Editing IPCC:AR6/SROCC/Chapter-1 (section)

== 1.3 Time Scales, Thresholds and Detection of Ocean and Cryosphere Change ==

<div id="article-1-3time-scales-thresholds-and-detection-of-ocean-and-cryosphere-change-block-1"></div>

It takes hundreds of years to millennia for the entire deep ocean to turn over (Matsumoto, 2007 <sup>[[#fn:r57|57]]</sup> ; Gebbie and Huybers, 2012 <sup>[[#fn:r58|58]]</sup> ), while renewal of the large ice sheets requires many thousands of years (Huybrechts and de Wolde, 1999 <sup>[[#fn:r59|59]]</sup> ). Long response times mean that the deep ocean and the large ice sheets tend to lag behind in their response to the rapidly changing climate at Earth’s surface, and that they will continue to change even after radiative forcing stabilises (e.g., Golledge et al., 2015 <sup>[[#fn:r60|60]]</sup> ; Figure 1.1a). Such ‘committed’ changes mean that some ocean and cryosphere changes are essentially irreversible on time scales relevant to human societies (decades to centuries), even in the presence of immediate action to limit further global warming (e.g., Section 4.2.3.5).

While some aspects of the ocean and cryosphere might respond in a linear (i.e., directly proportional) manner to a perturbation by some external forcing, this may change fundamentally when critical thresholds are reached. A very important example for such a threshold is the transition from frozen water to liquid water at around 0°C that can lead to rapid acceleration of ice-melt or permafrost thaw (e.g., Abram et al., 2013; Trusel et al., 2018 <sup>[[#fn:r61|61]]</sup> ). Such thresholds often act as tipping points, as they are associated with rapid and abrupt changes even when the underlying forcing changes gradually (Figure 1.1a, 1.1c). Tipping elements include, for example, the collapse of the ocean’s large-scale overturning circulation in the Atlantic (Section 6.7), or the collapse of the West Antarctic Ice Sheet though a process called marine ice sheet instability (Cross-Chapter Box 8 in Chapter 3; Lenton et al., 2008 <sup>[[#fn:r62|62]]</sup> ). Potential ocean and cryosphere tipping elements form part of the scientific case for efforts to limit climate warming to well below 2 o C (IPCC, 2018 <sup>[[#fn:r63|63]]</sup> ).

Anthropogenically forced change occurs against a backdrop of substantial natural variability (Figure 1.1b). The anthropogenic signal is already detectable in global surface air temperature and several other climate variables, including ocean temperature and salinity (IPCC, 2014 <sup>[[#fn:r64|64]]</sup> ), but short observational records and large year-to-year variability mean that formal detection is not yet the case for many expected ocean and cryosphere changes (Jones et al., 2016 <sup>[[#fn:r65|65]]</sup> ). ‘Time of Emergence’ refers to the time when anthropogenic change signals emerge from the background noise of natural variability in a pre-defined reference period Hawkins and Sutton, 2012; (Figure 1.1b; Section 5.2, Box 5.1). For some variables, (e.g., for those associated with ocean acidification), the current signals emerge from this natural variability within a few decades, whereas for others, such as primary production and expected Antarctic-wide sea ice decline, the signal may not emerge for many more decades even under high emission scenarios (Collins et al., 2013 <sup>[[#fn:r66|66]]</sup> ; Keller et al., 2014 <sup>[[#fn:r67|67]]</sup> ; Rodgers et al., 2015 <sup>[[#fn:r68|68]]</sup> ; Frölicher et al., 2016 <sup>[[#fn:r69|69]]</sup> ; Jones et al., 2016 <sup>[[#fn:r70|70]]</sup> ).

‘Detection and Attribution’ assesses evidence for past changes in the ocean and cryosphere, relative to normal/reference-interval conditions ( ''detection'' ), and the extent to which these changes have been caused by anthropogenic climate change or by other factors ( ''attribution'' ) (Bindoff et al., 2013 <sup>[[#fn:r71|71]]</sup> ; Cramer et al., 2014 <sup>[[#fn:r72|72]]</sup> ; Knutson et al., 2017 <sup>[[#fn:r73|73]]</sup> ; Figure 1.1d). Reliable detection and attribution is fundamental to our understanding of the scientific basis of climate change (Hegerl et al., 2010 <sup>[[#fn:r74|74]]</sup> ). For example, the main attribution conclusion of the IPCC 4th Assessment Report (AR4), in other words, that ‘most of the observed increase in global average temperatures since the mid-20th century is ''very likely'' due to the observed increase in anthropogenic greenhouse gas concentrations’, has had a strong impact on climate policy (Petersen, 2011 <sup>[[#fn:r75|75]]</sup> ). In AR5 this attribution statement was elevated to ‘ ''extremely likely'' ’ (Bindoff et al., 2013 <sup>[[#fn:r76|76]]</sup> ). Statistical approaches for attribution often involve using contrasting forcing scenarios in climate model experiments to detect the forcing that best explains an observed change (Figure 1.1d). In addition to passing the statistical test, a successful attribution also requires a firm process understanding. Confident attribution remains challenging though, especially when there are multiple or confounding factors that influence the state of a system (Hegerl et al., 2010 <sup>[[#fn:r77|77]]</sup> ). Particular challenges to detection and attribution in the ocean and cryosphere include the often short observational records (Section 1.8.1.1, Figure 1.3), which are particularly confounding given the long adjustment time scales to anthropogenic forcing of many properties of interest.

