Tropical cyclone cooling combats region-wide coral bleaching

Tropical cyclone cooling combats region-wide coralbleachingADAM D . CARR IGAN and MARJI PUOTINEN

Institute for Conservation Biology and Environmental Management and School of Earth and Environmental Sciences, University

of Wollongong, Wollongong, NSW, Australia

Abstract

Coral bleaching has become more frequent and widespread as a result of rising sea surface temperature (SST). During

a regional scale SST anomaly, reef exposure to thermal stress is patchy in part due to physical factors that reduce SST

to provide thermal refuge. Tropical cyclones (TCs – hurricanes, typhoons) can induce temperature drops at spatial

scales comparable to that of the SST anomaly itself. Such cyclone cooling can mitigate bleaching across broad areas

when well-timed and appropriately located, yet the spatial and temporal prevalence of this phenomenon has not

been quantified. Here, satellite SST and historical TC data are used to reconstruct cool wakes (n=46) across the Carib-

bean during two active TC seasons (2005 and 2010) where high thermal stress was widespread. Upon comparison of

these datasets with thermal stress data from Coral Reef Watch and published accounts of bleaching, it is evident that

TC cooling reduced thermal stress at a region-wide scale. The results show that during a mass bleaching event, TC

cooling reduced thermal stress below critical levels to potentially mitigate bleaching at some reefs, and interrupted

natural warming cycles to slow the build-up of thermal stress at others. Furthermore, reconstructed TC wave damage

zones suggest that it was rare for more reef area to be damaged by waves than was cooled (only 12% of TCs). Extend-

ing the time series back to 1985 (n = 314), we estimate that for the recent period of enhanced TC activity (1995–2010),the annual probability that cooling and thermal stress co-occur is as high as 31% at some reefs. Quantifying such

probabilities across the other tropical regions where both coral reefs and TCs exist is vital for improving our under-

standing of how reef exposure to rising SSTs may vary, and contributes to a basis for targeting reef conservation.

Keywords: climate change, coral bleaching, coral reef, cyclone cooling, hurricane, sea surface temperature, thermal refuge, ther-

mal stress, tropical cyclone

Received 22 July 2013; revised version received 25 December 2013 and accepted 6 January 2014

Introduction

Widespread concern for the future of the world’s coral

reefs (Hughes et al., 2003; Bellwood et al., 2004) has

been justified by reports of regional declines in coral

cover (Caribbean – Gardner et al., 2003; Indo-Pacific –Bruno et al., 2007; Great Barrier Reef – De’ath et al.,

2012). Indeed, where coral cover drops below 10% –common for the Caribbean – reefs may be unable to

sustain positive rates of carbonate production, leading

to an eventual loss of entire reef structures (Perry et al.,

2013). A major factor driving coral cover declines is a

combination of broad and local scale stressors, the

intervals between which are at times too short for full

recovery to occur, leading to a loss of resilience

(Hughes et al., 2003). In recent decades, thermal stress

has caused extensive coral mortality (Baird & Marshall,

2002) and morbidity (Mendes & Woodley, 2002) via

mass bleaching throughout the world’s reef provinces

(Hoegh-Guldberg, 1999; Wilkinson, 2008). With SST

already approaching critical thresholds in some regions

(Lough, 2012), mass bleaching events could become an

annual occurrence by mid-century (Donner et al., 2005;

van Hooidonk et al., 2013).

Prolonged exposure to elevated SST (thermal stress)

during the annual warm season (summer to autumn)

can harm corals throughout entire regions. Mass bleach-

ing occurs when anomalously high SST (~4 weeks at

1 °C above summer maximum Glynn, 1996; Hoegh-

Guldberg, 1999; Baker et al., 2008) trigger a breakdown

of the symbiosis between the coral host and its algal

symbionts (zoozanthellae). How reefs respond to a

given level and duration of thermal stress depends on

the physiology of individual corals and the symbionts

that reside within them (Grottoli et al., 2006; McClana-

han et al., 2007a; Oliver & Palumbi, 2011). Whether or

not thermal stress occurs, and how long it persists, at a

given location within a broad area of elevated SST

depends on the local physical environment. Thus, both

the distribution of bleaching from thermal stress, and

the distribution of thermal stress itself within area of

high SST, is inherently patchy. Identifying broad reefCorrespondence: A. D. Carrigan, tel. 0433071942, fax 02 4221 4250,

e-mail: [email protected]

1© 2014 John Wiley & Sons Ltd

Global Change Biology (2014), doi: 10.1111/gcb.12541

Global Change Biology

areas less likely to be exposed to thermal stress within

areas of predicted high SST nonetheless has been iden-

tified as a basis for prioritizing their conservation (for

example, Ban et al., 2012).

It has been well documented that various local scale

physical processes that cool SST, such as upwelling,

cold currents, and wind-driven mixing, create pockets

of thermal refuge within areas of otherwise elevated

SST (Glynn, 1996; Riegl & Piller, 2003; West & Salm,

2003; Skirving et al., 2006). Although it remains largely

unexplored, tropical cyclones (TCs – hurricanes,

typhoons) tracking near reefs when thermal stress is

high or on the rise can also provide refuge to reefs by

generating a ‘cool wake’ of lowered SST. During the

annual warm season, when SST is sufficiently high for

tropical cyclogenesis, TCs can lower SST by up to 10 °C(Chiang et al., 2011), though drops usually fall in the

range of 1°–6° (Price, 1981). Spatially, the extent of a

cool wake can rival the scale of a preexisting warm SST

anomaly. However, the total extent and shape of a cool

wake, as well as the magnitude of cooling within it,

varies considerably along a given TC’s track. These

variations arise from differences in the state of the

ocean and the TC itself. For example, the magnitude of

cooling is maximized where subsurface cold water is

readily accessible (i.e. shallow depth to mixed layer),

the TC is intense, and the TC moves slowly (Price, 1981;

Bender et al., 1993). After a TC passes, it takes an aver-

age of 40 days for the cooling signal to subside, though

SSTs rarely fully recover to pre-TC levels (Lloyd &

Vecchi, 2011; Vincent et al., 2012).

