Editing IPCC:AR6/SROCC/Chapter-6 (section)

== 6.5 Extreme ENSO Events and Other Modes of Interannual Climate Variability ==

<span id="key-processes-and-feedbacks-observations-detection-and-attribution-projections"></span>
=== 6.5.1 Key Processes and Feedbacks, Observations, Detection and Attribution, Projections ===

<div id="section-6-5-1-1extreme-el-nino-la-nina"></div>

<span id="extreme-el-niño-la-niña"></span>
==== 6.5.1.1 Extreme El Niño, La Niña ====

<div id="section-6-5-1-1extreme-el-nino-la-nina-block-1"></div>

AR5 (Christensen et al., 2013 <sup>[[#fn:r481|481]]</sup> ) and SREX do not provide a definition for an extreme El Niño but mention such events, especially in the context of the 1997–1998 El Niño and its impacts. AR5 and SREX concluded that confidence in any specific change in ENSO variability in the 21st century is low. However, they did note that due to increased moisture availability, precipitation variability associated with ENSO is likely to intensify. Since AR5 and SREX, there is now a limited body of literature that examines the impact of climate change on ENSO over the historical period.

Palaeo-ENSO studies suggest that ENSO was highly variable throughout the Holocene, with no evidence for a systematic trend in ENSO variance (Cobb et al., 2013 <sup>[[#fn:r482|482]]</sup> ) but with some indication that the ENSO variance over 1979–2009 has been much larger than that over 1590–1880 (McGregor et al., 2013 <sup>[[#fn:r483|483]]</sup> ). Palaeo-ENSO reconstruction for the past eight centuries suggests that central Pacific ENSO activity has increased between the last two decades (1980-2015; Liu et al., 2017b <sup>[[#fn:r484|484]]</sup> ), with an increasing number of central Pacific El Niño events compared to east Pacific El Niño events (Freund et al., 2019 <sup>[[#fn:r485|485]]</sup> ). Further proxy evidence exists for changes in the mean state of the equatorial Pacific in the last 2000 years (Rustic et al., 2015 <sup>[[#fn:r486|486]]</sup> ; Henke et al., 2017 <sup>[[#fn:r487|487]]</sup> ). Simulations using an Earth System Model indicate significantly higher ENSO variance during 1645–1715 than during the 21st century warm period, though it is unclear whether these simulated changes are realistic (Keller et al., 2015 <sup>[[#fn:r488|488]]</sup> ). For the 20th century, the frequency and intensity of El Niño events were high during 1951–2000, in comparison with the 1901–1950 period (Lee and McPhaden, 2010 <sup>[[#fn:r489|489]]</sup> ; Kim et al., 2014b <sup>[[#fn:r490|490]]</sup> ; Roxy et al., 2014 <sup>[[#fn:r491|491]]</sup> ). Current instrumental observational records are not long enough and the quality of data before 1950 is limited, to assert these changes with ''high confidence'' (Wittenberg, 2009 <sup>[[#fn:r492|492]]</sup> ; Stevenson et al., 2010 <sup>[[#fn:r493|493]]</sup> ) though the palaeo records mentioned here signal the emergence of a statistically significant increase in ENSO variance in recent decades.

Since SREX and AR5, an extreme El Niño event occurred in 2015–2016. This has resulted in significant new literature regarding physical processes and impacts but there are no firm conclusions regarding the impact of climate change on the event. The SST anomaly peaked toward the central equatorial Pacific causing floods in many regions of the world such as those in the west coasts of the USA and other parts of North America, some parts of South America close to Argentina and Uruguay, the UK and China (Ward et al., 2014 <sup>[[#fn:r494|494]]</sup> ; Ward et al., 2016 <sup>[[#fn:r495|495]]</sup> ; Zhai et al., 2016 <sup>[[#fn:r496|496]]</sup> ; Scaife et al., 2017 <sup>[[#fn:r497|497]]</sup> ; Whan and Zwiers, 2017 <sup>[[#fn:r498|498]]</sup> ; Sun and Miao, 2018 <sup>[[#fn:r499|499]]</sup> ; Yuan et al., 2018 <sup>[[#fn:r500|500]]</sup> ).

