http://www.diva-portal.org

This is the published version of a paper published in Environmental Research Letters.

Citation for the original published paper (version of record):

Yi, C., Pendall, E., Ciais, P. (2015)

Focus on extreme events and the carbon cycle.

Environmental Research Letters, 10(7): 070201

http://dx.doi.org/10.1088/1748-9326/10/7/070201

Access to the published version may require subscription.

N.B. When citing this work, cite the original published paper.

Permanent link to this version:http://urn.kb.se/resolve?urn=urn:nbn:se:su:diva-119194

This content has been downloaded from IOPscience. Please scroll down to see the full text.

Download details:

IP Address: 130.237.189.128

This content was downloaded on 03/07/2015 at 13:48

Please note that terms and conditions apply.

Focus on extreme events and the carbon cycle

View the table of contents for this issue, or go to the journal homepage for more

2015 Environ. Res. Lett. 10 070201

(http://iopscience.iop.org/1748-9326/10/7/070201)

Home Search Collections Journals About Contact us My IOPscience

iopscience.iop.org/page/termshttp://iopscience.iop.org/1748-9326/10/7http://iopscience.iop.org/1748-9326http://iopscience.iop.org/http://iopscience.iop.org/searchhttp://iopscience.iop.org/collectionshttp://iopscience.iop.org/journalshttp://iopscience.iop.org/page/aboutioppublishinghttp://iopscience.iop.org/contacthttp://iopscience.iop.org/myiopscience

Environ. Res. Lett. 10 (2015) 070201 doi:10.1088/1748-9326/10/7/070201

EDITORIAL

Focus on extreme events and the carbon cycle

ChuixiangYi1,2, Elise Pendall3 andPhilippeCiais4

1 School of Earth and Environmental Sciences, QueensCollege, CityUniversity ofNewYork, NY 11367,USA2 Department ofMeteorology, Bert BolinCentre for Climate Research, StockholmUniversity, Stockholm S-106 91, Sweden3 Hawkesbury Institute for the Environment, University ofWestern Sydney, Locked Bag 1797, Penrith,NSW2751, Australia4 Laboratoire des Sciences duClimat et de l’Environnement, CEACNRSUVSQ, 91191Gif sur Yvette, France

E-mail: cyi@qc.cuny.edu

Keywords: extreme events, ecological disturbance, droughts, hurricane, bark beetles, carbon cycle, climatemitigation

AbstractClimate physics indicates that warming climate is a likely cause of extremeweather andmore frequentand intense climate events. These extreme events can disrupt terrestrial carbon dynamics dramaticallyby triggering ecological disturbances and potentially forcing climate–carbon feedbacks. In this paperwe synthesize the findings of 26 papers that focus on collecting evidence and developing knowledge ofhow extreme events disturb terrestrial carbon dynamics.

1. Introduction

The World Meteorological Organization reportedrecently ‘that the world experienced unprecedentedhigh-impact climate extremes during the 2001–2010decade,whichwas thewarmest since the start ofmodernmeasurements in 1850 and continued an extendedperiod of pronounced global warming’ (WorldMeteor-ological Organization 2013). Even though since 2000,the warming rate is less than during the previous decade(the Hiatus period) extreme hot days have continued toincrease in frequency and severity (Seneviratneet al 2014). Although globalmean annual temperature isnot increasing every year, global decadalmean tempera-ture has increased every decade since 1970s. On theother hand, it is clearly evident that the frequency andintensity of extreme events of weather and climate havebeen increasing in this warming world (World Meteor-ological Organization 2011, Coumou and Rahm-storf 2012). Does global warming generate moreextreme weather events? There has been no directanswer yet. However, the link between extreme weatherand warming climate can be understood by the physicsof climate. In principle, the ability of atmosphere to holdwater is dictated by theClausius-Clapeyron equation;

e TT

T( ) 6.1094 exp

17.625

243.04, (1)s ⎜ ⎟

⎛⎝

⎞⎠= +

where, es is saturation vapor pressure in hPa, repre-senting the maximum amount of water vapor that air

can hold at a given temperature T (°C) (Ahrens 2000).The Clausius–Clapeyron equation (1) is derived fromthe first and second law of thermodynamics and hasbeen broadly verified under both laboratory andnatural conditions. It predicts that the water-holdingcapacity of the atmosphere increases by about 7% forevery 1 °C rise in temperature, which is close tosatellite-based observations that recorded a rate ofchange in the atmospheric water vapor content6% °C−1 (Kininmonth 2010).

Figure 1 illustrates conceptual links between globalwarming and extreme weather. Warming air hashigher saturation vapor pressure, i. e. larger size of the‘room’ for water to evaporate from soils and waterbodies and transpire from plants. In addition toincreased saturation vapor pressure in the atmo-sphere, warming directly increases plant evapo-transpiration demand, leading to more water emittedby vegetated surfaces to the atmosphere unless soilmoisture gets too depleted. This is likely to causeextreme precipitation events as the probability ofwater molecules condensing increases. Meanwhile,weather systems accompanied with extreme precipita-tion events can become more violent as more latentheat is released. It will also take longer to recharge theatmosphere withmoisture as its capacity to hold waterincreases. Thus, duration between rain eventsbecomes longer and hence drought would increase.We cannot be certain that any particular extremeweather event is caused by global warming. However,

OPEN ACCESS

RECEIVED

29April 2015

ACCEPTED FOR PUBLICATION

14May 2015

PUBLISHED

2 July 2015

Content from this workmay be used under theterms of theCreativeCommonsAttribution 3.0licence.