Such broad-scale (though spatially complex) and per-

sistent ocean cooling can slow or prevent the build-up

of thermal stress to critical thresholds and return SST to

climatological levels (Dare & McBride, 2011; Lloyd &

Vecchi, 2011). Likewise, cooling late in the annual

warm season can speed the onset of seasonal SST

reductions (Dare & McBride, 2011) and has the poten-

tial to prevent or minimize bleaching that is imminent

or underway. For example, lower thermal stress at

some reefs observed near TC tracks during the 2005

Caribbean mass bleaching event (Wilkinson & Souter,

2008; Eakin et al., 2010) reduced the duration and sever-

ity of bleaching (Manzello et al., 2007). Despite this, the

effect of TC cooling on thermal stress has not been

quantified on a regional basis, nor its prevalence across

coral reef areas examined. Consequently, most broad-

scale studies of stressors on reefs either do not consider

the effect of TC cooling on the likelihood of bleaching

(e.g., Maina et al., 2008, 2011) or do not model it

spatially explicitly (Edwards et al., 2011).

Although TC waves can cause catastrophic physical

damage to reefs, this is typically limited to patchy

areas within a narrow but variable swath along the

storm track (<100 km–Done, 1992; Fabricius et al.,

2008). Given that cool wakes tend to be much more

extensive than this (100s of km from the track – Stra-

mma et al., 1986), it is probable that more reef area

will often be beneficially cooled than damaged from a

given TC (as noted in Manzello et al., 2007). Further,

even if the overall frequency of global TCs remains

constant or declines slightly as currently predicted

(reviewed by Knutson et al., 2010), continued

increases in the frequency and intensity of ocean

warming across global coral reef areas will provide

ample opportunities for TCs to track near reefs when

thermal stress is high.

Here, we use satellite SST, historical TC data and a

meteorological model to reconstruct cool wakes and

catastrophic wave damage zones (n = 46) during two

active TC seasons (2005 and 2010) where high thermal

stress was prevalent across the Caribbean. Comparing

this to Coral Reef Watch’s (CRW) thermal stress data

and published accounts of bleaching, we assess the

extent to which thermal stress varied in response to TC

cooling. We further test whether the net effect of each

TC was positive (more cooling than damage) by recon-

structing the likely wave damage zone and spatially

comparing it to the associated cool wake. Extending the

time series to 1985–2010 (n = 314 storms), we estimate

the annual probability that cooling and thermal stress

will co-occur at Caribbean reefs.

Materials and methods

Study area

In 2005, pervasive anomalously warm SST in the Caribbean

caused widespread coral mortality from bleaching (Wilkin-

son & Souter, 2008; Eakin et al., 2010). This was coupled

with an unusually active TC season (Shein et al., 2006). Then,

in 2010, record setting SST again led to widespread bleach-

ing (e.g., Bayraktarov et al., 2012; del M�onaco et al., 2012),

accompanied by another active TC season (Blunden et al.,

2011). This combination of high thermal stress and TC activ-

ity frequently occurs across the region (Carrigan & Puotinen,

2011), particularly since the start of a multidecadal TC active

phase in 1995 (Bell & Chelliah, 2006). The Caribbean is

therefore an ideal setting to explore the relationship between

thermal stress and TC cooling on a regional basis. TCs and

elevated SST are possible within the North Atlantic in a

given year between June and December. We thus define a

combined ‘season’ of potential thermal stress and TC activity

as early (June–August) and late (September–December). We

consider coral reefs of the entire Greater Caribbean region

including the Gulf of Mexico, Florida, The Bahamas, and

Bermuda. Coral reef locations were sourced from www.

unep-wcmc.org (1 km resolution), and were aggregated into

~50 km reef cells (n = 477) to match the resolution of CRW’s

thermal stress data.

© 2014 John Wiley & Sons Ltd, Global Change Biology, doi: 10.1111/gcb.12541

2 A. D. CARRIGAN & M. PUOTINEN

Tropical cyclone cool wakes

For each TC (minimum 17 ms�1 wind speed) that tracked

through the Caribbean region in 2005 (n = 27) and 2010

(n = 19), sustained SST drops were calculated using the

National Oceanic and Atmospheric Administration’s (NOAA)

daily 1/4° optimum interpolation (OI) AVHRR SST product

(ftp://eclipse.ncdc.noaa.gov/pub/OI-daily/). The extent to

which post-TC conditions differed from ambient SST condi-

tions was established by comparing 14 day composite SST

images before (pre-TC) and after (post-TC) storm passage. Dif-

ferencing the pre- and post-TC SST images yielded a 14 day

mean SST decrease (DSST), providing a metric to capture

reductions in the both the magnitude and duration of thermal

stress. The date at which each TC began and ended was

derived separately for each grid cell in the study area to

ensure that the correct SST grids were used to calculate

averages. This was necessary because unusually long lasting

and/or fast moving TCs can traverse significant distances

(1000s of km) before dissipating, and thus the time at which

grid cells experience the onset and end of TC conditions may

vary. The resultant cool wakes were constrained to the radius

of gale force winds (17 ms�1 – TC footprint) to ensure cooling

was TC-induced. This is a reasonable assumption because ver-

tical mixing of the ocean from TC winds is responsible for as

much as 80% of the SST drop (Price, 1981; Huang et al., 2009).

However, a small component of observed cooling in the study

region could be caused by coastal upwelling (e.g. along the

South American coast – Gordon, 1967) or interaction with cold

core eddies which are known to enhance cooling within TC

wakes (Jaimes & Shay, 2009). Further, some TC related cooling

could occur beyond the TC footprint from advection of cooled

waters via TC-induced and existing regional currents. TC

track and gale radius data were obtained from NOAA’s

National Hurricane Center (NHC). Accumulated cooling over

a season in a given year was calculated by summing all

observed cooling at each pixel from June to December in that

year.