The main new body of literature concerns future projections of the frequency of occurrence and variability of extreme ENSO events with improved confidence (Cai et al., 2014a <sup>[[#fn:r501|501]]</sup> ; Cai et al., 2018 <sup>[[#fn:r502|502]]</sup> ). These studies define extreme El Niño events as those El Niño events which are characterised by a pronounced eastward extension of the west Pacific warm pool and development of atmospheric convection, and hence a rainfall increase of greater than 5 mm day -1 during December to February (above 90th percentile), in the usually cold and dry equatorial eastern Pacific (Niño 3 region, 150°W–90°W, 5°S–5°N; Cai et al., 2014a <sup>[[#fn:r503|503]]</sup> ), such as the 1982–1983, 1997–1998 and 2015–2016 El Niños (Santoso et al., 2017 <sup>[[#fn:r504|504]]</sup> ; Figure 6.5).

The background long-term warming puts the 2015–2016 El Niño among the three warmest in the instrumental records (24 El Niño events occurred during 1900–2018; Huang et al., 2016 <sup>[[#fn:r505|505]]</sup> ; Santoso et al., 2017 <sup>[[#fn:r506|506]]</sup> ). The 2015–2016 event can be viewed as the first emergence of an extreme El Niño in the 21st century – one which satisfies the rainfall threshold definition, but not characterised by the eastward extension of the west Pacific warm pool (L’Heureux et al., 2017 <sup>[[#fn:r507|507]]</sup> ; Santoso et al., 2017 <sup>[[#fn:r508|508]]</sup> ).

Based on the precipitation threshold, extreme El Niño frequency is projected to increase with the global mean temperatures ( ''medium confidence'' ) with a doubling in the 21st century under 1.5°C of global warming, from about one event every 20 years during 1891–1990, to one every 10 years (Cai et al., 2014a <sup>[[#fn:r509|509]]</sup> ; Figure 6.5). The increase in frequency continues for up to a century even after global mean temperature has stabilised at 1.5°C, thereby challenging the limits to adaptation, and hence indicates high risk even at the 1.5°C threshold (Wang et al., 2017 <sup>[[#fn:r510|510]]</sup> ; Hoegh-Guldberg et al., 2018 <sup>[[#fn:r511|511]]</sup> ). Meanwhile, the La Niña events also tend to increase in frequency and double under RCP8.5 (Cai et al., 2015 <sup>[[#fn:r512|512]]</sup> ), but indicate no further significant changes after global mean temperatures have stabilised (Wang et al., 2017 <sup>[[#fn:r513|513]]</sup> ). Particularly concerning is that swings from extreme El Niño to extreme La Niña (opposite of extreme El Niño) have been projected to occur more frequently under greenhouse warming (Cai et al., 2015 <sup>[[#fn:r514|514]]</sup> ). The increasing ratio of Central Pacific El Niño events to East Pacific El Niño events is projected to continue, under increasing emissions (Freund et al., 2019 <sup>[[#fn:r515|515]]</sup> ). Further, CMIP5 models indicate that the risk of major rainfall disruptions has already increased for countries where the rainfall variability is linked to ENSO variability. This risk will remain elevated for the entire 21st century, even if substantial reductions in global GHG emissions are made ( ''medium confidence'' ). The increase in disruption risk is caused by anthropogenic warming that drives an increase in the frequency and magnitude of ENSO events and also by changes in background SST patterns (Power et al., 2013 <sup>[[#fn:r516|516]]</sup> ; Chung et al., 2014 <sup>[[#fn:r517|517]]</sup> ; Huang and Xie, 2015 <sup>[[#fn:r518|518]]</sup> ). While many of these studies have adopted the precipitation view of an extreme El Nino, studies also indicate an increase in SST variability for events with their main SST anomalies in the east Pacific (Cai et al., 2018 <sup>[[#fn:r519|519]]</sup> ). Also, a role of cross-equatorial winds has been identified (Hu and Fedorov, 2018 <sup>[[#fn:r520|520]]</sup> )

<div id="section-6-5-1-1extreme-el-nino-la-nina-block-2"></div>

<span id="figure-6.5"></span>
<!-- START IMG -->
<!-- IMG TITLE -->
'''Figure 6.5'''

