegetated (IGBP 0 15ndash16) or wherevegetated land cover was not at least three times larger thananthropogenic land cover (IGBP 12ndash14)

To map trends in Prs over the 21ndash27 year periods first thegrowing season length (GSL) was estimated for each GIMMSgrid cell p The GSL was defined as two thirds of the periodbetween the latest lsquostart of greeningrsquo date and the earliest lsquostartof dormancyrsquo recorded between 2001 and 2004 in the MODISLand Cover Dynamics product (MOD12Q2 Zhang et al 2006Beck et al 2011c) For every year y Prs was then calculated ineach grid cell p as the highest mean NDVI observed in the 24NDVI values over a period of length GSLp (Berner et al 2011)

Prspy = max

(GSLminus1

t+GSLpsumi=t+1

NDVIpyi

where t = 0 1 2 24 minus GSL (1)

This moving window approach to estimating gross productivityis robust to changes in the timing of the growing season butwill not capture changes in its length if they are uncorrelatedwith changes in summer productivity To detect deterministicie non-stochastic trends in Prs we applied the Vogelsangtest (Vogelsang 1998) to each time series with statisticalsignificance set at α = 005 The test prevents autocorrelationin the series or abrupt disturbance events from generatingartificial trends (Goetz et al 2005)

21 Attribution of the trends

To discern how productivity responses differ between highlatitude land cover types tree cover as mapped by Hansen

et al (2003) using MODIS data was compared between areasof no positive or negative trends in Prs over the 1982ndash2008period In Alaskan boreal forest we further assessed whetherevergreen and deciduous vegetation dominance differentiallyinfluenced the observed changes in Prs by summarizing trendsalong a gradient of evergreen to deciduous vegetation coverusing a MODIS-derived map of forest composition producedand described by Beck et al (2011a) Areas that had burnedsince 1982 as delineated in a database of Alaskan fireperimeters produced by the Bureau of Land ManagementAlaska Fire Service (AFS httpagdcusgsgovdatablmfire)were identified and excluded from the analysis

To investigate the extent to which positive trends in tundraproductivity were attributable to increased shrub growth wetested on the North Slope of Alaska if summer (July andAugust) NDVI trends were more pronounced in areas ofgreater shrub density Thus yearly summer NDVI wassummarized along the gradient of shrub cover present onAlaskarsquos North Slope which was recently mapped as apercentage of the surface and at 30 m resolution (describedin more detail by Beck et al (2011b)) Summer NDVI wasused because the MOD12Q2 phenology data and the derivedPrs maps have gaps on the North Slope of Alaska and becausethe JulyndashAugust period represents the growing season well inthis region

The GIMMS-NDVI time series was also used to describeand compare vegetation recovery after fire in North Americaand Siberia A circumpolar data base of yearly burnedarea was compiled at GIMMS resolution using (a) theAlaska fire perimeter data set (1950ndash2007 httpagdcusgsgovdatablmfire) (b) the Canadian National Fire Data Base(NFDB 1950ndash2007 acquired from the Fire Research Groupat the Canadian Forest Service) (c) a gridded 500 mMODIS-derived monthly burned area product of Siberia forthe period 2000-9 (MCD45A1 httpmodis-fireumdeduBurned Area Productshtml) as well as two AVHRR-derived1 km fire disturbance data sets a first one used for 1996ndash9(Sukhinin et al 2004) and a second one used for 1992ndash3covering only central Siberia (George et al 2006) Theseregional fire data sets were resampled to express the yearly areaburned in each GIMMS grid cell in both regions and combinedwith the GIMMS Prs time series to create a chronosequence(set of different aged burned areas) describing mean regionalPrs as function of time since fire disturbance in boreal forests(FAO 2001) Grid cells were considered burned when gt75of their area was covered by a burn scar to ensure that the Prstime series of included areas was dominated by the effects ofdisturbance and recovery

3 Productivity changes between 1982 and 2008

Trends in summer NDVI generated from the AVHRR seriesof satellites as well as the MODIS NDVI series agree wellover their period of coincidence across the high latitudes ofNorth America (figure 1) Areas where trends in GIMMSand MODIS summer NDVI contradict each other are few andgenerally have poor statistical support for a linear trend insummer NDVI in either data set The only exception to that

Figure 1 Trends in summer (July and August) NDVI from 2001 to2008 and derived from the (a) GIMMS and (b) MOD13A3 (Hueteet al 2002) data sets measured as regressions slopes resulting fromordinary least squares regression of NDVI versus time and withuncertainty (gray shading) measured using the standard error relativeto the regression slope (c) The comparison of trends in the two datasets with uncertainty (gray shading) measured as the mean of therelative standard errors

is an area in northernmost Canada where vegetation cover isvery sparse and at a latitude where low illumination angleshamper the retrieval of consistent reflectance data from remotesensing (Shuai et al 2008) This area was excluded from furtheranalysis (figure 2)