Thermal stress indices

Thermal stress during the 2005 and 2010 seasons was mea-

sured using Coral Reef Watch’s (CRW) biweekly degree heat-

ing week (DHW) and hotspot thermal stress indices (Liu et al.,

2006). CRW hotspots measure instantaneous thermal stress,

whereas the DHW metric measures the accumulation of ther-

mal stress (based on a maximum monthly mean climatology)

over a rolling 12 week time period. Field data validation has

demonstrated that DHW values of 4 or more and values of 8

or more indicate likely and severe bleaching, respectively,

though local scale patterns of bleaching within these areas

have been observed to be patchy (McClanahan et al., 2007b).

DHW values just below this can affect coral growth, calcifica-

tion and fecundity (Hoegh-Guldberg, 1999; Mendes & Wood-

ley, 2002), as well as depleting energy reserves (Anthony et al.,

2008), leaving corals more susceptible to future disturbances

(e.g., disease; Bruno et al., 2007). Near-real time hotspot data

were accessed directly from CRW’s website (www.coralreef-

watch.noaa.gov) and retrospective annual maximum DHW

data (1985–2009) were provided by CRW. To assign values to

reef cells masked by land, spatial averaging was employed fol-

lowing Eakin et al. (2010). A time evolution of thermal stress

for the respective seasons was calculated by averaging Hot-

spot and DHW grids across Caribbean reef cells at a biweekly

time step from June to December. The influence of TC cooling

was examined by classifying the resultant time series into

reef cells with significant (DSST ≥ 1 °C) or insignificant

(DSST < 1 °C) accumulated cooling for the respective seasons.

To simulate what thermal stress levels would have been in the

absence of TCs during 2005 and 2010, accumulated seasonal

cooling was added to the annual maximum DHW grids.

Thermal stress and tropical cyclone history in theCaribbean

To examine the likelihood that cooling and thermal stress

coincided in a given year, sustained TC cooling was calculated

as above (DSST) for all TCs reaching gale-force intensity dur-

ing the 25 year period 1985–2010 (n = 314). For each year,

counts of individual cooling events with DSST ≥ 1 °C(2 weeks sustained drop) were compiled for all reef cells. This

threshold was chosen based on the assumption that a 2 weeks

sustained drop of 1 °C is capable of reducing a given reef

cell’s thermal stress level by at least 1 DHW and potentially

below thresholds for mild or severe bleaching. Note that this

represents a conservative approach. For example, a combina-

tion of several weak TCs that generate sub-1 degree cooling

early in the season could combine to keep thermal stress low

but would not be counted as cooling events. Annual retrospec-

tive maximum DHW data spanning 1985 to 2010 were used to

characterize historical thermal stress at reefs during low TC

(1985–1994) and high TC (1995–2010) periods (Bell & Chelliah,

2006).

A reef cell was assumed to be exposed to thermal stress if

the maximum annual DHW equalled or exceeded three in a

given year. Although the onset of bleaching is predicted to

occur at four DHW, a value of three was used to allow for the

possibility of a preventative TC cooling effect and sublethal

thermal stress relief. The average number of years for which

an interaction was likely (co-occurrence of at least one cooling

event and DHW ≥ 3) was calculated separately for the low

and high periods of TC activity. Following previous studies of

TC landfall probabilities (Tartaglione et al., 2003; Klotzbach,

2011), the Poisson probability formula (Pr(X ≥ 1) = 1–e�k) was

used to calculate the probability of an interaction for each

~50 km reef cell over the respective periods, where k is the

average number of years with at least one interaction.

Tropical cyclone net effects

The net effect of each TC was estimated by comparing the area

of reef cooled to the area likely damaged by waves. We focus

on catastrophic physical damage to corals (widespread break-

age and dislodgement of colonies, removal of entire sections

of reef framework) for which recovery may take decades to

centuries (Hughes & Connell, 1999). Here, to ensure cooling

was sufficiently intense to influence thermal stress at reefs,


CYCLONE COOLING COMBATS REGION-WIDE BLEACHING 3

only sustained drops of at least 1 °C were included in the cool

wake. To model the area of expected damage, the spatial dis-

tribution of TC winds for each storm were calculated every

hour along the track as 10-min maximum surface speeds using

a parametric wind model (Holland et al., 2010) that was driven

by TC data (e.g., central pressure, gale-force wind radius)

obtained from the NHC. A TC wave damage zone was then

reconstructed for each TC (n = 46) by: (i) modeling wind

speeds across the study area every hour during the storm, (ii)

calculating maximum wind speeds over the life of the storm

(as per Puotinen, 2007; Fabricius et al., 2008; De’ath et al.,

2012), and (iii) applying a threshold for damage identified

from field data in Gardner et al. (2005) which found coral

cover loss at sites where maximum TC winds met or exceeded

50 ms�1. Using this approach tends to overestimate the total

area of reef damage because actual coral colony exposure to

wave energy and susceptibility to physical damage within this

zone is highly variable, resulting in some false positives. The

net effect of each TC was calculated by subtracting the total

reef area located within the damage zone from the total reef

area located within the cool wake. Positive values indicate that

more reef area was cooled than damaged (net benefit) while

negative values indicate more reef area was damaged than

cooled (net cost).