Any further distribution ofthis workmustmaintainattribution to theauthor(s) and the title ofthework, journal citationandDOI.

© 2015 IOPPublishing Ltd

http://dx.doi.org/10.1088/1748-9326/10/7/070201mailto:cyi@qc.cuny.eduhttp://crossmark.crossref.org/dialog/?doi=10.1088/1748-9326/10/7/070201&domain=pdf&date_stamp=2015-07-02http://crossmark.crossref.org/dialog/?doi=10.1088/1748-9326/10/7/070201&domain=pdf&date_stamp=2015-07-02http://creativecommons.org/licenses/by/3.0http://creativecommons.org/licenses/by/3.0http://creativecommons.org/licenses/by/3.0

we are certain that global warming amplifies the riskfactors for extreme weather events from the principleof thermodynamics—hot air can hold more watermolecules. Global warming increases atmosphericevaporative demand, which is supported by observa-tional data; about half of the land in warm regions(annual temperature >16 °C) has been drying withincreasing temperature since the 1970s (Junget al 2010, Yi et al 2014).

The high risk consequence of these warming-induced extreme events is that they often lead to ecolo-gical disturbances which can affect land–atmosphereexchanges of carbon and water dramatically, poten-tially contributing to climate–carbon feedback (Coxet al 2013), illustrated in figure 2.Most climate-relateddisturbance events have negative impacts on the healthand carbon balance of ecosystems, especially in forests,because they kill trees, inducing a large carbon sourceto the atmosphere followed by a slow recovery sink(figure 3). Some ecosystems are more resistant to dis-turbance; for instance, deserts and grasslands are

adapted to water stress. However, mega-drought andheat-waves can cause damage to plants that results inmortality, andmore frequent drought can in the long-term change ecosystem structure (Breshears et al 2005,Edburg et al 2012;figure 3).

In addition to altering the carbon balance, extremeevents cause feedbacks to climate due to changedalbedo and roughness, and because of altered sensibleand latent heat emissions. Stand replacing fires innorthern regions causes an increase in winter albedodue to higher snow cover over the short vegetationthat covers the ground when the forest has burned.This has a net cooling climate effect (Randersonet al 2006). Extreme drought generally suppresses eva-potranspiration and increases sensible heat (Teulinget al 2013), which acts to increase air temperature anddecrease rain probability. This causes a positive feed-back on heat and drought. Thus, the direction of feed-back between ecological disturbance and globalwarming is challenging to predict, warranting the

Figure 1.Conceptual links betweenwarming climate and extremeweather.

Figure 2. Schematicmap of a potential climate-carbon feedback loop. This focus issue collects evidence on the links between extremeweather/climate and the carbon cycle.

2

Environ. Res. Lett. 10 (2015) 070201 CYi et al

investigations across a range of disturbance types andecosystems presented in this focus issue.

The aim of this focus issue is not to verify the exis-tence of feedbacks but to collect evidence on how theseextreme events, such as heat waves, droughts, hurri-canes and insect outbreaks change land–atmosphereexchanges, and to use quantitative approaches to iden-tify these events and explore driving forces behindtheir ecological consequences.

2.Drought

Fourteen papers in this focus provide lines of evidenceof how ongoing extreme droughts affect terrestrialecosystem exchange of CO2 with the atmosphere fromtower-based flux measurements and satellite-basedmodeling. Net ecosystem–atmosphere exchange(NEE) of CO2 is a small imbalance between two largeCO2 fluxes with opposite directions: uptake from theatmosphere by gross primary production (GPP) andrelease back to the atmosphere by terrestrial ecosystemrespiration (TER). These papers demonstrate thatimpact of extreme droughts on these two componentsis different and also varies from season to season andfrombiome to biome.

2.1. Spring droughtSoil moisture availability in spring is usually higherthan in summer because soil reservoirs have beenrecharged by winter precipitation, in many summer-dry northern regions. However, lower precipitationcombined with warmer winter temperatures can leadto spring droughts. Based on eddy-flux data collectedin Switzerland,Wolf et al (2013) found that grasslandsand forests respond differently to spring droughtsbecause these ecosystems exhibit different water use

strategies. The ratio of CO2 uptake (GPP) to water loss(transpiration) is called water use efficiency (WUE),which can be optimized at the ecosystem level underwater limited conditions as plants can regulate the sizeof stomatal openings (Claesson and Nycander 2013).The eddy-flux data indicate that forests increased theirWUE during the spring drought, while grasslands’WUE did not change substantially during the springdrought (Wolf 2013).

The response of ecosystems to drought is differentfrom biome to biome. For peatlands, drought condi-tions will not only reduce soil wetness (reducing GPP)but also lower the water table, increasing the depth ofthe aerobic respiration layer (increasing TER). Thedata collected in a south Swedish nutrient-poor peat-land by eddy-flux measurements indicate thatdroughts can turn a temperate peatland from a carbonsink to a carbon source (Lund et al 2012). The effect ofdrought on the peatlands depends on drought dura-tion. A prolonged drought in the early growing seasonresulted in reduced rates of GPP, while a short butsevere drought in the middle of the growing seasonresulted in increased rates of TER (Lund et al 2012),demonstrating the contrasting effects of extreme cli-mate events on components of the carbon cycle.