As the GIMMS record grows in length larger areasacross North America show statistically significant Prs trends(figures 3(a) and (c)) Strikingly statistical support is growingfor the earlier reported contrast between trends in Prs in thetundra and the boreal biome of North America (Goetz et al2005) with lsquogreeningrsquo and lsquobrowningrsquo trends increasinglydominating the tundra and boreal biomes respectively In

Figure 2 Trends in remotely sensed gross productivity (Prs)between 1982 and 2008 Gray shading indicates the trend wasnon-deterministic based on a Vogelsang test (α = 005) Areas inwhite were excluded from the analysis

contrast the proportion of the Siberian tundra and boreallandscape that shows a deterministic trend in Prs has remainedrelatively constant (figures 3(b) and (d)) Nonetheless anincreasingly negative forest Prs response is discernable inSiberia (figure 3(d)) although it is much less widespread thanin North America Indeed in the Eurasian boreal biome Prsincreases are still about twice as common as Prs decreasesLower tree cover in parts of the Siberian boreal zone mightbe partly responsible for this the FAO (2001) vegetationclass lsquoboreal tundra woodlandrsquo had just 15 [SD = 14] percent tree cover in Eurasia versus 29 [SD = 16] per centin North America (tree cover differences between the twocontinents for the polar zone boreal coniferous forest andboreal mountain system were 6 minus4 and 4 respectively) Assuch this zone of sparse tree cover which occurs south of thetundra in Siberia and is here included in the boreal biome isactually a transition zone from tundra to more densely forestedareas However the boreal tundra woodland class representonly 14 of boreal Eurasia and productivity increases aremore often observed in areas of high tree cover in Eurasiathan in North America (figure 4) Extensive increases inPrs are observed in central and eastern Siberia suggestingenvironmentally driven productivity shifts In contrast someof the spatially scattered increases observed in western Siberiamight be due to limitations in the land cover map used to maskagricultural landscapes combined with forest recovery afteragricultural land abandonment around the end of the Soviet era(de Beurs et al 2009)

4 Boreal forests

In general we observe a similar pattern of Prs responseswith regard to tree cover in North America and Siberia

Figure 3 Aerial fraction of the non-anthropogenic vegetated landscape displaying statistically significant deterministic changes in remotelysensed gross productivity (Prs) when considering progressively longer time series since 1982 Tundra and boreal biomes were outlined usingFAO (2001)

Figure 4 MODIS tree cover in areas with and without statisticallysignificant trends in Prs between 1982 and 2008 in North Americanand Siberian tundra and boreal areas as defined by a globaleco-floristic zone map (see text for sources) KruskalndashWallis testsshowed statistically significant differences ( p lt 0001) in tree coverbetween areas with positive and negative Prs trends in both NorthAmerican and Eurasia

Arctic tundra and areas at the tundra-forest ecotone whichare characterized by low tree cover tend to show increasingPrs whereas decreases in Prs are associated with the moredensely forested areas (figure 4) which is consistent withearlier findings (Bunn and Goetz 2006 Beck et al 2011c)The increase in Prs at the forest-tundra ecotone is expected asthe bioclimatic envelope for tree growth moves pole-ward andis in line with observed expansion of larch and tall shrubs attreeline in Siberia and Alaska respectively (Silapaswan et al

2001 Lloyd et al 2003) Areas of increasing and decreasingPrs differ less in tree cover in Eurasia than in North Americadue to the widespread greening response observed in Centraland Western Siberia This area is unique in the sense thatit is underlain by continuous permafrost and yet forestedand it roughly coincides with the range of Dahurian Larch(Larix gmelinii) (Osawa et al 2010) Lloyd et al (2011)found that trends in NDVI in the floodplains of this areareflect the prevalence of positive and negative responses oftree growth to climate warming Moreover they noted thatevergreen tree species of pine and spruce which have a moresoutherly distribution than larch colonize larch dominatedareas in central Siberia (Kharuk et al 2008)

Model projections for the 21st century predict that underscenarios of moderate warming (lt2 C) the limited depthof permafrost thaw and a positive larch-permafrost feedbackcould prevent the substitution of larch forests with evergreenneedle-leaf trees (Tchebakova et al 2009 Shuman et al 2011Zhang et al 2011) However beyond this warming thresholdlarch forest resilience declines and transition to evergreenfunctional types would persist Regardless forest-steppe andsteppe ecosystems are projected to expand northwards intothe current southern extent of the Siberian forest (Vygodskayaet al 2007 Tchebakova et al 2009) If the pattern observedin Siberia is associated with a shift toward a greater coverof evergreen species in taiga forests it would have profoundimplications for the exchange of carbon and energy betweenthe land surface and the atmosphere (Bonan 2008) Furtheranalysis of primary productivity responses along temperatureand forest composition gradients in Eurasia will contributeto assessing and anticipating these projections but might beconstrained by our understanding of permafrost dynamics andchange