<span id="figure-6.5-frequency-of-extreme-el-nino-southern-oscillation-enso-events-adapted-from-cai-et-al.-2014a.-a-december-to-february-mean-meridional-sea-surface-temperature-sst-gradient-x-axis-5on10on-210oe270oe-minus-2.5os2.5on-210oe270oe-versus-equatorial-pacific-anomalous-rainfall-y-axis-5os5on-210oe270oe.-data-from-only-those-coupled-model-intercomparison-project-phase-5-cmip5-models"></span>
<!-- IMG CAPTION -->
'''Figure 6.5 | Frequency of extreme El Niño Southern Oscillation (ENSO) events, adapted from Cai et al. (2014a). (a) December to February mean meridional sea surface temperature (SST) gradient (x-axis: 5oN–10oN, 210oE–270oE minus 2.5oS–2.5oN, 210oE–270oE) versus equatorial Pacific anomalous rainfall (y-axis: 5oS–5oN, 210oE–270oE). Data from only those Coupled Model Intercomparison Project Phase 5 (CMIP5) models […]'''
<!-- IMG FILE -->
[[File:74279e093d8124aa346fc5422faae8c0 IPCC-SROCC-CH_6_5.jpg]]

Figure 6.5 | Frequency of extreme El Niño Southern Oscillation (ENSO) events, adapted from Cai et al. (2014a). (a) December to February mean meridional sea surface temperature (SST) gradient (x-axis: 5oN–10oN, 210oE–270oE minus 2.5oS–2.5oN, 210oE–270oE) versus equatorial Pacific anomalous rainfall (y-axis: 5oS–5oN, 210oE–270oE). Data from only those Coupled Model Intercomparison Project Phase 5 (CMIP5) models that capture the observed relationship between Pacific SST and rainfall are shown. Black dots are from observations with extreme El Niño and extreme La Niña years indicated. The horizontal line denotes the threshold of 5 mm day–1 for an extreme event. (b)Histogram showing the relative frequency of rainfall rates. The vertical line denotes the 5 mm day–1 threshold. Higher counts of extreme events under the Representative Concentration Pathway (RCP)8.5 scenario suggest an increase in the frequency of extreme El Niño under global warming.

<!-- END IMG -->
<div id="section-6-5-1-2indian-ocean-basin-wide-warming-and-changes-in-indian-ocean-dipole-iod-events"></div>

<span id="indian-ocean-basin-wide-warming-and-changes-in-indian-ocean-dipole-iod-events"></span>
==== 6.5.1.2 Indian Ocean Basin-wide Warming and Changes in Indian Ocean Dipole (IOD) Events ====

<div id="section-6-5-1-2indian-ocean-basin-wide-warming-and-changes-in-indian-ocean-dipole-iod-events-block-1"></div>

The Indian Ocean has experienced consistent warming from the surface to 2,000 m during 1960–2015, with most of the warming occurring in the upper 300 m (Cheng et al., 2015 <sup>[[#fn:r521|521]]</sup> ; Nieves et al., 2015 <sup>[[#fn:r522|522]]</sup> ; Cheng et al., 2017 <sup>[[#fn:r523|523]]</sup> ; Gnanaseelan et al., 2017 <sup>[[#fn:r524|524]]</sup> ). New historical ocean heat content (OHC) estimates show an abrupt increase in the Indian Ocean upper 700 m OHC after 1998, contributing to more than 28% of the global ocean heat gain, despite representing only about 12% of the global ocean area (Cheng et al., 2017 <sup>[[#fn:r525|525]]</sup> ; Makarim et al., 2019 <sup>[[#fn:r526|526]]</sup> ). The tropical Indian Ocean SST has warmed by 1.04°C during 1950–2015, while the tropical SST warming is 0.83°C and the global SST warning is 0.65°C. More than 90% of the surface warming in the Indian Ocean has been attributed to changes in GHG emissions (Dong et al., 2014 <sup>[[#fn:r527|527]]</sup> ), with the heat redistributed in the basin via local ocean and atmospheric dynamics (Liu et al., 2015b <sup>[[#fn:r528|528]]</sup> ), the ITF (Section 6.6.1; Susanto et al., 2012 <sup>[[#fn:r529|529]]</sup> ; Sprintall and Revelard, 2014 <sup>[[#fn:r530|530]]</sup> ; Lee et al., 2015b <sup>[[#fn:r531|531]]</sup> ; Susanto and Song, 2015 <sup>[[#fn:r532|532]]</sup> ; Zhang et al., 2018 <sup>[[#fn:r533|533]]</sup> ) and the Walker circulation (Roxy et al., 2014 <sup>[[#fn:r534|534]]</sup> ; Abish et al., 2018 <sup>[[#fn:r535|535]]</sup> ).