The 2010 spring drought in southwestern Chinawas the most severe and long-lasting of the last halfcentury. Satellite-based observations show that severeand extended spring droughts substantially reducedthe enhanced vegetation index (EVI) and GPP andboth did not recover from drought stress until August(Zhang et al 2012). This spring drought event also low-ered the water levels of rivers, reservoirs and lakes, andeven dried up some water bodies, which affected agri-cultural irrigation and hence crop production. Othersatellite-based observations also show that springdrought induced by the El Niño/La Niña-Southern

Figure 3.Hypothetical framework showing how ecological disturbance (red arrow) can perturb the ecosystem carbon cycle. First,vegetation is killed and its potential for C uptake via gross primary production (GPP, green line) declines. Disturbancemay enhanceClosses from ecosystem respiration (TER, dark blue solid line) if increased inputs of dead biomass stimulate decomposition.Alternatively, TERmay be reduced as autotrophic respiration is eliminated (dark blue dashed line). Net ecosystemproduction (NEP,light blue line), or net C storage on land, declines rapidly but recoversmore slowly, even up to decades. Eventually, gains and losses arebalanced and the ecosystem returns to steady state (Odum1969). Based on afigure in Edburg et al (2012), originally published inFrontiers in Ecology and the Environment.

3

Environ. Res. Lett. 10 (2015) 070201 CYi et al

Oscillation (ENSO) delayed vegetation onset time andreduced annual net primary production (NPP) in thesubtropical island of Taiwan (Chang et al 2013).

2.2. Late-summer droughtKolb et al (2013) found that impacts of extreme late-summer drought on net ecosystem production (NEP)of semi-arid forest in the southern west United State(SWUS) differed between disturbed and undisturbedecosystems based on six-year eddy-covariance mea-surements. The late-summer drought shifted AugustNEP at the undisturbed site from a carbon sink tosource because the reduction of GPP (70%) exceededthe reduction of TER (35%). At the burned site, late-summer drought shifted August NEP from a weakcarbon source to neutral because the reduction of TER(40%) exceeded the reduction of GPP (20%). Theresults illustrate the sensitivity of semi-arid forestcarbon sequestration to extreme drought and suggestthat warming-associated drought may interact withdisturbances such as fire, possibly weakening semi-arid forest carbon sinks.

The contrasting impact of drought on GPP andTER between grazed and ungrazed desert steppe wasalso evidenced by data collected with tower-basededdy covariance approach on the Mongolian Plateau(Shao et al 2013). This two-year dataset indicates thatgrazing can play a positive role in biophysical regula-tion of carbon fluxes in a desert steppe. The positiveeffects of grazing on the uptake of CO2 in semiaridsteppe were also reported by Kang et al (2013).Although both these examples were short, the analyseswill be useful for grazing management. Nevertheless,additional long-term fluxmeasurements are needed tounderstand the mechanisms underlying the interac-tions between extreme climate, ecological disturbance,and carbon cycling.

2.3. Long-termdrought eventsBabst et al (2012) did clustering analysis of tree-ringgrowth data at ∼1000 sites across Europe that showedregionally consistent growth patterns driven by cli-mate regimes and species–specific growth character-istics. Fifteen regional, synthetic chronologies thatrepresent radial tree growth anomalies across Europewere reconstructed over the past ∼500 years. Theanalysis revealed 18 extreme tree-growth events dur-ing the pre-instrumental and 2 events (1947/48 and1976) in the instrumental period, which may be beencaused by extreme late-summer drought conditions.

The complexity of interactions between droughtand vegetation response is demonstrated in a 26-yearstudy that analyzed vegetation growth based on Nor-malizedDifference Vegetation Index (NDVI) in China(Xu et al 2012). Extreme events in temperature, pre-cipitation and drought based on Palmer DroughtSeverity Index (PDSI) appear to have increased acrossChina over the past three decades. Associated with the

increasing trend in heat waves and drought events, adecreasing trend in the extreme lowNDVI events (lowvegetation growth) was observed. However, differentregions in China experienced contrasting temporalpatterns. Although extreme drought events increasedin the 2000s in both northern and southern China,extreme lowNDVI events keep decreasing in southernChina, while the decrease is stalled or even reversed innorthern China. Apparently, drought does not impactvegetation productivity in a consistent fashion acrosscontinents; prior land use such as grazing may need tobe considered to reveal detailedmechanisms.

3. Extremeweather

3.1.HurricaneThe impact of catastrophic hurricanes on life loss andproperty damage has received considerable attention,but environmental impacts are often neglected. Hurri-canes create major ecological disturbances by killinghundreds ofmillions of trees and adding plant litter onthe forest floor, and potentially increasing risk of forestfire conditions. The drastic ecological disturbance byhurricanes is expected to have significant impacts onregional forest carbon balances. In this focus issue, wereport case-studies for hurricane impacts on carboncycling in tropical savannas and temperate forestecosystems (Hutley et al 2013, Vargas 2012, Fisket al 2013).

3.1.1. Hurricane-savanna-fire interactionsHutley et al (2013) estimated the ecological conse-quences of the mega-cyclone Monica that damaged ordestroyed 140 million trees across the coastal and sub-coastal region of north Australia in 2006. They usedthe moderate resolution imaging spectroradiometer(MODIS) GPP (MOD17A2) to track spatial andtemporal patterns between six years preceding and sixyears following the cyclone Monica. They found thatthere was a significant decline in the region’s GPP(directly proportional to its capacity to sequestercarbon) relative to the pre-cyclone rates, and thisreduced productivity persisted for four years post-cyclone. The most important ecological consequencesof hurricanes is that the hundreds of millions of treeskilled or damaged will be converted into CO2 andreturned back to the atmosphere sooner by fire or laterby decomposition. Hutley et al (2013) examined theimpact that the massive increase in fuel load (vegeta-tion debris) following Cyclone Monica would have onthe region’s fire regime. They found that fire frequencywas extreme in the year following the cyclone, but firefrequency more or less returned to pre-cyclone condi-tions in the years that followed—an unexpected result.