Figure 5 Proportion of area in boreal Alaska showing decreases orincreases in remotely sensed gross productivity (Prs) or nodeterministic trend in productivity over the period 1982ndash2008 alonga gradient from evergreen to deciduous tree dominance as mappedby Beck et al (2011a) The deciduous fraction ranges from 0which represents purely evergreen stands to 100 which representspurely deciduous stands

The presence of evergreen and deciduous tree functionaltypes across North Americarsquos boreal forest has not yet beenassessed in relation to the observed patterns of Prs decline(browning) A recent comparison of changes in Prs and whiteas well as black spruce growth trends captured in tree-ringsrevealed positive correlation between the two observationaldata records confirming that Prs trends in boreal Alaskaaccurately reflect decreased spruce productivity caused byclimate warming (Beck et al 2011c) In line with theseobservations we observe expansion of lsquobrowningrsquo in areasdominated by spruce in this region (for areas not recentlydisturbed by fire) (figure 5) In other words decliningspruce productivity appears to be the major driver of thedownward Prs trend in Alaskan boreal forest Nonethelessdecreasing Prs is observed in areas that are largely dominatedby deciduous trees as well albeit much less frequentlyMulti-year eddy covariance measurements show that droughtinduces decreases in primary productivity in both deciduousand evergreenndashdeciduous forests (Welp et al 2007) and thatnet carbon exchange in well-drained deciduous ecosystems isparticularly drought sensitive (Grant et al 2009) Additionalin situ observations of growth in these deciduous boreal treesincluding tree ring analyses are needed to understand theirresponse to climate change of recent decades and how theymight fare in the future The latter question is particularlypertinent if continued climate warming in the coming decadesincreases spruce mortality and associated shifts in the fireregime limit spruce forest regeneration (Johnstone and Chapin2006 Beck et al 2011c) As a consequence the suitabilityof future environmental conditions for currently non-dominantboreal species such as deciduous trees in North America andnorthward migrating temperate species across the boreal biomewill determine the rate and extent of any future biome shift(Lucht et al 2006)

Figure 6 Time series of summer GIMMS-NDVI on the North Slopeof Alaska stratified by shrub cover as mapped by Beck et al (2011b)Sample sizes are as follows lt25 n = 1471 25ndash50 n = 97050ndash75 n = 1326 gt75 n = 1989

5 Arctic tundra and shrubs

Tundra areas well north of tree line in both North America andnorthern Eurasia continue to display increased productivitywhich in North America has become more pervasive in thelast decade (figure 2) This phenomenon has been repeatedlydiscussed in the context of increased shrub growth (Jia et al2003 Walker et al 2003 Bunn et al 2005 Goetz et al 2005Bunn and Goetz 2006) following documented shrub expansionover the last half century through repeat photography (Sturmet al 2001 Tape et al 2006) More recent comparisons ofshrub growth rings and summer NDVI time series in Siberiaindicate significant correlation in some shrub species (Forbeset al 2010) but not in others (Blok et al 2011) Experimentalresearch further shows that deciduous shrubs (ie dwarf birchalder and willow species) respond to multi-year warmingwith increased growth but this response is shared with otherfunctional types particularly graminoid species (Walker et al2006)

A summary of summer NDVI time series along a shrubcover gradient on the North Slope of Alaska indicates thatthe remotely sensed lsquogreeningrsquo trend is not unique to shrub-dominated areas (figure 6) Although shrubs are completelyabsent from relatively few areas on the North Slope of Alaska(Beck et al 2011b) this observation indicates that the Prs trendsin tundra areas are not the result of changes in productivity ofshrub vegetation alone but rather a response shared by multiplefunctional types of vegetation in this biome The latter wasalso suggested by earlier reports of high consistency in thetemporal trends in annual peak NDVI across three temperaturezones (Jia et al 2003) This result implies significant changesin Arctic terrestrial ecosystems can be expected irrespective ofshrub presence a finding also recently noted with respect toalbedo changes (Loranty et al 2011)

Figure 7 (a) Chronology of growing season NDVI as a function oftime since burning (b) The number of GIMMS grid cells available ineach year of the chronosequences Note the much smaller samplesize in the Eurasian data due to the more limited extent of the firedatabase

6 The influence of fire disturbance

In burned forests of boreal North America NDVI recoveredto pre-burn levels within 5ndash10 years (figure 7(a)) whichis comparable to earlier estimates of NDVI-recovery timesfollowing fire (Hicke et al 2003 Goetz et al 2006) NDVIincreases following fire persisted for 15ndash20 years howeverand exceeded pre-burn levels between 20 and 40 years afterthe fire This rebound in productivity is consistent with peakproductivity in intermediate-aged boreal forest successionas well as with eddy covariance estimates of gross primaryproduction along a chronosequence of burned boreal forestsites (Goulden et al 2011) In Alaska this intermediate-aged productivity peak has been related to greater cover ofdeciduous trees (Johnstone et al 2004 Beck et al 2011aAlexander et al 0000)