The dynamic processes related to the projected changes in IOD under global warming have a large inter-model spread (Cai et al., 2013 <sup>[[#fn:r536|536]]</sup> ). The frequency of extreme positive IOD events are projected to increase by almost a factor of three, from a one-in-seventeen-year event in the 20th century to a one-in-six-year event in the 21st century ( ''low confidence'' ). The bias in the CMIP5 models and internal variability could enlarge the projected increase in the extreme positive IOD events (Li et al., 2016a <sup>[[#fn:r537|537]]</sup> ; Hui and Zheng, 2018 <sup>[[#fn:r538|538]]</sup> ). The increase in IOD events is not linked to the change in the frequency of El Niño events but instead to mean state change—with weakening of both equatorial westerly winds and eastward oceanic currents in association with a faster warming in the western than the eastern equatorial Indian Ocean (Cai et al., 2014b <sup>[[#fn:r539|539]]</sup> ). A combination of extreme ENSO and IOD events has led to a northward shift in the Intertropical Convergence Zone (ITCZ) during 1979–2015, which is expected to increase further in the future (Freitas et al., 2017 <sup>[[#fn:r540|540]]</sup> ).

<span id="impacts-on-human-and-natural-systems"></span>
=== 6.5.2 Impacts on Human and Natural Systems ===

<div id="section-6-5-2impacts-on-human-and-natural-systems-block-1"></div>

Increasing frequency of extreme ENSO and IOD events have the potential to have widespread impacts on natural and human systems in many parts of the globe. Though the occurrence of the extreme 2015–2016 El Niño has produced a large body of literature, it is still not clear how climate change may have altered such an impact, nor how such impacts might change in the future with increasing frequency of extreme ENSO events. We highlight here some studies that have attempted to assess the joint impact of mean change and variability. In addition to observed high variability of rainfall, severe weather events and impacts on TCs activity (Yonekura and Hall, 2014 <sup>[[#fn:r541|541]]</sup> ; Zhang and Guan, 2014 <sup>[[#fn:r542|542]]</sup> ; Wang and Liu, 2016 <sup>[[#fn:r543|543]]</sup> ; Zhan, 2017 <sup>[[#fn:r544|544]]</sup> ), extreme El Nino events have substantial impacts on natural systems which include those on marine ecosystems (Sanseverino et al., 2016 <sup>[[#fn:r545|545]]</sup> ; Mogollon and Calil, 2017 <sup>[[#fn:r546|546]]</sup> ; Ohman et al., 2017 <sup>[[#fn:r537|537]]</sup> ), such as severe and repeated bleaching of corals (Hughes et al., 2017a <sup>[[#fn:r548|548]]</sup> ; Hughes et al., 2017b <sup>[[#fn:r549|549]]</sup> ; Eakin et al., 2018 <sup>[[#fn:r550|550]]</sup> ), and glacial growth and retreat (Thompson et al., 2017 <sup>[[#fn:r551|551]]</sup> ). On the other hand, impacts on human, including managed systems are: increased incidences of forest fires (Christidis et al., 2018b <sup>[[#fn:r552|552]]</sup> ; Tett et al., 2018 <sup>[[#fn:r553|553]]</sup> ), degraded air quality (Koplitz et al., 2015 <sup>[[#fn:r554|554]]</sup> ; Chang et al., 2016 <sup>[[#fn:r555|555]]</sup> ; Zhai et al., 2016 <sup>[[#fn:r556|556]]</sup> ) such as the dense haze over most parts of Indonesia and the neighbouring countries in Southeast Asia as a result of prolonged Indonesian wildfires, thus imposing adverse impacts on public health in the affected areas (Koplitz et al., 2015 <sup>[[#fn:r557|557]]</sup> ; WMO, 2016), decreased agricultural yields in many parts of the globe (e.g., in most of the Pacific Islands countries, Thailand, eastern and southern Africa and others which had resulted food insecurity, particularly in eastern and southern Africa (UNSCAP, 2015; WMO, 2016; Christidis et al., 2018b <sup>[[#fn:r558|558]]</sup> ; Funk et al., 2018 <sup>[[#fn:r559|559]]</sup> ), and regional uptick in the number of reported cases of plague and hantavirus in Colorado and New Mexico, cholera in Tanzania, dengue in Brazil and Southeast Asia (Anyamba et al., 2019 <sup>[[#fn:r560|560]]</sup> ) and Zika virus in South America (Caminade et al., 2017 <sup>[[#fn:r561|561]]</sup> ), including increases in heat stroke cases (Christidis et al., 2018b <sup>[[#fn:r562|562]]</sup> ). Substantial economic losses had resulted from droughts and floods across various parts of the globe due to teleconnections. For instance, direct losses of 10 billion USD (Sun and Miao, 2018 <sup>[[#fn:r563|563]]</sup> ; Yuan et al., 2018 <sup>[[#fn:r564|564]]</sup> ) and 6.5 billion USD (Christidis et al., 2018b <sup>[[#fn:r565|565]]</sup> ) were estimated to have been incurred from severe urban inundation in cities along the Yangtze River in China and the extreme drought in Thailand, respectively.