3.1.2. Effect of hurricanes on the soil carbon fluxVargas (2012) observed large pulses of soil CO2 fluxafter hurricane Wilma in two-year time series of soil

4

Environ. Res. Lett. 10 (2015) 070201 CYi et al

carbon dioxide efflux measurements from two soil-sensor nodes. He reported that in the year followinghurricane Wilma, annual soil carbon dioxide emis-sions were more than 3.9 kgCm−2 but decreased to1.7 kgCm−2 for the second post-hurricane year.Vargas attributed the increase in the soil carbon fluxpost-Wilma to tree damage depositing nitrogen-rich,easily decomposable organic matter and higher basalsoil CO2 efflux rates.Moreover, GPPwas not impactedimmediately after the hurricane but during thedrought period of the first year, a depletion of non-structural carbon reserves that likely were used for a‘rapid’ recovery of canopy cover (e.g., leaf area index)during thefirst 2months following the hurricane.

3.1.3. Carbon footprint of historical hurricanes ineasternUSThe case studies conducted by Hutley et al (2013) andVargas (2012) have demonstrated that a single hurri-cane event can exert a large disturbance in forestcarbon balance, but forests have strong resilience torecover. In the short- to medium-term, hurricanesrelease carbon due to leaves and litters can be easilydecomposed, while over the longer term, GPP will beincreased from forest re-growth. Therefore, the recov-ery times of the two components, GPP and TER, offorest carbon balance from hurricane disturbances toapproaching the normal rates are not the same (seefigure 3). This imbalance betweenGPP andTERmightaffect the ability of forests to absorb carbon in a long-term perspective. Fisk et al (2013) investigated it bymodeling the reconstructed maps of tree mortalityfrom hurricanes impacting the eastern US from 1851to 2000. They found how long it takes for the forest re-growth to recover the carbon loss depends on the typeof forest. Small disturbances in fast-growing regionscan recover quickly, but large disturbances or slowergrowth rates can result in recovery that takes morethan a century. Fisk et al (2013) concluded that onaverage, tropical cyclones contribute a net carbonsource over latter half of the 19th century due to highstorm activity and the existence of larger forests duringthat period. However, throughout much of the 20thcentury a regional carbon sink is estimated resultingfromperiods of forest recovery exceeding damage.

3.2. Ice stormsTerrestrial ecosystems in temperate to subtropicalregions are vulnerable to the invasion of extreme coldevents like ice storms. These cold events are commonin East Asia and can cause massive structural damageto forests and often define the northern boundaries ofkey tree species. Sun et al (2012) investigated theimpact of themassive 2008Chinese ice stormon forestecosystems in southern China and found that whileforest structures were substantially changed by the icestorm, forest greenness, as measured by satellite,recovered fairly quickly overall; however, moderately

damaged forests recovered the slowest as a conse-quence of salvage logging. These cold events can alsointerrupt ecosystem potential productivity dramati-cally. Huang et al (2013) observed the large reductionof GPP from Qianyanzhou Ecological ExperimentalStation in southern China as a result of ice storms in2005 and 2008.

3.3. Extreme snow events3.3.1. Extreme short snow seasonClimate change is likely to shorten the duration ofsnow cover in mountainous regions such as theEuropean Alps. In 2011, Europe experienced extre-mely warm spells in late winter. These events did notspare the Alps, and the Italian Aosta Valley region sawthe earliest snowmelt for 65 years and the shortestsnow season for 83 years. Galvagno et al (2013)examined how the short snow season modified thecarbon balance of subalpine grassland in the region byeddy-covariance measurements. They observed thatthe extremely short snow season in the Alps led to a100% increase in the annual net CO2 uptake. Theyfound, interestingly, this larger carbon sink wasattributable to smaller carbon dioxide emissions byTER during the shorter winter, rather than toenhanced plant activity and GPP during the longersummer.

3.3.2. Summer drought worsens impacts of bitter winterinMongolia Plateau‘Dzud’ in theMongolian Plateau refers to awinter withextreme cold, heavy snowfall, reduced availability offorage and widespread mortality of livestock (Johnet al 2013). Recently, the dzuds and droughts in theMongolian Plateau have become more serious. Parti-cularly when a dzud is preceded by a summer drought,the combined summer drought–dzud often leads toeven higher mortality of livestock (Fernández-Gimé-nez et al 2012). John et al (2013) identified suchextreme events in the Mongolian Plateau during theperiod of 2000–2010 using precipitation and PDSIdata, and examined the vegetation response to theseextreme events using anomalies of vegetation indices.They concluded that the grassland biome on theplateau was more resilient to the extreme events thanthe desert biome.

4. Bark beetles

Global warming and drought are associated withexpanding populations of forest insects and pathogens(Logan et al 2003), which cause widespread treemortality and cascading ecosystem level disturbanceimpacts (Edburg et al 2012). Severe outbreaks of barkbeetles may coincide with or follow extreme droughts(Breshears et al 2005), as stressed trees are unable todefend themselves from the attacking insects. Disrup-tion to the carbon cycle by beetle-induced mortality

5

Environ. Res. Lett. 10 (2015) 070201 CYi et al

has been compared with the carbon losses by forestfires in Canada (Kurz et al 2008). In this focus issue, weprovide four reports on climate-associated insect out-breaks and their impacts on carbon cycling in NorthAmerica.