A much less consistent influence of fire disturbanceon the NDVI chronosequence is visible in northern Eurasia(figure 7(a)) This is partly due to the more limited Eurasianfire databases providing a smaller sample size (figure 7(b))despite the larger extent of the northern Eurasian boreal domain(taiga) Nonetheless there is a much less pronounced post-disturbance Prs drop in Eurasia relative to North America

which may reflect differences in tree mortality the density oftree cover or both Similarly a recovery signal is not evidentin the time series confirming observations in central Siberia byCuevas-Gonzalez et al (2009) showing MODIS NDVI had notreturned to pre-burn levels after 13 years of post-fire vegetationrecovery Unfortunately we cannot rule out the possibilitythat these patterns are artifacts of a high prevalence of falsepositives in the Eurasian fire databases

A shift in dominant tree species during successioncommon in North America generally does not occur inSiberian larch forestsmdashrather biomass recovery dependsprimarily on regrowth of larch trees (Zyryanova et al 2010)However very low recruitment in burn scars has beenreported in Siberian larch forests when a thick moss and dufflayer remains after burning (Sofronov and Volokitina 2009)Furthermore Siberian larch lack fire-resistant seed conesthus recruitment depends on masting events and is highlyvariable between years Together the variation in post-firesoil conditions and seed availability generate large variabilityin forest regeneration in the larch dominated portion of borealSiberia We note however that the variable degree of humanmanagement in many of the burned areas in Eurasia as well asthe potential occurrence of other forms of disturbance than fire(Krankina et al 2004) emphasize the need for further researchinto these post-disturbance recovery dynamics

7 Conclusions

Earlier reported contrasts between productivity trends in Arctictundra and boreal forest biomes have amplified between2002 and 2008 In tundra areas consistent greening trendsare continuing in both North America and Eurasia InNorth America the proportion of tundra areas increasing inproductivity has steadily grown since 1982 reaching 32of non-barren areas in 2008 This greening trend appearsunrelated to shrub density indicating that primary productivityis increasing across a range of functional vegetation typesBoreal forest areas in both North America and Eurasiaincreasingly show declining Prs the so-called lsquobrowningrsquophenomenon that has been tied to increasing drought stress

Areas displaying decreasing productivity tend to havedenser forest cover although some areas of high tree coverin Eurasia do show increasing productivity In Alaskaproductivity decreases are most prevalent in areas dominatedby coniferous trees They are observed in areas with relativelyhigh broadleaf deciduous tree cover as well albeit lessfrequently warranting further research into the health andproductivity of these less widespread forest types In borealNorth America broadleaf deciduous trees and shrubs arealso often prominent in vegetation succession after fire andtheir biomass accumulation most likely contributes to therecovery of NDVI values after fire The absence of a shiftin dominant tree species post-fire and a wide range of larchrecruitment conditions emerge as a potential causes for aless distinct NDVI-recovery pattern across boreal forests ofnorthern Eurasia

Acknowledgments

We would like to thank Greg Fiske and Logan Bernerfor assistance with data pre-processing John Little of theCanadian Forest Service for providing the Canadian firedatabase and Jorge Pinzon for providing the GIMMS 3Gdata We acknowledge support from the National ScienceFoundation (0902056) the NOAA Carbon Cycle Scienceprogram (NA08OAR4310526) and the NASA Carbon program(NNX08AG13G) to SJG We thank Michael Loranty HeatherAlexander and two anonymous referees for comments on themanuscript and for fielding questions

References

Alexander H D Mack M C Goetz S J and Beck P S A Effects ofalternative successional trajectories on carbon pools withinboreal forests of Interior Alaska Ecosphere in review

Angert A Biraud S Bonfils C Henning C C Buermann WPinzon J Tucker C J Fung I and Field C B 2005 Drier summerscancel out the CO2 uptake enhancement induced by warmersprings Proc Natl Acad Sci USA 102 10823ndash7

Bala G Caldeira K Wickett M Phillips T J Lobell D BDelire C and Mirin A 2007 Combined climate and carbon-cycleeffects of large-scale deforestation Proc Natl Acad Sci USA104 6550ndash5

Barber V A Juday G P and Finney B P 2000 Reduced growth ofAlaskan white spruce in the twentieth century fromtemperature-induced drought stress Nature 406 668ndash73

Barrett K Kasischke E McGuire A and Hoy E 2011 Potential shiftsin dominant forest cover in interior Alaska driven by variationsin fire severity Ecol Appl at press (doi10189010-08961)

Beck P S A Goetz S J Mack M C Alexander H Jin YRanderson J T and Loranty M M 2011a The impacts andimplications of an intensifying fire regime on Alaskan borealforest composition and albedo Glob Change Biol 17 2853ndash66

Beck P S A Horning N Goetz S J Loranty M M and Tape K 2011bShrub cover on the North Slope of Alaska a circa 2000 baselinemap Arct Antarct Alpine Res 43 355ndash63