Results

Tropical cyclone cooling during thermal stress events in2005 and 2010

During both 2005 and 2010, the most severe thermal

stress was situated in the southeast Caribbean (Fig. 1c,

f), while in 2010, the ‘warm pool’ of water extended fur-

ther south and westward, adjacent to the north coast of

South America (Fig. 1f). These warm areas coincide

with a general absence (2005–Fig. 1b) and low levels

(2010–Fig. 1e) of accumulated TC cooling. Higher levels

of cooling elsewhere may have prevented warm pools

from horizontally advecting further westward (via

major westward flowing currents) or developing alto-

gether. Similarly, concentrations of high accumulated

cooling (2005 – Belize, Mexico to Florida coast in

Fig. 1b; 2010 - Belize in Fig. 1e) correspond to generally

low levels of thermal stress (Fig. 1c, f). However, if

thermal stress is sufficiently elevated, levels can remain

high even after TC-induced cooling, as was evident for

the Bahamas in 2005. Red circles and the red square

(a) (b) (c)

(d) (e) (f)

Fig. 1 Tropical cyclone (TC) cooling and thermal stress near coral reefs: Caribbean, 2005 and 2010. Accumulated TC-induced cooling

(b) and (e), accumulated thermal stress (c) and (f) and tracks of TCs with recorded winds of at least gale force (17 ms�1; (a) and (d) for

2005 and 2010 seasons, respectively. Accumulated cooling is the sum of all sustained (2 week average) TC-induced sea surface tempera-

ture drops that occurred in each 28 km cell throughout Caribbean during the warm season (June–December) in a given year. Thermal

stress is the maximum degree heating week (DHW) value to occur in each 50 km cell over the season. For TC tracks, line thickness

denotes storm intensity (dashed lines depict winds <33 ms�1) and color indicates whether a TC formed early (pre-August; gray) or late

(post-August; black) in the TC season. In all panels, coral reef locations are depicted in orange. The location of Manzello et al. (2007)

study sites from 2005 are shown at Florida Reef Tract (red circles) and the US Virgin Islands (red square) in panels a-c.



show the relative location of the sites where in situ mea-

surements of bleaching and SST change were taken by

Manzello et al. (2007) during the 2005 season at the

Florida Reef Tract and the US Virgin Islands, respec-

tively. TC activity (Fig. 1a) and associated cooling

(Fig. 1b) were clearly absent for the latter where bleach-

ing was observed to be more severe and persistent,

with extremely high levels of accumulated thermal

stress (Fig. 1c). In contrast, moderate levels of cooling

(Fig. 1b) from several TCs (Fig. 1a) was evident at pix-

els surrounding the Florida Reef Tract sites, at which

accumulated thermal stress peaked at the level capable

of causing bleaching (4 DHW – Fig. 1c). Further,

although bleaching was reported throughout the Carib-

bean during 2005 from a range of sources (Fig. 1 in

Eakin et al., 2010), it was most widespread where ther-

mal stress was highest and TC cooling was lowest (far

southeast). Similarly in 2010, thermal stress was con-

centrated in the southeast, where there was bleaching

reported at reefs with minimal TC cooling (Columbia –Bayraktarov et al., 2012; Venezuela – del del M�onaco

et al., 2012, Lesser Antilles –reefbase.org).The timing of TC activity and thermal stress was

similar in each season, peaking in late summer/early

autumn (post-August) as SSTs in the tropics approached

their seasonal maximums. In 2005, the strongest TCs

(solid lines) are concentrated in the Western Caribbean

(Fig. 1a), as opposed to the open North Atlantic Ocean

in 2010 (Fig. 1d). This corresponds to the concentration

of highest cooling near Mexico, Florida, the Bahamas,

and Cuba in 2005 (Fig. 1b) and in the open ocean in

2010 (Fig. 1e). On the other hand, weaker TCs are

found throughout the Caribbean and North Atlantic

during both seasons. Although weak TCs do not always

generate notable cooling, their lower wind speeds typi-

cally result in less severe and more localized damage to

reefs (Fabricius et al., 2008).

In both 2005 and 2010, the maximum duration and

magnitude of thermal stress was lower at reef areas

(~50 km cells containing reefs) cooled (accumulated

seasonal DSST ≥ 1 °C) by nearby TCs (Fig. 2). A clear

divergence in accumulated thermal stress, measured by

the DHW metric, is evident between cooled reef cells

(dashed lines) and those with negligible accumulated

cooling (<1 °C–solid lines). Beginning in mid-August

2005 (Fig. 2a) and late-September 2010 (Fig. 2b), the

accumulation of DHWs at cooled reef areas plateaued

while DHWs at uncooled reef areas continued to rise.

As a result, TC-cooled reef areas experienced lower

peak thermal stress levels and shorter exposure times

to thermal stress than those reefs not cooled (dashed

lines below and/or to the left of solid lines on Fig. 2).

For both years, this meant the peak of the mean DHW

curve dropped to or below four (threshold at which

bleaching may occur) for TC-cooled reef areas. Rapid

drops in CRW’s hotspot metric (light gray lines on

Fig. 2), which measures thermal stress at a given time,

reflect the influence of individual TCs and follow a sim-

ilar trend of lower values at reef areas cooled by nearby

TCs. The cumulative effect of TC cooling becomes evi-

dent later in the season in both cases because most of

the strong TCs (thick lines on Fig. 1a, d) formed after

August.

To model likely thermal stress levels in the absence

of TCs, accumulated seasonal cooling was added to the

maximum DHW grids (Fig. 3). Nearly 75% (2005 –Fig. 3a) and 90% (2010 – Fig. 3b) of reef pixels were

likely cooled to some degree by nearby TCs (nonwhite

areas on pie charts). This relief was sufficient to either

prevent exceeding the bleaching threshold (green areas)

or reduce the severity of potential bleaching (yellow

areas) for about a quarter (2005– Fig. 3a) and one-fifth

(2010 – Fig. 3b) of reef cells during the respective sea-

sons. Although TC-induced cooling affected more reef

cells in 2010 than in 2005, a greater proportion of reef

areas were cooled enough to prevent or reduce the

severity of bleaching in 2005 (green, yellow and pink

areas in Fig. 3a).

(a)

(b)

Fig. 2 Biweekly progression of thermal stress and tropical

cyclone (TC) cooling: Caribbean, 2005 and 2010. Time evolution

of mean hotspot (gray lines) and mean DHW (black lines)

biweekly thermal stress data at ~50 km cells containing reefs

(n = 477; reef cells) across the Caribbean for the 2005 (a) and

2010 (b) Atlantic warm seasons (June–December). Reef cells

with an accumulated TC-induced cooling (based on 2 week sus-

tained drop) of 1 °C or greater are depicted by dashed lines

while solid lines represent reef cells with negligible TC-induced

cooling.



