Global and regional surface cooling in a warming climate: a multi-model analysis

Abstract

to variations in global surface temperature. The likelihood of decadal-scale non-warming periods decrease with global warming, firstly at the low latitude region stretching eastward from the tropical Atlantic towards the western Pacific. The North Atlantic and Southern Oceans have largest likelihood of non-warming decades in a warming world. Keywords Hiatus · CMIP5 · Global warming · Decadal variability MotivationA central issue in climate research in recent years has been the apparent paradox that global surface temperature has not been increasing in tandem with increasing emissions of human-induced greenhouse gases. This paradox has generated a lot of attention among researchers (e.g., Easterling and Wehner 2009;Foster and Rahmstorf 2011;Katsman and van Oldenborgh 2011;Santer et al. 2011Santer et al. , 2014Trenberth and Fasullo 2013;Huber and Knutti 2014;Maher et al. 2014;Risbey et al. 2014;Watanabe et al. 2014) and the general public (Tollefson 2014). It has also been used to cast doubts about the reliability of climate research in general and climate models in particular (Showstack 2014) since the ensemble mean of the models do not reproduce the so-called surface temperature hiatus.A large suite of possible factors influencing the recent hiatus in the global surface temperature have been identified, including: the possibility of too-high model sensitivity to increased greenhouse gas (GHG) forcing (Flato et al. 2013); decadal-scale ocean uptake and storage of heat in the Pacific Ocean (e.g., Meehl et al. 2011;Kosaka and Xie 2013;Meehl et al. 2013;Trenberth and Fasullo 2013;England et al. 2014;Hu et al. 2015), in the Atlantic Ocean Abstract Instrumental temperature records show that the global climate may experience decadal-scale periods without warming despite a long-term warming trend. We analysed 17 global climate models participating in phase 5 of the Coupled Model Intercomparison Project (CMIP5), identifying the likelihood and duration of periods without warming in the four Representative Concentration Pathway (RCP) scenarios RCP2.6, RCP4.5, RCP6.0 and RCP8.5, together with the preindustrial control and historical simulations. We find that non-warming periods may last 10, 15 and 30 years for RCP8.5, RCP6.0 and RCP4.5, respectively. In the models, anomalous ocean heat uptake and storage are the main factors explaining the decadalscale surface temperature hiatus periods. The low-latitude East Pacific Ocean is a key region for these variations, acting in tandem with basin-scale anomalies in the sea level pressure. During anomalously cold decades, roughly 35-50 % of the heat anomalies in the upper 700 m of the ocean are located in the Pacific Ocean, and 25 % in the Atlantic Ocean. Decadal-scale ocean heat anomalies, integrated over the upper 700 m, have a magnitude of about 7.5 × 10 21 J. This is comparable to the ocean heat uptake needed to maintain a 10 year period without increasing surface temperature under global warming. On sub-decadal time scales the Atlantic, Pacific and Southern Oceans all have the a...

Hiatus‐like decades in the absence of equatorial Pacific cooling and accelerated global ocean heat uptake

“…Only 1 out of 330 decades over the 2041–2060 period is a hiatus decade, and no hiatus decade is simulated over 2081–2100. This is qualitatively in agreement with previous studies [e.g., Easterling and Wehner , ; Maher et al ., ; Roberts et al ., ; Medhaug and Drange , ].…”

The rogue nature of hiatuses in a global warming climate

“…Following the null hypothesis of climate variability (i.e., SAT has a white noise distribution), we can schematically explain the likelihood of a decadal hiatus as the ratio of the intensity of (internally generated) decadal climate variability to the intensity of the (externally forced) global warming trend ( Figure 1). Hence, in the context of global warming, the relevant scientific question does not seem to be about the existence of a hiatus but rather seem to be about its likelihood [Maher et al, 2014;Schurer et al, 2015;Roberts et al, 2015;Risbey et al, 2015;Medhaug and Drange, 2015]. This is what we investigate here.…”

“…The SAT is filtered using a low-pass filter with a cutoff frequency set at 10 years and based on a simple step function (i.e., removing all frequency components above the cutoff frequency but not affecting lower frequencies). A probability density function of the time derivative of this filtered SAT is built to represent the likelihood of warming and cooling events equal to or longer than 10 years (other minimum durations have been widely tested in Figure S3; for more exhaustive discussion on this point we refer the reader to Medhaug and Drange [2015]). The decadal internal density distribution is evaluated through the 20-model mean.…”

Feeding the world into the future – food and nutrition security: the role of food science and technology

“…The potential exists for over-focusing on particular outcomes (eg the gravest, extreme temperature increase scenarios) at the expense of preparing and planning for a broader range of future climate trends. Additionally, the application of global climate change predictions to food security is further complicated by the proven nonuniform regional climate trends that occur within an overall changing climate (Medhaug & Drange 2015). This last point is critically important in terms of food security globally in that evidence-based decisions must be made regionally if crop and nutritional policies and practices are to account for changes in climate.…”