Beck P S A Jonsson P Hoslashgda K A Karlsen S R Eklundh L andSkidmore A K 2007 A ground-validated NDVI dataset formonitoring vegetation dynamics and mapping phenology inFennoscandia and the Kola Peninsula Int J Remote Sens28 4311ndash30

Beck P S A Juday G P Alix C Barber V A Winslow S ESousa E E Heiser P Herriges J D and Goetz S J 2011c Changesin forest productivity across Alaska consistent with biome shiftEcol Lett 14 373ndash9

Berner L Beck P S A Bunn A Lloyd A H and Goetz S J 2011Evaluation of high-latitude coniferous forest growth using asatellite-derived spectral vegetation index J Geophys ResBiogeosci 116 G01015

Blok D Sass-Klaassen U Schaepman-Strub G Heijmans M M P DSauren P and Berendse F 2011 What are the main climatedrivers for shrub growth in Northeastern Siberian tundraBiogeosci 8 1169ndash79

Bonan G B 2008 Forests and climate change forcings feedbacksand the climate benefits of forests Science 320 1444ndash9

Bond-Lamberty B Peckham S D Ahl D E and Gower S T 2007 Fireas the dominant driver of central Canadian boreal forest carbonbalance Nature 450 89ndash92

Bunn A G and Goetz S J 2006 Trends in satellite-observedcircumpolar photosynthetic activity from 1982 to 2003 theinfluence of seasonality cover type and vegetation densityEarth Interact 10 1ndash19

Bunn A G Goetz S J and Fiske G J 2005 Observed and predictedresponses of plant growth to climate across Canada GeophysRes Lett 32 L16710

Chapin F S Fastie C L Viereck L A Ott R A Adams P C Mann DVan Cleve K and Johnstone J F 2006 Successional processes inthe Alaskan boreal forest Alaskarsquos Changing Boreal Forested F S Chapin III M W Oswood K Van Cleve L A Viereck andD L Verbyla (New York Oxford University Press) pp 100ndash20

Chapin F S III et al 2000 Arctic and boreal ecosystems of westernNorth America as components of the climate system GlobChange Biol 6 211ndash23

Chapin F S III et al 2005 Role of land-surface changes in Arcticsummer warming Science 310 657ndash60

Chapin F S III Randerson J T McGuire A D Foley J A andField C B 2008 Changing feedbacks in the climatendashbiospheresystem Front Ecol Environ 6 313ndash20

Cuevas-Gonzalez M Gerard F Balzter H and Riano D 2009Analysing forest recovery after wildfire disturbance in borealSiberia using remotely sensed vegetation indices Glob ChangeBiol 15 561ndash77

de Beurs K M Wright C K and Henebry G M 2009 Dual scale trendanalysis for evaluating climatic and anthropogenic effects on thevegetated land surface in Russia and Kazakhstan Environ ResLett 4 045012

Euskirchen E S McGuire A D Chapin F S III and Rupp T S 2010The changing effects of Alaskarsquos boreal forests on the climatesystem Can J Forest Res 40 1336ndash46

FAO 2001 Global ecological zoning for the global forest resourcesassessment 2000 Forest Resources AssessmentmdashWP 56 Rome

Forbes B C Fauria M M and Zetterberg P 2010 Russian Arcticwarming and lsquogreeningrsquo are closely tracked by tundra shrubwillows Glob Change Biol 16 1542ndash54

Friedl M A et al 2002 Global land cover mapping from MODISalgorithms and early results Remote Sens Environ 83 287ndash302

George C Rowland C Gerard F and Balzter H 2006 Retrospectivemapping of burnt areas in Central Siberia using a modificationof the normalised difference water index Remote Sens Environ104 346ndash59

Gillett N P Weaver A J Zwiers F W and Flannigan M D 2004Detecting the effect of climate change on Canadian forest firesGeophys Res Lett 31 L18211

Goetz S J Bunn A G Fiske G J and Houghton R A 2005Satellite-observed photosynthetic trends across boreal NorthAmerica associated with climate and fire disturbance Proc NatlAcad Sci USA 102 13521ndash5

Goetz S J Fiske G J and Bunn A G 2006 Using satellite time-seriesdata sets to analyze fire disturbance and forest recovery acrossCanada Remote Sens Environ 101 352ndash65

Goulden M L McMillan A M S Winston G C Rocha A VManies K L Harden J W and Bond-Lamberty B P 2011 Patternsof NPP GPP respiration and NEP during boreal forestsuccession Glob Change Biol 17 855ndash71

Grant R F Barr A G Black T A Margolis H A Dunn A LMetsaranta J Wang S McCaughey J H and Bourque C A 2009Interannual variation in net ecosystem productivity of Canadianforests as affected by regional weather patternsmdashaFluxnet-Canada synthesis Agric Forest Meterol 149 2022ndash39

Groisman P and Soja A J 2009 Ongoing climatic change in northernEurasia justification for expedient research Environ Res Lett4 045002

Groisman P Y et al 2007 Potential forest fire danger over northernEurasia changes during the 20th century Glob Planet Change56 371ndash86