Extreme climate events (e.g., marine heatwaves or storm surges) push a system to near or beyond the ends of its normally observed range ( Seneviratne et al., 2012 <sup>[[#fn:r78|78]]</sup> ; Figure 1.1b; Chapter 6;). Extremes can be very costly in terms of loss of life, ecosystem destruction, and economic damage. In a system affected by climate change, the recurrence and intensity of these extreme events can change much faster and have greater impacts than changes of the average system state (Easterling et al., 2000 <sup>[[#fn:r79|79]]</sup> ; Parmesan et al., 2000 <sup>[[#fn:r80|80]]</sup> ; Hughes et al., 2018 <sup>[[#fn:r81|81]]</sup> ). Of particular concern are ‘compound events’, when the joint probability of two or more properties of a system is extreme at the same time or closely connected in time and space (Cross-Chapter Box 5 in Chapter 1; Sections 4.3.4, 6.8). Such a compound event is given, for example, when marine heatwaves co-occur with very low nutrient levels in the ocean potentially resulting in extreme impacts (Bond et al., 2015 <sup>[[#fn:r82|82]]</sup> ). The interconnectedness of the ocean and cryosphere (Section 1.2.2) can also lead to cascading effects where changes in one element trigger secondary changes in completely different but connected elements of the systems, including its socioeconomic aspects. (Figure 1.1e). An example is the large change in ocean productivity triggered by the changes in circulation and iron inputs induced by the large outflow of melt waters from Greenland (Kanna et al., 2018 <sup>[[#fn:r83|83]]</sup> ). New methodologies for attributing extreme events and the risks they bring to climate change have emerged since AR5 (Trenberth et al., 2015 <sup>[[#fn:r84|84]]</sup> ; Stott et al., 2016 <sup>[[#fn:r85|85]]</sup> ; Kirchmeier-Young et al., 2017 <sup>[[#fn:r86|86]]</sup> ; Otto, 2017 <sup>[[#fn:r87|87]]</sup> ), especially also for the attribution of individual events through an assessment of the fraction of attributable risk (Figure 1.1f).

<div id="article-1-3time-scales-thresholds-and-detection-of-ocean-and-cryosphere-change-block-2"></div>

<span id="figure-1.1"></span>
<!-- START IMG -->
<!-- IMG TITLE -->
'''Figure 1.1'''

<span id="figure-1.1-schematic-of-key-concepts-associated-with-changes-in-the-ocean-and-cryosphere.-a-differing-responses-of-systems-to-gradual-forcing-e.g.-linear-delayed-abrupt-nonlinear.-b-evolution-of-a-dynamical-system-in-time-revealing-both-natural-unforced-variability-and-a-response-to-a-new-e.g.-anthropogenic-forcing.-key-concepts-include-i-the"></span>
<!-- IMG CAPTION -->
'''Figure 1.1 | Schematic of key concepts associated with changes in the ocean and cryosphere. (a) Differing responses of systems to gradual forcing (e.g., linear, delayed, abrupt, nonlinear). (b) Evolution of a dynamical system in time, revealing both natural (unforced) variability and a response to a new (e.g., anthropogenic) forcing. Key concepts include (i) the […]'''
<!-- IMG FILE -->
[[File:8b492d9f897a12efa0a3f7c1ab09f20d IPCC-SROCC-CH_1_1-e1574938357439-1.jpg]]

Figure 1.1 | Schematic of key concepts associated with changes in the ocean and cryosphere. (a) Differing responses of systems to gradual forcing (e.g., linear, delayed, abrupt, nonlinear). (b) Evolution of a dynamical system in time, revealing both natural (unforced) variability and a response to a new (e.g., anthropogenic) forcing. Key concepts include (i) the time of emergence and (ii) extreme events near or beyond the observed range of variability. (c) Tipping points and the change of their behaviour through time in response to, for example, anthropogenic change (adapted from Lenton et al., 2008). The two minima represent two stable fixed points, separated by a maximum representing an unstable fixed point, acting as a tipping point. The ball represents the state of the system with the red dash line indicating the stability of the fixed point and the system’s response time to small perturbations. (d) Detection and attribution, i.e., the statistical framework used to determine whether a change occurs or not (detection), and whether this detected change is caused by a particular set of forcings (e.g., greenhouse gases) (attribution). (e) Cascading effects, where changes in one part of a system inevitably affect the state in another, and so forth, ultimately affecting the state of the entire system. These cascading effects can also trigger feedbacks, altering the forcing. (f) Event attribution and fraction of attributable risk. The blue (orange) probability density function shows the likelihood of the occurrence of a particular value of a climate variable of interest under natural (present = including anthropogenic forcing) conditions. The corresponding areas above the threshold indicate the probabilities Pnat and Pant of exceedance of this threshold. The fraction of attributable risk (given by FAR = 1 – Pant/Pnat ) indicates the likelihood that a particular event has occurred as a consequence of anthropogenic change (adapted from Stott et al., 2016).

<!-- END IMG -->
<span id="changes-in-the-ocean-and-cryosphere"></span>