Probability of future interactions

In the Caribbean, the current active TC period coupled

with increasing levels of thermal stress makes it proba-

ble that TC-induced cooling will prevent or mitigate

bleaching at some reefs in future seasons. To explore

how the likelihood of this interaction varies spatially,

the probability that high thermal stress and TC cooling

coincided at each reef cell in a given season was calcu-

lated for both the inactive (i.e. low thermal stress/low

TC activity; Fig. 4a) and the active (i.e. high thermal

stress/high TC activity; Fig. 4b) periods. The resultant

probabilities show that, during the active period, some

reef areas are more likely to experience TC cooling dur-

ing anomalously high SST than others and the location

of such reefs varies spatially. During the inactive per-

iod, however, the chance of TC cooling during high

thermal stress was uniformly low (Fig. 4a). The highest

probabilities (up to 31% – light blue and green squares

on Fig. 4b) were found in the north around the Baha-

mas and Bermuda, and in southeast along the Lesser

Antilles.

Net effect of Caribbean tropical cyclones

A comparison of cooling and modeled wave damage

near reefs during the 2005 and 2010 seasons reveals that

less than one-third of TCs produced a net effect,

whether positive (22%) or negative (4%–Fig. 5). The

remaining TCs had no effect because neither any

(a) (b)

Fig. 3 Estimated effect of tropical cyclone (TC) cooling on accumulated thermal stress: Caribbean, 2005 (a) and 2010 (b). Colors indicate

the difference in thermal stress that would have occurred at each pixel in the absence of TC cooling: an increase in thermal stress insuf-

ficient to cause bleaching (blue), enough to result in bleaching (DHW > = 4 – green), enough to make bleaching become severe (DHW >

8 – yellow), and enough to raise thermal stress from none to severe bleaching (pink). Pie diagrams represent the proportion of 56 km

reef pixels (n = 477) for which each level of change in thermal stress applies. Coral reef locations are shown in black on the maps.

(a) (b)

Fig. 4 Probability of beneficial tropical cyclone (TC) cooling in the Caribbean, 1985–2010. Beneficial cooling is assumed to occur when

high thermal stress (DHW ≥3) and a TC cooling event of at least 1 °C co-occur in a given year at 56-km reef cells in the Caribbean. Prob-

abilities are presented separately for the low TC/thermal stress period (1985–1994); (a) and the recent high TC/thermal stress period

(1995–2010; (b). The maximum probability was 0.31.



cooling nor damaging waves they generated occurred

near reefs, particularly in 2010. Positive effects (net ben-

efits) were five times more prevalent than negative

effects. While over a quarter of the TCs generated

winds that exceeded the damage threshold (50 ms�1) at

some point along their tracks, the storms tended to be

weak (dashed rather than solid lines on Fig. 5) when

tracking near reefs, especially during 2010. Conse-

quently, only 9% of the TCs produced damage zones

that encompassed reefs.

Discussion

This study builds on previous work to provide the first

spatially explicit examination of the extent to which

TC-induced cooling reduced region-wide thermal

stress within a season. Prior work exploring the rela-

tionship between TC cooling and thermal stress at

coral reefs either assumed uniform cooling within the

zone of TC-induced gale-force winds (Carrigan & Puo-

tinen, 2011), or that cooling from any TC tracking near

reefs was sufficient to prevent bleaching (Edwards

et al., 2011). Notably, we used satellite SST and histori-

cal TC data to delineate the magnitude and extent of

actual cooling generated by each TC within a given

season. This is important as the magnitude of cooling

is rarely uniform within the wake, and the extent and

spatial configuration of wakes vary considerably. Simi-

larly, Edwards et al. (2011) improved upon earlier

studies that assumed TC reef damage occurs within a

set distance of a TC track (Woodley, 1992; Treml et al.,

1997; Puotinen, 2004) by accounting for known higher

wind speeds to the right of the track (in northern

hemisphere). We extend this by reconstructing actual

wind fields for each TC using meteorological models.

This is necessary because the spatial distribution of

high winds is variable between and within TCs of

comparable maximum intensity due to differences in

translation speed (Shapiro, 1983) and size of circulation

(Merrill, 1984) along the track. Further, though ocean-

ographers have long used SST data to detect and ana-

lyze wakes (e.g., Sriver & Huber, 2007; Lloyd &

Vecchi, 2011), this study is the first to examine SST

changes with the specific intent of identifying cooling

sufficiently intense and persistent to benefit reefs.

Using this methodology, we confirm that TC cooling

can significantly lower thermal stress across an entire

coral reef region (the Caribbean) and that this occurs

frequently enough near some reefs to provide intermit-

tent thermal refuge. The scope for such cooling to

occur at reefs in the Caribbean and beyond will likely

increase even if the frequency of TCs remains constant

or declines slightly as predicted by global models

(reviewed by Knutson et al., 2010). This will happen

because episodes of anomalously warm SST will

become more frequent and widespread (though at a

spatially variable rate) in the coming decades (Donner

et al., 2005; van Hooidonk et al., 2013), and TCs regu-

larly track near many of the world’s reefs, especially

when SST is at or near the seasonal maximum. Thus,

understanding how future risks to reefs will vary

requires further examination of the current prevalence

of beneficial TC cooling across the globe.

