An increase in marine heatwaves without significant changes in surface ocean temperature variability

Abstract

Marine heatwaves (MHWs)—extremely warm, persistent sea surface temperature (SST) anomalies causing substantial ecological and economic consequences—have increased worldwide in recent decades. Concurrent increases in global temperatures suggest that climate change impacted MHW occurrences, beyond random changes arising from natural internal variability. Moreover, the long-term SST warming trend was not constant but instead had more rapid warming in recent decades. Here we show that this nonlinear trend can—on its own—appear to increase SST variance and hence MHW frequency. Using a Linear Inverse Model to separate climate change contributions to SST means and internal variability, both in observations and CMIP6 historical simulations, we find that most MHW increases resulted from regional mean climate trends that alone increased the probability of SSTs exceeding a MHW threshold. Our results suggest the need to carefully attribute global warming-induced changes in climate extremes, which may not always reflect underlying changes in variability.

Historical & future maximum ocean temperatures

“…In order to investigate whether these significant changes in Figure 5 were due to mean-SST trends or to changes in interannual variability, Figure 6 shows the ensemblemean interannual SST variance, its change from 1982-2020 versus 2061-2100, and its change from 1982-2020 versus 2061-2100 after detrending. The absence of an increase in interannual variability in the latter case, and that we don't find scale changes for the detrended anomalies data, suggest that the apparent increase in σ is due to increasing warming trends over 2061-2100 in those regions, as in [13,36].…”

“…), the models generally do not predict substantial changes nor agree on the sign of change, i.e. the 90% confidence intervals there include zero over almost all of the ocean, or the change is due to aggregating SSTs over a period with a warming trend artificially increasing the interannual variability [36] in the case of σ raw (see Materials and Methods). The location parameter for the raw SST data increases almost everywhere between the observation period 1982-2020 and 2061-2100, both under SSP1-2.6 and SSP5-8.5 (Figure 3a,b).…”

“…Robust increases in the scale parameter are simulated for the raw data in the tropical Atlantic and Indian Ocean under the SSP5-8.5 scenario (but not SSP1-2.6); these appear to be due to increases in the warming trend rather than interannual variability changes (Materials and Methods) [36], so we do not focus on them here.…”

Ensemble Spread Behavior in Coupled Climate Models: Insights From the Energy Exascale Earth System Model Version 1 Large Ensemble

“…Understanding the magnitude of both forced trends and internal variability is crucial for accurately quantifying the risks of climate‐driven extremes (Deser et al., 2020; Huang & Stevenson, 2021; Xu et al., 2022). The two are additionally likely related: as anthropogenic climate impacts progress during the twenty‐first century, modulations in background conditions are expected to modify both the processes operating during natural low‐frequency climate variability (see examples in the Pacific by Joh & Di Lorenzo, 2017; Joh et al., 2021; Liguori & Di Lorenzo, 2018) and the associated impacts (Fasullo et al., 2018).…”

Record-breaking persistence of the 2022/23 marine heatwave in the Mediterranean Sea

“…Marine heatwaves (MHWs) are persistent and exceptional ocean warming events at a given location exceeding the 'expected' conditions, typically defined according to a reference climatology (Hobday et al 2016). MHWs occurrence is ubiquitous at the global scale, with severe events detected with increasing frequency in recent years in different regions, such as the Tasman Sea (Kajtar et al 2022), the north-west Pacific (Di Lorenzo andMantua 2016, Li et al 2023), the southern Atlantic (Manta et al 2018), the Indian Ocean (Saranya et al 2022) and the Mediterranean Sea (Marullo and Guarracino 2003), among others. Nowadays, it is widely recognized that some of the properties (i.e.…”