Over the last two decades, a widespread outbreakof bark beetles (Dendroctonus spp.) has affected tens ofmillions of hectares of conifer forests (Edburget al 2012). In an Idaho pine forest, aboveground car-bon stocks in trees were reduced by up to 50%, asshown by multispectral imaging and LIDAR (Brightet al 2012). Most of the area in this region was moder-ately affected by the beetles, although small areasexperienced severe mortality. An inventory-basedanalysis over the western United States found that theamount of carbon in trees killed by beetles exceededthat lost to forest fires (Hicke et al 2013). These dis-turbances represent a loss of about 10% of the totalcarbon stock inWestern US forests, comparable to theamount removed by logging activities. Although theimpact of bark beetle-induced mortality is visuallydramatic, their influence on carbon and water cyclingmay be less than anticipated. Using eddy covariance,Reed et al (2014) found that maximum carbon uptakeby a Wyoming pine forest was not strongly altereddespite the progression of mortality from 30% tonearly 80% in the tower footprint. They attributed theresilience of the ecosystem to enhanced light use effi-ciency by the surviving trees as the canopy opened.

The effects of defoliation on carbon cycling wereevaluated using an ecosystem demography model thatrepresents fine-scale canopy structure to scale up phy-siological processes (Medvigy et al 2012). The modelindicated that net ecosystem production decreasedlinearly with increasing defoliation by gypsy moths,but that the carbon cycle responses interacted withdrought impacts. Drought-induced tree mortality waslower in scenarios including defoliation disturbance(Medvigy et al 2012), possibly due to compensatingprocesses such as enhanced water or nutrient avail-ability at intermediate disturbance levels (Reedet al 2014).

5.Quantitative approaches

How to quantify relationships between the extremeevents and their ecological consequences is a knowl-edge gap. In this focus issue, a few quantitativeapproaches have been demonstrated.

5.1. Tail ApproachThe ecological consequences of extreme climate eventscan be mapped from remote sensing images withsharper changes in vegetation greenness and leaf areaindex, or can be identified by a large reduction incarbon uptake (GPP) with eddy-covariance measure-ments. Extreme events occur with small probabilitybut with large amplitudes of changes in variables

characterizing the state of the terrestrial biosphere(Zscheischler et al 2014). These extreme events can bedefined as the occurrence of certain values in the tailsof the probability distribution of the anomalies of thevariables, similar to extreme weather defined by IPCC(Seneviratne et al 2012). We call this method the‘tailapproach’. The question is how long the tails can be cutto define extreme events.

Zscheischler et al (2014) applied this approach tothe terrestrial carbon balance by defining extremes tobe outside a certain threshold q, which is defined by apercentile (i.e. 1%,…,10%) on the absolute values ofthe anomalies. They then defined an extreme event byspatiotemporally contiguous points whose values arelarger than q (positive extremes) and smaller than −q(negative extremes), respectively. Zscheischler et al(2014) reanalyzed four global MODIS-based GPPdatasets over the last 30 years and found that positive(extreme excess in carbon uptake) and negative(extreme reduction in photosynthetic carbon uptake)GPP extremes occurring on 7% of the spatiotemporaldomain explained 78%of the global interannual varia-tion in GPP. Most of the negative GPP extremes (lar-ger than positive extremes) were attributed todroughts.

The tail approach depends on the shape of theprobability distribution of the anomalies, and on theirseverity. The identification of extreme events requiresthat the shapes of probability distributions are similarso that comparison and analysis cross all sites can bemade. However, it is impossible to have similar dis-tribution of probability of the anomalies on a globalscale due to heterogeneous properties of vegetationfrom region to region. Liu et al (2013) developed a newmethod of using the Box–Cox transformation (Boxand Cox 1964) to convert non-normally distributeddataset into normally distributed dataset. They appliedthe new method to biweekly time series of global nor-malized difference vegetation index (NDVI) from1982 to 2006 to investigate the extreme events on aglobal scale. They found that extreme NDVI eventswere aggregated in Amazonia and in the semi-arid andsemi-humid regions in low andmiddle latitudes, whilethey seldom occurred in high latitudes. This NDVIdata analysis indicates that precipitation was a majorcontrol on the NDVI extremes in low and middle lati-tudes, providing a consistent picture as by FLUXNETdata synthetic analysis (Yi et al 2010, 2014).

5.2. Risk approachRisk analysis is not a new method, but applying it tothe understanding of interactions between extremeevents and ecological resources is challenging.van Oijen et al (2013) introduced a simple probabil-istic method that can be used to assess ecosystem risksimposed by extreme climatic events. In this approach,risk is quantified as the product of two components:the probability of hazardous conditions and the

6

Environ. Res. Lett. 10 (2015) 070201 CYi et al

vulnerability of ecosystems to those hazardous condi-tions. The challenge in this approach is to determinethe threshold values of hazardous climate conditions,which varies between ecosystems and can also evolvewith time. They have used their method to analyze therisk of drought in European forests under recent(1971–2000) and future (2071–2100) climate condi-tions. Vulnerability is defined as the mean differencein primary vegetation productivity between non-drought and drought years.

5.3. Perfect-deficit approachA perfect-deficit approach developed by Yi et al (2012)is different from the widely used, traditional methodof the tails approach (extreme fluctuations relative tomean values). The nature of the perfect-deficitapproach is data-based. During observational period,a perfect growth (carbon uptake) curve can beconstructed by the best growth of ecosystems acrossthe years at the site. Thus, the growth curves of eachyear can be compared with the perfect growth curve.The difference between the perfect-growth curve andyearly growth curve is defined as deficit. Yi et al (2012)derived this approach based on GPP data measuredfrom five grassland tower sites. They found stronglinks between GPP deficits and drought index. Thisapproach can be applied to any continuous dataset ofecosystem–climate interactions. Therefore, the sameextreme events can be cross-verified by differentindependent datasets.