Widespread, persistent high thermal stress during an

active TC season provides the best chance for TC

cooling to lower thermal stress at reefs below the

(a) (b)

Fig. 5 The distribution of tropical cyclone (TC) tracks and their net effect in 2005 (a) and 2010 (b). Blue (red) tracks show positive (nega-

tive) net cooling effects, and black tracks show a neutral effect. TC track line thickness is proportional to storm intensity (dashed lines

depict winds <33 ms�1). Gray grid cells delineate 50-km areas that contain reefs across the Greater Caribbean (n = 477).



critical levels generally known to cause bleaching. The

annual probability of this was as high as 31% for some

Caribbean reefs. However, our data demonstrate that

even when cooling is widespread at times of high ther-

mal stress within a given reef region, it may not be high

enough or occur close enough to all reefs to provide

thermal refuge. In other words, one cannot assume that

all reefs located near cyclones at times of thermal stress

are protected – instead, explicit spatial modeling using

actual SST data is required to identify where protection

occurred as was done here. Nonetheless, predicted ris-

ing SST (IPCC, 2013) combined with continued high

levels of TC activity from a warm phase of the Atlantic

Multidecadal Oscillation (Goldenberg et al., 2001; Bell

& Chelliah, 2006) imply that chances for such relief will

persist or increase across the Caribbean in the coming

decades. This is less certain in other ocean basins where

either TC activity or thermal stress (or both) have gen-

erally been low (Carrigan & Puotinen, 2011). Support-

ing this are the uniformly low probabilities of TC

cooling and high thermal stress (<10%) found across

the Caribbean during the inactive TC period prior to

1995. It should be noted, however, that these probabili-

ties neglect to consider the full benefit of preventative

cooling from TC interruption of warming cycles (early

season cooling) and acceleration of the onset of seasonal

SST drops (late season cooling). The latter was evident

during the 2005 and 2010 Caribbean seasons by the

greater disparity between thermal stress levels at

cooled vs. noncooled reef pixels later in the season.

Developing a robust method for measuring this effect

will be particularly important for global regions with

less frequent immediate stress relief.

Clearly, any perceived benefit from TC cooling is

irrelevant for a coral community also destroyed by TC

wave action. Yet in 2005 and 2010, our results show that

a greater area of Caribbean reef was cooled than dam-

aged (22% vs. 4%). One reason for this is the relatively

stringent wind speed threshold indicated by field data

to define the wave damage zones for the Caribbean

(maximum winds >50 ms�1 - Gardner et al., 2005 vs.

33 ms�1 and 40 ms�1 for the Great Barrier Reef – Fabri-

cius et al., 2008). This suggests that coral vulnerability

to wave damage is relatively low in the Caribbean, per-

haps due to frequent TC activity favoring the survival

of wave resistant forms. Indeed, small dome-shaped

corals like Porites astreoides (Green et al., 2008) that tend

to be more wave resistant (Madin, 2005) are more abun-

dant across the region than the more vulnerable

branching Acroporid corals (Aronson & Precht, 2001).

Explicit spatial modeling of TC cool wakes and damage

zones in other regions is thus needed to deter-

mine whether the Caribbean case is typical. Ocean

acidification, on the other hand, is expected to increase

coral vulnerability to wave damage (Langdon & Atkin-

son, 2005; Silverman et al., 2009) by reducing calcifica-

tion rates (Madin & Connolly, 2006) and cementation

(Manzello et al., 2008). This may be exacerbated by high

variability in CO2 conditions at reefs as natural short

term diurnal variation in CO2 is superimposed on the

long term increased mean CO2 levels (Shaw et al.,

2013).

As coral bleaching becomes more frequent and wide-

spread, the need to identify areas of potential thermal

refuge for which to prioritize conservation efforts

becomes increasingly urgent. This study demonstrates

that TC-induced cooling can provide intermittent ther-

mal refuge at a broad spatial scale in the Caribbean.

Spatially explicit modeling of TC cooling and subse-

quent estimation of the likelihood it will coincide with

high thermal stress across the rest of the world’s reefs

is needed to assist in optimizing strategies that promote

the survival of coral reefs under global climate change.

Acknowledgements

This study was funded by the GeoQuEST Research Centre, theInstitute for Conservation Biology and Environmental Manage-ment and the Faculty of Science, all at the University of Wollon-gong. Thanks to Stuart Phinn and Tim McClanahan for usefulcomments on the thesis chapter (AC) on which the paper isbased. Thanks to Helen McGregor, Jeff Maynard, and two anon-ymous reviewers for critical feedback which improved thismanuscript. Technical support was provided by the SpatialAnalysis Laboratories at the University of Wollongong, and theSchool of Earth Sciences at The Ohio State University.

References

Anthony KRN, Kline DI, Diaz-Pulido G, Dove S, Hoegh-Guldberg O (2008) Ocean

acidification causes bleaching and productivity loss in coral reef builders. Proceed-

ings of the National Academy of Science, 105, 17442–17446.

Aronson RB, Precht WF (2001) White-band disease and the changing face of Carib-

bean coral reefs. Hydrobiologia, 460, 25–38.

Baird AH, Marshall PA (2002) Mortality, growth and reproduction in scleractinian

corals following bleaching on the great barrier reef. Marine Ecology Progress Series,

237, 133–141.

Baker AC, Glynn PW, Riegl B (2008) Climate change and coral reef bleaching: an eco-

logical assessment of long-term impacts, recovery trends and future outlook. Estu-

arine Coastal and Shelf Science, 80, 435–471.

Ban NC, Pressey RL, Weeks S (2012) Conservation objectives and sea-surface

temperature anomalies in the great barrier reef. Conservation Biology, 26, 799–

809.

Bayraktarov E, Pizarro V, Eidens C, Wilke T, Wild C (2012) Upwelling mitigates coral

bleaching in the Columbian Caribbean. Proceedings of the 12th International Coral

Reef Symposium. Cairns, Australia.

Bell GD, Chelliah M (2006) Leading tropical modes associated with interannual and

multidecadal fluctuations in north Atlantic hurricane activity. Journal of Climate,

19, 590–612.

Bellwood DR, Hughes TP, Folke C, Nystrom M (2004) Confronting the coral reef

crisis. Nature, 429, 827–833.

Bender MA, Ginis I, Kurihara Y (1993) Numerical simulations of tropical cyclone-

ocean interaction with a high-resolution coupled model. Journal of Geophysical

Research-Atmospheres, 98, 23245–23263.

Blunden J, Arndt DS, Baringer MO (2011) State of the climate in 2010. Bulletin of the

American Meteorological Society, 92, s1–s266.



Bruno JF, Selig ER, Casey KS et al. (2007) Thermal stress and coral cover as drivers of

coral disease outbreaks. PLoS Biology, 5, 1220–1227.