Hansen M C DeFries R S Townshend J R G Carroll MDimiceli C and Sohlberg R A 2003 Global percent tree cover ata spatial resolution of 500 meters first results of the MODISvegetation continuous fields algorithm Earth Interact 7 1ndash15

Hicke J A Asner G P Kasischke E S French N H F Randerson J TJames Collatz G Stocks B J Tucker C J Los S O and

Field C B 2003 Postfire response of North American borealforest net primary productivity analyzed with satelliteobservations Glob Change Biol 9 1145ndash57

Higuera P E Brubaker L B Anderson P M Brown T AKennedy A T and Hu F S 2008 Frequent fires in ancient shrubtundra implications of paleorecords for Arctic environmentalchange PLoS One 3 e0001744

Hinzman L et al 2005 Evidence and implications of recent climatechange in northern Alaska and other Arctic regions ClimChange 72 251ndash98

Huete A Didan K Miura T Rodriguez E P Gao X and Ferreira L G2002 Overview of the radiometric and biophysical performanceof the MODIS vegetation indices Remote Sens Environ83 195ndash213

Jia G J Epstein H E and Walker D A 2003 Greening of ArcticAlaska 1981ndash2001 Geophys Res Lett 30 2067

Johnstone J and Chapin F 2006 Effects of soil burn severity onpost-fire tree recruitment in boreal forest Ecosystems 9 14ndash31

Johnstone J F Chapin F S Foote J Kemmett S Price K andViereck L A 2004 Decadal observations of tree regenerationfollowing fire in boreal forests Can J Forest Res 34 267ndash73

Johnstone J F Chapin F S III Hollingsworth T N Mack M CRomanovsky V and Turetsky M R 2010a Fire climate changeand forest resilience in interior Alaska Can J Forest Res40 1302ndash12

Johnstone J F Hollingsworth T N Chapin F S and Mack M C 2010bChanges in fire regime break the legacy lock on successionaltrajectories in Alaskan boreal forest Glob Change Biol16 1281ndash95

Joly K Jande R R Meyers C R and Cole M J 2007 Changes invegetative cover on Western Arctic Herd winter range from1981 to 2005 potential effects of grazing and climate changeRangifer 17 199ndash207

Kasischke E S and Turetsky M R 2006 Recent changes in the fireregime across the North American boreal regionmdashspatial andtemporal patterns of burning across Canada and AlaskaGeophys Res Lett 33 L09703

Kasischke E S Verbyla D L Rupp T S McGuire A D Murphy K AJandt R Barnes J L Hoy E E Duffy P A Calef M andTuretsky M R 2010 Alaskarsquos changing fireregimemdashimplications for the vulnerability of its boreal forestsCan J Forest Res 40 1313ndash24

Kharuk V Ranson K and Dvinskaya M 2008 Evidence of evergreenconifer invasion into larch dominated forests during recentdecades in Central Siberia Eurasian J Forest Res 10 163ndash71

Krankina O N Sun G Shugart H H Kharuk V Bergen K MMasek J G Cohen W B Oetter D R and Duane M V 2004Northern Eurasia remote sensing of boreal forests in selectedregions Land Change Science Observing Monitoring andUnderstanding Trajectories of Change on the Earthrsquos Surface(Remote Sensing and Digital Image Processing vol 6) edG Gutman A C Janetos C O Justice E F MoranJ F Mustard R R Rindfuss D Skole B L Turner II andM A Cochrane (Dordrecht Springer) pp 123ndash38

Lloyd A H Bunn A G and Berner L 2011 A latitudinal gradient intree growth response to climate warming in the Siberian taigaGlob Change Biol 17 1935ndash45

Lloyd A H Rupp T S Fastie C L and Starfield A M 2002 Patternsand dynamics of treeline advance on the Seward PeninsulaAlaska J Geophys Res Atmos 107 8161

Loranty M M Goetz S J and Beck P S A 2011 Tundra vegetationeffects on pan-Arctic albedo Environ Res Lett 6 024014

Lotsch A Friedl M A Anderson B T and Tucker C J 2005 Responseof terrestrial ecosystems to recent northern hemispheric droughtGeophys Res Lett 32 L06705

Lucht W Schaphoff S Erbrecht T Heyder U and Cramer W 2006Terrestrial vegetation redistribution and carbon balance underclimate change Carbon Balance Manage 1 6

Lynch J A Clark J S Bigelow N H Edwards M E and Finney B P2003 Geographical and temporal variations in fire history inboreal ecosystems in Alaska J Geophys Res Atmos 108 8152

Macdonald S E and Fenniak T E 2007 Understory plant communitiesof boreal mixedwood forests in western Canada naturalpatterns and response to variable-retention harvesting ForestEcol Manage 242 34ndash48

Mack M Treseder K Manies K Harden J Schuur E Vogel JRanderson J and Chapin F 2008 Recovery of aboveground plantbiomass and productivity after fire in mesic and dry blackspruce forests of interior Alaska Ecosystems 11 209ndash25