Wei et al (2014) applied the perfect deficitapproach to assess the effects of climate extremes onforest carbon assimilation capacity based on FLUX-NET data and MODIS GPP datasets. They found astrong link between periods of drought and reducedcapacity of trees to absorb carbon. In particular,broadleaf (evergreen and deciduous) trees were verysensitive, whereas coniferous trees were less affected.This is because most broadleaf trees are adapted togenerally well watered and good nutrient conditions,whereas needles are a specific adaptation that evolvedto maximize carbon uptake in water-and tempera-ture-limited regions. Huang et al (2013) applied theperfect-deficit approach to explore how potential pro-ductivity of coniferous plantation forest was sig-nificantly reduced by extreme drought and ice storm.They found that temperature deficits rather thandroughts weremain drivers for theGPP deficits.

5.4. ClimatemitigationAs illustrated in figure 1, more frequent extremeweather and drought events are likely to be associatedwith the warming climate. In this focus issue, we havecollected evidence on how these warming-associatednatural disturbances interrupted current terrestrialcarbon dynamics. The disruption of the terrestrialcarbon dynamics may trigger the potential climate–carbon feedbacks illustrated in figure 2. Le Page et al

(2013) used the Global Change Assessment Model(GCAM) to test what successful climate mitigationshould be undertaken with the natural disturbances.They studied scenarios of stable, increasing anddecreasing natural disturbance rates, with all modelruns stabilizing atmospheric carbon dioxide at a levelthat is expected to keep global temperature increasewithin a certain range above pre-industrial levels by2095. They found that in the case of increaseddisturbance rates, resulting terrestrial emissionswouldhave to be balanced by further deployment of low-carbon technologies for society to achieve a given levelof mitigation. Under a carbon market policy, thisimplies increased carbon prices and mitigation costsup to 2.5 times higher in the most extreme scenario(doubling disturbance rates). As a case study, Camp-bell et al (2012) examined the strategic effects of futuremountaintop coal mining on the carbon budget of thesouthern Appalachian forest region by ecosystemmodeling andfield experiment data.

Acknowledgements

CY is grateful to the International MeteorologicalInstitute for supporting him as a Rossby Fellow tocomplete this synthesis and review.

References

AhrensCD2000Essentials ofMeteorology: An Invitation to theAtmosphere 4th Edn (PacificGrove, CA: Brooks/Cole)

Babst F, CarrerM, Poulter B,Urbinati C,Neuwirth B and FrankD2012 500 years of regional forest growth variability and linksto climatic extreme events in Europe Environ. Res. Lett. 7045705

BoxGEP andCoxDR1964An analysis of transformations J. R.Stat. Soc. Ser.B 26 211–52

BreshearsDD et al 2005Regional vegetation die-off in response toglobal-change-type drought Proc. Natl Acad. Sci. USA 10215144–8

Bright BC,Hicke J A andHudakAT2012 Landscape-scale analysisof aboveground tree carbon stocks affected bymountain pinebeetles in IdahoEnviron. Res. Lett. 7 045702

Campbell J E, Fox J F andActon PM2012 Terrestrial carbon lossesfrommountaintop coal mining offset regional forest carbonsequestration in the 21st century Environ. Res. Lett. 7045701

ChangC-T,WangH-C andHuangC-Y 2013 Impacts of vegetationonset time on the net primary productivity in amountainousisland in Pacific Asia Environ. Res. Lett. 8 045030

Claesson J andNycander J 2013Combined effect of global warmingand increasedCO2-concentration on vegetation growth inwater-limited conditions EcologicalModelling 256 23–30

CoumouD andRahmstorf S 2012A decade ofweather extremesNat. Clim. Change 2 491–6

Cox PM, PearsonD, Booth BB, Friedlingstein P,HuntingfordC,Jones CDand LukeCM2013 Sensitivity of tropical carbon toclimate change constrained by carbon dioxide variabilityNature 494 341–4

Edburg S L et al 2012Cascading impacts of bark beetle-caused treemortality on coupled biogeophysical and biogeochemicalprocesses Frontiers in Ecology and the Environment 10 416–24

Fisk J P,Hurtt GC,Chambers JQ, ZengH,DolanKA andNegrón-Juárez R I 2013The impacts of tropical cyclones on

7

Environ. Res. Lett. 10 (2015) 070201 CYi et al

http://dx.doi.org/10.1088/1748-9326/7/4/045705http://dx.doi.org/10.1088/1748-9326/7/4/045705http://dx.doi.org/10.1073/pnas.0505734102http://dx.doi.org/10.1073/pnas.0505734102http://dx.doi.org/10.1073/pnas.0505734102http://dx.doi.org/10.1073/pnas.0505734102http://dx.doi.org/10.1088/1748-9326/7/4/045702http://dx.doi.org/10.1088/1748-9326/7/4/045701http://dx.doi.org/10.1088/1748-9326/7/4/045701http://dx.doi.org/10.1088/1748-9326/8/4/045030http://dx.doi.org/10.1016/j.ecolmodel.2013.02.007http://dx.doi.org/10.1016/j.ecolmodel.2013.02.007http://dx.doi.org/10.1016/j.ecolmodel.2013.02.007http://dx.doi.org/10.1038/nclimate1452http://dx.doi.org/10.1038/nclimate1452http://dx.doi.org/10.1038/nclimate1452http://dx.doi.org/10.1038/nature11882http://dx.doi.org/10.1038/nature11882http://dx.doi.org/10.1038/nature11882http://dx.doi.org/10.1890/110173http://dx.doi.org/10.1890/110173http://dx.doi.org/10.1890/110173

the net carbon balance of easternUS forests (1851–2000)Environ. Res. Lett. 8 045017