Carrigan AD, Puotinen ML (2011) Assessing the potential for tropical cyclone

induced sea surface cooling to reduce thermal stress on the world’s coral reefs.

Geophysical Research Letters, 38, L23604.

Chiang TL, Wu CR, Oey LY (2011) Typhoon Kai-Tak: an ocean’s perfect storm. Journal

of Physical Oceanography, 41, 221–233.

Dare RA, McBride JL (2011) Sea Surface Temperature Response to Tropical Cyclones.

Monthly Weather Review, 139, 3798–3808.

De’ath G, Fabricius KE, Sweatman H, Puotinen ML (2012) The 27-year decline of coral

cover on the great barrier reef and its causes. Proceedings of the National Academy of

Sciences of the United States of America, 109, 17995–17999.

Done TJ (1992) Effects of tropical cyclone waves on ecological and geomorphological

structures on the great barrier reef. Continental Shelf Research, 12, 859.

Donner SD, Skirving WJ, Little CM, Oppenheimer M, Hoegh-Guldberg O (2005) Glo-

bal assessment of coral bleaching and required rates of adaptation under climate

change. Global Change Biology, 11, 2251–2265.

Eakin CM, Morgan JA, Heron SF et al. (2010) Caribbean Corals in Crisis: record Ther-

mal Stress, Bleaching, and Mortality in 2005. PLoS ONE, 5, e13969.

Edwards HJ, Elliott IA, Eakin CM et al. (2011) How much time can herbivore protec-

tion buy for coral reefs under realistic regimes of hurricanes and coral bleaching?

Global Change Biology, 17, 2033–2048.

Fabricius KE, De’ath G, Puotinen ML, Done T, Cooper TF, Burgess SC (2008) Distur-

bance gradients on inshore and offshore coral reefs caused by a severe tropical

cyclone. Limnology and Oceanography, 53, 690–704.

Gardner TA, Cote IM, Gill JA, Grant A, Watkinson AR (2003) Long-term region-wide

declines in Caribbean corals. Science, 301, 958–960.

Gardner TA, Cote IM, Gill JA, Grant A, Watkinson AR (2005) Hurricanes and Carib-

bean coral reefs: impacts, recovery patterns, and role in long-term decline. Ecology,

86, 174–184.

Glynn PW (1996) Coral reef bleaching: facts, hypotheses and implications. Global

Change Biology, 2, 495–509.

Goldenberg SB, Landsea CW, Mestas-Nunez AM, GrayWM (2001) The recent increase

in Atlantic hurricane activity: causes and implications. Science, 293, 474–479.

Gordon AL (1967) Circulation of the Caribbean sea. Journal of Geophysical Research, 72,

6207.

Green DH, Edmunds PJ, Carpenter RC (2008) Increasing relative abundance of Por-

ites astreoides on Caribbean reefs mediated by an overall decline in coral cover.

Marine Ecology-Progress Series, 359, 1–10.

Grottoli AG, Rodrigues LJ, Palardy JE (2006) Heterotrophic plasticity and resilience in

bleached corals. Nature, 440, 1186–1189.

Hoegh-Guldberg O (1999) Climate change, coral bleaching and the future of the

world’s coral reefs. Marine And Freshwater Research, 50, 839–866.

Holland GJ, Belanger JI, Fritz A (2010) A revised model for radial profiles of hurri-

cane winds. Monthly Weather Review, 138, 4393–4401.

van Hooidonk R, Maynard JA, Planes S (2013) Temporary refugia for coral reefs in a

warming world. Nature Climate Change, 3, 508–511.

Huang PS, Sanford TB, Imberger J(2009) Heat and turbulent kinetic energy budgets

for surface layer cooling induced by the passage of Hurricane Frances (2004). Jour-

nal of Geophysical Research-Oceans, 114, C12023.

Hughes TP, Connell JH (1999) Multiple stressors on coral reefs: a long-term perspec-

tive. Limnology and Oceanography, 44, 932–940.

Hughes TP, Baird AH, Bellwood DR et al. (2003) Climate change, human impacts,

and the resilience of coral reefs. Science, 301, 929–933.

IPCC (2013) Summary for policymakers. In: Climate Change 2013: The Physical Science

Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovern-

mental Panel on Climate Change (eds Stocker TF, Qin D, Plattner G-K, Tignor M,

Allen SK, Boschung J, Nauels A, Xia Y, Bex V and Midgley PM). Cambridge Uni-

versity Press, Cambridge, UK/New York, NY, USA.

Jaimes B, Shay LK (2009) Mixed layer cooling in mesoscale oceanic eddies during

hurricanes Katrina and Rita. Monthly Weather Review, 137, 4188–4207.

Klotzbach PJ (2011) El Nino-Southern oscillation’s impact on Atlantic basin hurri-

canes and U.S Landfalls. Journal of Climate, 24, 1252–1263.

Knutson TR, McBride JL, Chan J et al. (2010) Tropical cyclones and climate change.

Nature Geoscience, 3, 157–163.

Langdon C, Atkinson MJ (2005) Effect of elevated pCO(2) on photosynthesis and cal-

cification of corals and interactions with seasonal change in temperature/irradi-

ance and nutrient enrichment. Journal of Geophysical Research-Oceans, 110, C09S07.

Liu G, Strong AE, Skirving WJ, Arzayus LF (2006) Overview of noaa coral reef watch pro-

gram’s near-real-time satellite global coral bleaching monitoring activities. Proceedings

of the 10th International Coral Reef Symposium. ICRS, Okinawa, Japan.

Lloyd ID, Vecchi GA (2011) Observational evidence for oceanic controls on hurricane

intensity. Journal of Climate, 24, 1138–1153.

Lough JM (2012) Small change, big difference: sea surface temperature distributions

for tropical coral reef ecosystems, 1950-2011. Journal of Geophysical Research C:

Oceans, 117, C09018.

Madin JS (2005) Mechanical limitations of reef corals during hydrodynamic distur-

bances. Coral Reefs, 24, 630–635.