ENSO events affect TCs activity through variations in the low-level wind anomalies, vertical wind shear, mid-level relative humidity, steering flow, the monsoon trough and the western Pacific subtropical high in Asia (Yonekura and Hall, 2014 <sup>[[#fn:r566|566]]</sup> ; Zhang and Guan, 2014 <sup>[[#fn:r567|567]]</sup> ). The subsurface heat discharge due to El Niño can intensify TCs in the eastern Pacific (Jin et al., 2014 <sup>[[#fn:r568|568]]</sup> ; Moon et al., 2015b <sup>[[#fn:r569|569]]</sup> ). TCs are projected to become more frequent (~20–40%) during future-climate El Niño events compared with present climate El Niño events ( ''medium confidence'' ), and less frequent during future-climate La Niña events, around a group of small island nations (for example, Fiji, Vanuatu, Marshall Islands and Hawaii) in the Pacific (Chand et al., 2017 <sup>[[#fn:r570|570]]</sup> ). The Indian Ocean basin-wide warming has led to an increase in TC heat potential in the Indian Ocean over the last 30 years, however the link to the changes in the frequency of TCs is not robust (Rajeevan et al., 2013 <sup>[[#fn:r571|571]]</sup> ).

During the early stages of an extreme El Niño event (2015–2016 El Niño), there is an initial decrease in atmospheric CO 2 concentrations over the tropical Pacific Ocean, due to suppression of equatorial upwelling, reducing the supply of CO 2 to the surface (Chatterjee et al., 2017 <sup>[[#fn:r572|572]]</sup> ), followed by a rise in atmospheric CO 2 concentrations due reduced terrestrial CO 2 uptake and increased fire emissions (Bastos et al., 2018 <sup>[[#fn:r573|573]]</sup> ). It is not clear how a future increase in the frequency extreme events would modulate the carbon cycle on longer decadal time scales.

Studies on projections of changes in ENSO impacts or teleconnections are rather limited. Nevertheless, Power and Delage (2018) provide a multi-model assessment of CMIP5 models and their simulated changes in the precipitation response to El Niño in the future (Figure 6.6). They identify different combinations of changes that might further impact natural and human systems. El Niño causes either positive or negative precipitation anomalies in diverse regions of the globe. Dry El Niño teleconnection anomalies may be further strengthened by, either mean climate drying in the region (Amazon, Central America and Australia in June to August (JJA)), or a strengthening of the El Niño dry teleconnection, or both. Conversely, wet El Niño teleconnections can be further strengthened by either increases in mean precipitation (East Africa and southeastern South America in December to February (DJF)) or a strengthening of the El Niño wet teleconnection (southeastern South America in JJA), or both (Tibetan Plateau, DJF). However, a present day dry El Niño response may be dampened by a wet mean response (South, East and Southeast Asia in JJA) or a wet present day El Niño response may be weakened by a dry mean change (Southern Europe/Mediterranean and West Coast South America in JJA). Finally, changes in the mean and El Niño response may be in the opposite direction (Southeast Asia, JJA and Central North America, DJF). Such changes could have an impact on phenomena such as wildfires (Fasullo et al., 2018 <sup>[[#fn:r575|575]]</sup> ). However, in many other regions that are currently impacted by El Niño, e.g., regions of South America, studies have found no significant changes in the ENSO-precipitation relationship (Tedeschi and Collins, 2017 <sup>[[#fn:r576|576]]</sup> ) and agreement between models for many regions suggests ''low confidence'' in projections of teleconnection changes (Yeh et al., 2018 <sup>[[#fn:r577|577]]</sup> ).