Fernández-GiménezME, Batkhishig B andBatbuyan B 2012Cross-boundary and cross-level dynamics increase vulnerability toseverewinter disasters (dzud) inMongoliaGlob. Environ.Change 22 836–51

GalvagnoM et al 2013 Phenology and carbon dioxide source/sinkstrength of a subalpine grassland in response to anexceptionally short snow seasonEnviron. Res. Lett. 8 025008

Hicke J A,Meddens A JH, AllenCDandKoldenCA2013Carbonstocks of trees killed by bark beetles andwildfire in thewesternUnited States Environ. Res. Lett. 8 035032

HuangK,Wang S, ZhouL,WangH,LiuY andYangF2013Effects ofdrought and ice rainonpotential productivity of a subtropicalconiferous plantation from2003 to 2010basedon eddy covariancefluxobservationEnviron. Res. Lett.8 035021

Hutley LB, Evans B J, Beringer J, CookGD,Maier SWandRazon E2013 Impacts of an extreme cyclone event on landscape-scalesavannafire, productivity and greenhouse gas emissionsEnviron. Res. Lett. 8 045023

JungM et al 2010Recent decline in the global landevapotranspiration trend due to limitedmoisture supplyNature 467 951–4

JohnR, Chen J, Ou-Yang Z-T, Xiao J, Becker R, SamantaA,Ganguly S, YuanWandBatkhishigO 2013Vegetationresponse to extreme climate events on theMongolian Plateaufrom2000 to 2010Environ. Res. Lett. 8 035033

KangXM et al 2013 Effects of grazing onCO2 balance in a semiaridsteppe: field observations andmodeling Journal of Soils andSediments 13 1012–23

KininmonthW2010Clausius-clapeyron and the regulation ofglobal warming Fisica E 26 61–70

KolbT,Dore S andMontes-HeluM2013 Extreme late-summerdrought causes neutral annual carbon balance insouthwestern ponderosa pine forests and grasslands Environ.Res. Lett. 8 015015

KurzWA et al 2008Mountain pine beetle and forest carbonfeedback to climate changeNature 452 987–90

Le Page Y et al 2013 Sensitivity of climatemitigation strategies tonatural disturbancesEnviron. Res. Lett. 8 015018

LiuG, LiuH andYin Y 2013Global patterns ofNDVI-indicatedvegetation extremes and their sensitivity to climate extremesEnviron. Res. Lett. 8 025009

Logan J, Regniere J and Powell J A 2003Assessing the impacts ofglobal warming on forest pest dynamics Frontiers in Ecologyand the Environment 1 130–7

LundM,Christensen TR, Lindroth A and Schubert P 2012 Effects ofdrought conditions on the carbon dioxide dynamics in atemperate peatlandEnviron. Res. Lett. 7 045704

MedvigyD, ClarkK L, SkowronskiN S and Schäfer KVR2012Simulated impacts of insect defoliation on forest carbondynamics Environ. Res. Lett. 7 045703

OdumEP1969The strategy of ecosystem development Science 164262–70

ReedDE, Ewers B E and Pendall E 2014 Impact ofmountain pinebeetle inducedmortality on forest carbon andwaterfluxesEnviron. Res. Lett. 9 105004

Randerson J T et al 2006The impact of boreal forestfire on climatewarming Science 314 1130–2

Seneviratne S I et al 2012Changes in climate extremes and theirimpacts on the natural physical environment 109–230Managing the Risks of Extreme Events andDisasters to AdvanceClimate Change Adaptation: A Special Report ofWorkingGroups I and II of the Intergovernmental Panel on ClimateChange

Seneviratne S I et al 2014No pause in the increase of hottemperature extremesNature Clim. Change 4 161–3

ShaoC,Chen J and Li L 2013Grazing alters the biophysicalregulation of carbon fluxes in a desert steppeEnviron. Res.Lett. 8 025012

SunY,Gu L,DickinsonRE andZhouB 2012 Forest greenness afterthemassive 2008Chinese ice storm: integrated effects ofnatural processes and human intervention Environ. Res. Lett.7 035702

TeulingA J et al 2013 Evapotranspiration amplifies Europeansummer droughtGeophys. Res. Lett. 40 2071–5

Vargas R 2012How a hurricane disturbance influences extremeCO2fluxes and variance in a tropical forestEnviron. Res. Lett.7 035704

vanOijenM, Beer C, CramerW,RammigA, ReichsteinM,Rolinski S and Soussana J-F 2013Anovel probabilistic riskanalysis to determine the vulnerability of ecosystems toextreme climatic events Environ. Res. Lett. 8 015032

Wei S et al 2014Data-based perfect-deficit approach tounderstanding climate extremes and forest carbonassimilation capacityEnviron. Res. Lett. 9 065002

Wolf S, EugsterW,AmmannC,HäniM, Zielis S,Hiller R, Stieger J,ImerD,Merbold L andBuchmannN2013Contrastingresponse of grassland versus forest carbon andwaterfluxes tospring drought in Switzerland Environ. Res. Lett. 8 035007

WorldMeteorological Organization (WMO) 2013The globalclimate 2001–2010: a decade of climate extremes Summaryreport (Geneva:WorldMeteorological Organization) no 1119http://library.wmo.int/pmb_ged/wmo_1119_en.pdf