Mack M C Schuur E A G Bret-Harte M S Shaver G R andChapin F S III 2004 Ecosystem carbon storage in Arctic tundrareduced by long-term nutrient fertilization Nature 431 440ndash3

McGuire A D Anderson L G Christensen T R Dallimore S Guo LHayes D J Heimann M Lorenson T D Macdonald R W andRoulet N 2009 Sensitivity of the carbon cycle in the Arctic toclimate change Ecol Monogr 79 523ndash55

Myneni R B Keeling C D Tucker C J Asrar G and Nemani R R1997 Increased plant growth in the northern high latitudes from1981ndash1991 Nature 386 698ndash702

Osawa A Zyryanova O A Matsuura Y Kajimoto T andWein R W (ed) 2010 Permafrost Ecosystems Siberian LarchForests (Dordrecht Springer)

Piao S Wang X Ciais P Zhu B Wang T A O and Liu J I E 2011Changes in satellite-derived vegetation growth trend intemperate and boreal Eurasia from 1982 to 2006 Glob ChangeBiol 17 3228ndash39

Randerson J T et al 2006 The impact of boreal forest fire on climatewarming Science 314 1130ndash2

Shaver G R Billings W D Chapin F S III Giblin A ENadelhoffer K J Oechel W C and Rastetter E B 1992 Globalchange and the carbon balance of Arctic ecosystems Bioscience42 433ndash41

Shaver G R and Chapin F S III 1991 Production biomassrelationships and element cycling in contrasting Arcticvegetation types Ecol Monogr 61 1ndash31

Shuai Y Schaaf C B Strahler A H Liu J and Jiao Z 2008 Qualityassessment of BRDFalbedo retrievals in MODIS operationalsystem Geophys Res Lett 35 L05407

Shuman J K Shugart H H and OrsquoHalloran T L 2011 Sensitivity ofSiberian larch forests to climate change Glob Change Biol17 2370ndash84

Silapaswan C S Verbyla D L and McGuire A 2001 Land coverchange on the Seward Peninsula the use of remote sensing toevaluate the potential influences of climate warming onhistorical vegetation dynamics Can J Remote Sens 27 542ndash54

Sofronov M A and Volokitina A V 2009 Wildfire ecology incontinuous permafrost zone Permafrost Ecosystems SiberianLarch Forests ed A Osawa O A Zyryanova Y MatsuuraT Kajimoto and R W Wein (Dordrecht Springer) pp 59ndash82

Soja A J Tchebakova N M French N H F Flannigan M DShugart H H Stocks B J Sukhinin A I Parfenova E IChapin F S III and Stackhouse J P W 2007 Climate-inducedboreal forest change predictions versus current observationsGlob Planet Change 56 274ndash96

Sturm M Douglas T Racine C and Liston G E 2005a Changingsnow and shrub conditions affect albedo with globalimplications J Geophys Res Biogeosci 110 13

Sturm M Holmgren J McFadden J P Liston G Chapin F S III andRacine C H 2001 Snowndashshrub interactions in Arctic tundra ahypothesis with climatic implications J Clim 14 336ndash44

Sturm M Schimel J Michaelson G Welker J M Oberbauer S FListon G E Fahnestock J and Romanovsky V E 2005b Winterbiological processes could help convert Arctic tundra toshrubland Bioscience 55 17ndash26

Sukhinin A I et al 2004 AVHRR-based mapping of fires in Russianew products for fire management and carbon cycle studiesRemote Sens Environ 93 546ndash64

Tape K Sturm M and Racine C 2006 The evidence for shrubexpansion in northern Alaska and the pan-Arctic Glob ChangeBiol 12 686ndash702

Tape K D Lord R Marshall H-P and Ruess R W 2010Snow-mediated ptarmigan browsing and shrub expansion inArctic Alaska Ecoscience 17 186ndash93

Tchebakova N M et al 2009 The effects of climate permafrost andfire on vegetation change in Siberia in a changing climateEnviron Res Lett 4 045013

Tucker C J Pinzon J Brown M Slayback D Pak E Mahoney RVermote E and El Saleous N 2005 Extended AVHRR 8-kmNDVI data set compatible with MODIS and SPOT vegetationNDVI data Int J Remote Sens 26 4485ndash98

Turetsky M R Kane E S Harden J W Ottmar R D Manies K LHoy E and Kasischke E S 2011 Recent acceleration of biomassburning and carbon losses in Alaskan forests and peatlandsNature Geosci 4 27ndash31

Verbyla D L 2008 The greening and browning of Alaska baded on1982ndash2003 satellite data Global Ecol Biogeogr 17 547ndash55

Vogelsang T J 1998 Trend function hypothesis testing in the presenceof serial correlation Econometrica 66 123ndash48

Vygodskaya N N et al 2007 Ecosystems and climate interactions inthe boreal zone of northern Eurasia Environ Res Lett2 045033