Along with extreme El Niño events, abrupt warming in the Indian Ocean and extreme IOD events have largely altered the Asian and African monsoon, impacting the food and water security over these regions. As a response to rising global SSTs and partially due to extreme El Niño events, the NH summer monsoon showed substantial intensification during 1979–2011, with an increase in rainfall by 9.5% per degree Celsius of global warming (Wang et al., 2013 <sup>[[#fn:r578|578]]</sup> ). However, the Indian summer monsoon circulation and rainfall exhibits a statistically significant weakening since the 1950s. This weakening has been hypothesised to be a response to the Indian Ocean basin-wide warming (Mishra et al., 2012 <sup>[[#fn:r579|579]]</sup> ; Roxy et al., 2015 <sup>[[#fn:r580|580]]</sup> ) and also to increased aerosol emissions (Guo et al., 2016 <sup>[[#fn:r581|581]]</sup> ) and changes in land use (Paul et al., 2016 <sup>[[#fn:r582|582]]</sup> ). Warming in the north Indian Ocean has resulted in increasing fluctuations in the southwest monsoon winds and a three-fold increase in extreme rainfall events across central India (Roxy et al., 2017 <sup>[[#fn:r583|583]]</sup> ). The frequency and duration of heatwaves have increased over the Indian subcontinent, and these events are associated with the Indian Ocean basin-wide warming and frequent El Niños (Rohini et al., 2016 <sup>[[#fn:r584|584]]</sup> ). In April 2016, as a response to the extreme El Niño, Southeast Asia experienced surface air temperatures that surpassed national records, increased energy consumption, disrupted agriculture and resulted in severe human discomfort (Thirumalai et al., 2017 <sup>[[#fn:r585|585]]</sup> ). A strong negative IOD event in 2016 led to large climate impact on East African rainfall, with some regions recording below 50% of normal rainfall, leading to devastating drought, food insecurity and unsafe drinking water for over 15 million people in Somalia, Ethiopia and Kenya.

<div id="section-6-5-2impacts-on-human-and-natural-systems-block-2"></div>

<span id="figure-6.6"></span>
<!-- START IMG -->
<!-- IMG TITLE -->
'''Figure 6.6'''

<span id="figure-6.6-schematic-figure-indicating-future-changes-in-el-nino-teleconnections-based-on-the-study-of-power-and-delage-2018.-the-background-pattern-of-sea-surface-temperature-sst-anomalies-oc-are-averaged-from-june-2015-to-august-2015-panel-a-and-december-2015-to-february-2016-panel-b-during-the-most-recent-extreme-el"></span>
<!-- IMG CAPTION -->
'''Figure 6.6 | Schematic figure indicating future changes in El Niño teleconnections based on the study of Power and Delage (2018). The background pattern of sea surface temperature (SST) anomalies (oC) are averaged from June 2015 to August 2015 (panel a) and December 2015 to February 2016 (panel b), during the most recent extreme El […]'''
<!-- IMG FILE -->
[[File:f08939c94162605d6103183bb40c3ccd IPCC-SROCC-CH_6_6.jpg]]

Figure 6.6 | Schematic figure indicating future changes in El Niño teleconnections based on the study of Power and Delage (2018) <sup>[[#fn:r589|589]]</sup> . The background pattern of sea surface temperature (SST) anomalies (oC) are averaged from June 2015 to August 2015 (panel a) and December 2015 to February 2016 (panel b), during the most recent extreme El Niño event (anomalies computed with respect to 1986–2005). Symbols indicate present day teleconnections for El Niño events. Black arrows indicate if there is a model consensus on change in mean rainfall in the region. Red arrows indicate if there is a model consensus on change in the rainfall anomaly under a future El Niño event. Direction of the arrow indicates whether the response in precipitation is increasing (up) or decreasing (down). Significance is determined when two-thirds or more of the models agree on the sign.