Madin JS, Connolly SR (2006) Ecological consequences of major hydrodynamic

disturbances on coral reefs. Nature, 444, 477–480.

Maina J, Venus V, McClanahan TR, Ateweberhan M (2008) Modelling susceptibility

of coral reefs to environmental stress using remote sensing data and GIS models.

Ecological Modelling, 212, 180–199.

Maina J, McClanahan TR, Venus V, Ateweberhan M (2011) Global gradients of coral

exposure to environmental stresses and implications for local management. PLoS

ONE, 68, e23064.

Manzello DP, Brandt M, Smith TB, Lirman D, Hendee JC, Nemeth RS (2007) Hurri-

canes benefit bleached corals. Proceedings of the National Academy of Sciences of the

United States of America, 104, 12035–12039.

Manzello DP, Kleypas JA, Budd DA, Eakin CM, Glynn PW, Langdon C (2008) Poorly

cemented coral reefs of the eastern tropical Pacific: possible insights into reef

development in a high-CO(2) world. Proceedings of the National Academy of Sciences

of the United States of America, 105, 10450–10455.

McClanahan TR, Ateweberhan M, Muhando CA, Maina J, Mohammed MS (2007a)

Effects of climate and seawater temperature variation on coral bleaching and

mortality. Ecological Monographs, 77, 503–525.

McClanahan TR, Ateweberhan M, Sebastian CR, Graham NAJ, Wilson SK, Brugge-

mann JH, Guillaume MMM (2007b) Predictability of coral bleaching from synoptic

satellite and in situ temperature observations. Coral Reefs, 26, 695–701.

Mendes JM, Woodley JD (2002) Effect of the 1995-1996 bleaching event on polyp

tissue depth, growth, reproduction and skeletal band formation in Montastraea

annularis. Marine Ecology Progress Series, 235, 93–102.

Merrill RT (1984) A comparison of large and small tropical cyclones. Monthly Weather

Review, 112, 1408–1418.

del M�onaco C, Haiek G, Narciso S, Galindo M (2012) Massive bleaching of coral reefs

induced by the 2010 ENSO, Puerto Cabello, Venezuela. Revista de Biologia Tropical,

60, 527–538.

Oliver TA, Palumbi SR (2011) Do fluctuating temperature environments elevate coral

thermal tolerance? Coral Reefs, 30, 429–440.

Perry CT, Murphy GN, Kench PS, Smithers SG, Edinger EN, Steneck RS, Mumby PJ

(2013) Caribbean-wide decline in carbonate production threatens coral reef

growth. Nature communications, 4, 1402.

Price JF (1981) Upper ocean response to a hurricane. Journal of Physical Oceanography,

11, 153–175.

Puotinen ML (2004) Tropical cyclones in the great barrier reef region, 1910-1999: a

first step towards characterising the disturbance regime. Australian Geographical

Studies, 42, 378–392.

Puotinen ML (2007) Modelling the risk of cyclone wave damage to coral reefs using

GIS: a case study of the great barrier reef, 1969-2003. International Journal Of

Geographical Information Science, 21, 97–120.

Riegl B, Piller WE (2003) Possible refugia for reefs in times of environmental stress.

International Journal of Earth Sciences, 92, 520–531.

Shapiro LJ (1983) The asymmetric boundary layer flow under a translating hurricane.

Journal of Atmospheric Science, 40, 1984–1998.

Shaw EC, McNeil BI, Tilbrook B, Matear R, Bates ML (2013) Anthropogenic changes

to seawater buffer capacity combine with natural reef metabolism induce extreme

future coral reef CO2 conditions. Global Change Biology, 19, 1632–1641.

Shein KA, Waple AM, Menne MJ et al. (2006) State of the climate in 2005. Bulletin of

the American Meteorological Society, 87, S6–S102.

Silverman J, Lazar B, Cao L, Caldeira K, Erez J (2009) Coral reefs may start

dissolving when atmospheric CO(2) doubles. Geophysical Research Letters, 36,

L05606.

Skirving WJ, Heron M, Heron S (2006) The hydrodynamics of a bleaching event:

implications for management and monitoring. In: Coral Reefs and Climate Change –

Science and Management (ed. Phinney JT), pp. 145–161. American Geophysical

Union, Washington DC.

Sriver RL, Huber M (2007) Observational evidence for an ocean heat pump induced

by tropical cyclones. Nature, 447, 577–580.

Stramma L, Cornillon P, Price JF (1986) Satellite observations of sea surface cooling

by hurricanes. Journal of Geophysical Research-Oceans, 91, 5031–5035.

Tartaglione CA, Smith SR, O’Brien JJ (2003) ENSO impact on hurricane landfall prob-

abilities for the Caribbean. Journal of Climate, 16, 2925–2931.



Treml E, ColganM, KeevicanM (1997)Hurricane Disturbance and Coral Reef Development:

A Geographic Information System (Gis) Analysis of 501 Years of Hurricane Data from the

Lesser Antilles. Proceedings of the 8th International Coral Reef Symposium. Smithso-

nianTropical Research Institute, Balboa, Panama.

Vincent EM, Lengaigne M, Madec G et al. (2012) Processes setting the characteristics

of sea surface cooling induced by tropical cyclones. Journal of Geophysical Research-

Oceans, 117, C02020.

West JM, Salm RV (2003) Resistance and resilience to coral bleaching: implications for

coral reef conservation and management. Conservation Biology, 17, 956–967.

Wilkinson C (2008) Status of coral reefs of the world: 2008. Global Coral Reefs

Monitoring Network and Reef and Rainforest Research Centre, Townsville,

Australia.

Wilkinson C, Souter D, 2008 Status of Caribbean Coral Reefs After Bleaching and Hurri-

canes in 2005. Global Coral Reef Monitoring Network, and Reef and Rainforest

Research Centre, Townsville.

Woodley JD (1992) The incidence of hurricanes on the north coast of Jamaica since

1870: are the classic reef descriptions atypical? Hydrobiologia, 247, 133–138.



Documents

Tropical cyclone cooling combats region-wide coral bleaching