Walker D A et al 2003 Phytomass LAI and NDVI in northernAlaska relationships to summer warmth soil pH plantfunctional types and extrapolation to the circumpolar ArcticJ Geophys Res 108 8169

Walker M D et al 2006 Plant community responses to experimentalwarming across the tundra biome Proc Natl Acad Sci103 1342ndash6

Wannebo-Nilsen K Bjerke J W Beck P S A and Toslashmmervik H 2010Epiphytic macrolichens in spruce plantation and native birchforests along a coast-inland gradient in Norway Boreal Env Res15 43ndash57

Welp L R Randerson J T and Liu H P 2007 The sensitivity of carbonfluxes to spring warming and summer drought depends on plantfunctional type in boreal forest ecosystems Agric ForestMeterol 147 172ndash85

Werner R A Raffa K F and Illman B L 2006 Dynamics ofphytophagous insects and their pathogens in Alaskan borealforests Alaskarsquos Changing Boreal Forest ed F S Chapin IIIM W Oswood K Van Cleve L A Viereck andD L Verbyla (New York Oxford University Press) pp 133ndash46

Wookey P A et al 2009 Ecosystem feedbacks and cascade processesunderstanding their role in the responses of Arctic and alpineecosystems to environmental change Glob Change Biol15 1153ndash72

Zhang N Yasunari T and Ohta T 2011 Dynamics of the larchtaiga-permafrost coupled system in Siberia under climatechange Environ Res Lett 6 024003

Zhang X Friedl M A and Schaaf C B 2006 Global vegetationphenology from moderate resolution imaging spectroradiometer(MODIS) evaluation of global patterns and comparison with insitu measurements J Geophys Res 111 G04017

Zyryanova O A Abaimov A P Bugaenko T N and Bugaenko N N2010 Recovery of forest vegetation after fire disturbancePermafrost Ecosystems Siberian Larch forests ed A Osawa OA Zyryanova Y Matsuura T Kajimoto andR W Wein (Dordrecht Springer) pp 83ndash96

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf

Corrigendum Satellite observations of high northern latitude vegetation productivity changes between 1982 and 2008 ecological variability and regional differences

References

Corrigendum Satellite observations ofhigh northern latitude vegetationproductivity changes between 1982 and2008 ecological variability and regionaldifferences2011 Environ Res Lett 6 045501

Pieter S A Beck and Scott J Goetz

E-mail pbeckwhrcorg

Received 26 March 2012Accepted for publication 26 March 2012Published 18 April 2012Online at stacksioporgERL7029501

Due to an error in our implementation of the Vogelsangtest (Vogelsang 1998) the criterion used to determinewhere trends in remotely sensed gross productivity (Prs)were deemed deterministic at α = 005 was overly le-nient This resulted in an overestimation of the totalgeographical area exhibiting deterministic Prs trends at

Figure 2 Trends in remotely sensed gross productivity (Prs) between 1982 and 2008 Gray shading indicates the trend wasnon-deterministic based on a Vogelsang test (α = 005 in the top panel and α = 01 in the bottom panel) Areas in white were excludedfrom the analysis as described in the text

α = 005 This error affected figures 2 3 4 and 5 butdoes not modify any of the conclusions drawn from theanalysis

Corrected versions of the figures are included belowIn addition but unrelated titles of the panels in figure 3

were inadvertently omitted They are included here

11748-932612029501+03$3300 ccopy 2012 IOP Publishing Ltd Printed in the UK

References

E-mail pbeckwhrcorg

1 Introduction

Prspy = max

(GSLminus1

t+GSLpsumi=t+1

NDVIpyi

4 Boreal forests

7 Conclusions

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

References

E-mail pbeckwhrcorg

1 Introduction

Prspy = max

(GSLminus1

t+GSLpsumi=t+1

NDVIpyi

4 Boreal forests

7 Conclusions

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

References

E-mail pbeckwhrcorg

1 Introduction

Prspy = max

(GSLminus1

t+GSLpsumi=t+1

NDVIpyi

4 Boreal forests

7 Conclusions

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

E-mail pbeckwhrcorg

1 Introduction

Prspy = max

(GSLminus1

t+GSLpsumi=t+1

NDVIpyi

4 Boreal forests

7 Conclusions

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

E-mail pbeckwhrcorg

1 Introduction

Prspy = max

(GSLminus1

t+GSLpsumi=t+1

NDVIpyi

4 Boreal forests

7 Conclusions

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

Prspy = max

(GSLminus1

t+GSLpsumi=t+1

NDVIpyi

4 Boreal forests

7 Conclusions

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

Prspy = max

(GSLminus1

t+GSLpsumi=t+1

NDVIpyi

4 Boreal forests

7 Conclusions

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

4 Boreal forests

7 Conclusions

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

7 Conclusions

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

7 Conclusions

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

7 Conclusions

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

Acknowledgments

References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

1 Introduction



4 Boreal forests



7 Conclusions

Acknowledgments

References

erl12_2_029501pdf


References

Documents

Satellite observations of high northern latitude vegetation