<!-- END IMG -->
<span id="risk-management-and-adaptation-1"></span>
=== 6.5.3 Risk Management and Adaptation ===

<div id="section-6-5-3risk-management-and-adaptation-block-1"></div>

Risk management of ENSO events has focussed on two main aspects: better prediction and early warning systems, and better mechanisms for reducing risks to agriculture, infrastructure, fisheries and aquaculture, wildfire and flood management. Extreme ENSO events are rare, with three such events since 1950 and they are difficult to predict due to the different drivers influencing them (Puy et al., 2017 <sup>[[#fn:r586|586]]</sup> ). The impacts of ENSO events also vary between events and between the different regions affected (Murphy et al., 2014 <sup>[[#fn:r587|587]]</sup> ; Fasullo et al., 2018 <sup>[[#fn:r588|588]]</sup> ; Power and Delage, 2018 <sup>[[#fn:r589|589]]</sup> ) however, there is limited literature on the change in the impacts of extreme ENSO compared to other ENSO events. In addition, there are also no specific risk management and adaptation strategies for human and natural systems for more extreme events other than what is in place for ENSO events (see also Chapter 4, Section 4.4 for the response to sea level change, an observed impact of ENSO). A first step in risk management and adaptation is thus to better understand the impacts these events have and to identify conditions that herald such extreme events that could be used to better predict extreme ENSO events.

Monitoring and forecasting are the most developed ways to manage extreme ENSOs. Several systems are already in place for monitoring and predicting seasonal climate variability and ENSO occurrence. However, the sustainability of the observing system is challenging and currently the Tropical Pacific Observing System 2020 (TPOS 2020) has the task of redesigning such a system, with ENSO prediction as one of its main objectives. These systems could be further elaborated to include extreme ENSO events. Westerly wind events in the Western Tropical Pacific, (Lengaigne et al., 2004 <sup>[[#fn:r590|590]]</sup> ; Chen et al., 2015a <sup>[[#fn:r591|591]]</sup> ; Fedorov et al., 2015 <sup>[[#fn:r592|592]]</sup> ) strong easterly wind events in the tropical Pacific (Hu and Fedorov, 2016 <sup>[[#fn:r593|593]]</sup> ; Puy et al., 2017 <sup>[[#fn:r594|594]]</sup> ), nonlinear interaction between air-sea fluxes and atmospheric deep convection (Bellenger et al., 2014 <sup>[[#fn:r595|595]]</sup> ; Takahashi and Dewitte, 2016 <sup>[[#fn:r596|596]]</sup> ) and advection of mean temperature by anomalous eastward zonal currents (Kim and Cai, 2014 <sup>[[#fn:r597|597]]</sup> ) are some of the factors that play an important role in the evolution of extreme ENSO events, which can be considered while improving the monitoring and forecasting system.

Despite the specificity of each extreme El Niño event, their forecasting is expected to improve through monitoring of recently identified precursory signals that peak in a window of two years before the event (Varotsos et al., 2016 <sup>[[#fn:r598|598]]</sup> ). An early warning system for coral bleaching associated, among other stressors, with extreme ENSO heat stress is provided by the NOAA Coral Reef Watch service with a 5 km resolution (Liu et al., 2018 <sup>[[#fn:r599|599]]</sup> ). The impacts of ENSO-associated extreme heat stress are heterogeneous, indicating the influence of other factors either biotic such as coral species composition, local adaptation by coral taxa reef depth or abiotic such as local upwelling or thermal anomalies (Claar et al., 2018 <sup>[[#fn:r600|600]]</sup> ). When identified and quantified, these factors can be used for risk analysis and risk management for these ecosystems.

In principle, it is easier to transfer the financial risk associated with extreme ENSO events through, for example, insurance products or other risk transfer instruments such as Catastrophe Bonds, than for more moderate events. An accurate prediction system is not required, but the measurement of these events, and quantification of likely impacts is required. As in other types of insurance systems, this can be done through, for example, calculations of average annual losses associated with extreme ENSO, and the design of appropriate financial instruments. Examples of research that can support the design of risk transfer instruments include Anderson et al. (2018) and Gelcer et al. (2018) for specific crops yields, and Aguilera et al. (2018) and Broad et al. (2002) for specific fisheries. Several risk transfer instruments have been implemented to deal with ENSO impacts, including parametric insurance based on SSTs for heavy rainfall damages, and another scheme for agricultural damages, both in Peru. Other examples include forecast-based financial aid (Red Cross Climate Centre, 2016). More broadly, other forms of risk management and governance can be designed with better information about the likely impacts of extreme ENSO events (e.g., Vignola et al., 2018).

<span id="inter-ocean-exchanges-and-global-change"></span>