Future changes of circulation types and their effects on surface air temperature and precipitation in the SMHI large ensemble

Abstract

Being strongly influenced by internal climate variability, the atmospheric circulation response to greenhouse gas forcing in the future climate is uncertain. This study addresses atmospheric circulation through representative circulation types (CTs) and investigates the CTs’ changes with respect to frequency and effect on surface temperature and precipitation over a pan-Scandinavian domain. The analysis is based on the Swedish Meteorological and Hydrological Institute Large Ensemble performed with EC-Earth3: 50-member ensembles from one historical and six scenario simulations for the twenty-first century are used to assess internal variability and significance of changes. Although the results show no strong future changes in the CTs’ sea level pressure patterns, CT frequency changes suggest a future extension of summer conditions towards spring and autumn and point towards a clearer distinction between summer and winter. The present-climate CT effect on surface temperature is found to generally weaken which is consistent with a general decrease in temperature variability. Largest-scale and strongest temperature effect changes are projected between March and May for CT3, a cyclone towards the east of the domain associated with a domain-wide cooling effect that will likely be decreased towards the end of the twenty-first century. Similarly but of opposite sign, the CT effect on precipitation is strengthening as a result of the increased precipitation variability. Here, largest-scale and strongest effect enhancement is found for CT6, thus increasing its wetting effect over southern Scandinavia and drying effect west of the Scandes. Changes are generally largest towards the end of the twenty-first century and tend to increase with the forcing strength, thus maximizing for SSP585.

Relationship Between Circulation Types and Extreme Precipitation Over Scandinavia Is Stable Under Climate Change

“…We apply the SANDRA algorithm as implemented in the COST733 software package (see Philipp et al., 2010) to daily mean SLP fields over a pan‐Scandinavian domain (0–30°E, 50–73°N) as described in Hansen et al. (2023). Data from the whole year has been used together.…”

“…To circumvent the computationally challenging task of clustering the whole SMHI-LENS, we first compute a reference set of 10 SLP CTs from a combined 120-year data set over the same domain covering the historical and future periods. As described in Hansen et al (2023), this combined data set is taken from regional climate model simulations and covers similar temporal and ensemble variability. Afterward, each daily SMHI-LENS SLP field from the historical as well as from the two future periods is assigned to the representative CTs based on the smallest Euclidean distance.…”

“…Even the future change signals are compared to the present-day ensemble spread to indicate whether the future climate forced response is larger than the expected perturbations resulting from the internal climate variability. Following Hansen et al (2023) (and equivalent to, e.g., Chapman & Walsh, 2007;Christensen et al, 2007), a signal is considered significant if S/N > 1.…”

“…CT10 shows a mostly westerly flow with a strong SLP gradient indicating strong winds impinging on the Scandes. The resulting uplift of moist air from the North Atlantic induces not only wetter-thannormal conditions upstream and drier-than-normal conditions downstream of the mountains in general (Hansen et al, 2023), but also precipitation extremes over the mountains as can be seen from Figure 3a. The probability of extreme precipitation in CT10 is particularly high over the Scandes in spring (20%), but also in autumn and winter (Figures 3d and 3e).…”

“…The probability of extreme precipitation in CT10 is particularly high over the Scandes in spring (20%), but also in autumn and winter (Figures 3d and 3e). CT10 occurs only very rarely in summer (not shown; see Hansen et al (2023)) and therefore The probability is expressed as the significant deviation from the 1% climatological probability (see text). (b) ERA5; the probability is expressed as the deviation from the 1% climatological probability.…”