WorldMeteorological Organization 2011Weather Extremes in aChanging Climate: Hindsight on Foresight (Geneva:WorldMeteorological Organization) no 1075

XuX, Piao S,WangX, ChenA,Ciais P andMyneni RB 2012 Spatio-temporal patterns of the area experiencing negativevegetation growth anomalies inChina over the last threedecadesEnviron. Res. Lett. 7 035701

YiC,Wei S andHendreyG2014Warmingclimate extendsdryness-controlledareasof terrestrial carbonsequestrationSci.Rep.45472

YiC et al 2012Climate extremes and grassland potentialproductivity Environ. Res. Lett. 7 035703

YiC et al 2010Climate control of terrestrial carbon exchange acrossbiomes and continents Environ. Res. Lett. 5 1748–9326

Zhang L, Xiao J, Li J,WangK, Lei L andGuoH2012The 2010 springdrought reduced primary productivity in southwesternChinaEnviron. Res. Lett. 7 045706

Zscheischler J et al 2014A few extreme events dominate globalinterannual variability in gross primary production Environ.Res. Lett. 9 035001

8

Environ. Res. Lett. 10 (2015) 070201 CYi et al

http://dx.doi.org/10.1088/1748-9326/8/4/045017http://dx.doi.org/10.1016/j.gloenvcha.2012.07.001http://dx.doi.org/10.1016/j.gloenvcha.2012.07.001http://dx.doi.org/10.1016/j.gloenvcha.2012.07.001http://dx.doi.org/10.1088/1748-9326/8/2/025008http://dx.doi.org/10.1088/1748-9326/8/3/035032http://dx.doi.org/10.1088/1748-9326/8/3/035021http://dx.doi.org/10.1088/1748-9326/8/4/045023http://dx.doi.org/10.1038/nature09396http://dx.doi.org/10.1038/nature09396http://dx.doi.org/10.1038/nature09396http://dx.doi.org/10.1088/1748-9326/8/3/035033http://dx.doi.org/10.1007/s11368-013-0675-5http://dx.doi.org/10.1007/s11368-013-0675-5http://dx.doi.org/10.1007/s11368-013-0675-5http://dx.doi.org/10.1088/1748-9326/8/1/015015http://dx.doi.org/10.1038/nature06777http://dx.doi.org/10.1038/nature06777http://dx.doi.org/10.1038/nature06777http://dx.doi.org/10.1088/1748-9326/8/1/015018http://dx.doi.org/10.1088/1748-9326/8/2/025009http://dx.doi.org/10.1890/1540-9295(2003)001[0130:ATIOGW]2.0.CO;2http://dx.doi.org/10.1890/1540-9295(2003)001[0130:ATIOGW]2.0.CO;2http://dx.doi.org/10.1890/1540-9295(2003)001[0130:ATIOGW]2.0.CO;2http://dx.doi.org/10.1088/1748-9326/7/4/045704http://dx.doi.org/10.1088/1748-9326/7/4/045703http://dx.doi.org/10.1126/science.164.3877.262http://dx.doi.org/10.1126/science.164.3877.262http://dx.doi.org/10.1126/science.164.3877.262http://dx.doi.org/10.1126/science.164.3877.262http://dx.doi.org/10.1088/1748-9326/9/10/105004http://dx.doi.org/10.1126/science.1132075http://dx.doi.org/10.1126/science.1132075http://dx.doi.org/10.1126/science.1132075http://dx.doi.org/10.1038/nclimate2145http://dx.doi.org/10.1038/nclimate2145http://dx.doi.org/10.1038/nclimate2145http://dx.doi.org/10.1088/1748-9326/8/2/025012http://dx.doi.org/10.1088/1748-9326/7/3/035702http://dx.doi.org/10.1002/grl.50495http://dx.doi.org/10.1002/grl.50495http://dx.doi.org/10.1002/grl.50495http://dx.doi.org/10.1088/1748-9326/7/3/035704http://dx.doi.org/10.1088/1748-9326/8/1/015032http://dx.doi.org/10.1088/1748-9326/9/6/065002http://dx.doi.org/10.1088/1748-9326/8/3/035007http://library.wmo.int/pmb_ged/wmo_1119_en.pdfhttp://dx.doi.org/10.1088/1748-9326/7/3/035701http://dx.doi.org/10.1038/srep05472http://dx.doi.org/10.1038/srep05472http://dx.doi.org/10.1088/1748-9326/7/3/035703http://dx.doi.org/10.1088/1748-9326/5/3/034007http://dx.doi.org/10.1088/1748-9326/5/3/034007http://dx.doi.org/10.1088/1748-9326/5/3/034007http://dx.doi.org/10.1088/1748-9326/7/4/045706http://dx.doi.org/10.1088/1748-9326/9/3/035001

1. Introduction2. Drought2.1. Spring drought2.2. Late-summer drought2.3. Long-term drought events

3. Extreme weather3.1. Hurricane3.1.1. Hurricane-savanna-fire interactions3.1.2. Effect of hurricanes on the soil carbon flux3.1.3. Carbon footprint of historical hurricanes in eastern US

3.2. Ice storms3.3. Extreme snow events3.3.1. Extreme short snow season3.3.2. Summer drought worsens impacts of bitter winter in Mongolia Plateau

4. Bark beetles5. Quantitative approaches5.1. Tail Approach5.2. Risk approach5.3. Perfect-deficit approach5.4. Climate mitigation

AcknowledgementsReferences