Seasonal Forecasts of the Twentieth Century

Abstract

Capsule Summary New seasonal retrospective forecasts for 1901-2010 show that skill for predicting ENSO, NAO and PNA is reduced during mid-century periods compared to earlier and more recent high-skill decades.

Assessing the robustness of multidecadal variability in Northern Hemisphere wintertime seasonal forecast skill

“…Whilst all these ensembles use prescribed SSTs, in a recent study Weisheimer et al . (2020) presented a coupled hindcast using a similar set‐up to the ASF‐20C but coupled to a version of the NEMO ocean model (and initialised from a coupled reanalysis). This coupled hindcast experiment also shows similar variations in NAO and PNA correlation skill over the twentieth century, with a distinct minimum during the mid‐century period, demonstrating that this variability in skill is not merely a feature of atmospheric model ensembles with prescribed SST boundary conditions.…”

“…This is most likely related to a weakening of the influence of the ENSO on the predictable signal during the middle of the twentieth century (e.g., O'Reilly et al, 2017), which is evident in the signal-to-noise ratio of the different ensembles, and in particular the larger ensembles analysed here, which exhibit lower predictable signals of extratropical large-scale circulation anomalies during the mid-century period. Whilst all these ensembles use prescribed SSTs, in a recent study Weisheimer et al (2020) presented a coupled hindcast using a similar set-up to the ASF-20C but coupled to a version of the NEMO ocean model (and initialised from a coupled reanalysis). This coupled hindcast experiment also shows similar variations in NAO and PNA correlation skill over the twentieth century, with a distinct minimum during the mid-century period, demonstrating that this variability in skill is not merely a feature of atmospheric model ensembles with prescribed SST boundary conditions.…”

Nonstationary weather and water extremes: a review of methods for their detection, attribution, and management

“…The UNSEEN trends approach has the potential to detect nonstationarities in a range of climate extremes and can be applied to centurylong seasonal hindcasts (e.g. Weisheimer et al, 2017Weisheimer et al, , 2020. Furthermore, pooled members from dynamically downscaled climate model large ensembles have been used to estimate precipitation extremes and their nonstationarity in the past (Kirchmeier-Young and Poschlod et al, 2021) and to estimate the likelihood of historic droughts under present-day greenhouse gases (GHGs; Cowan et al, 2020).…”

Impact of Eurasian autumn snow on the winter North Atlantic Oscillation in seasonal forecasts of the 20<sup>th</sup> century

“…Using shorter time scales, the probability of "cherry picking" a period of increased correlation and sampling co-variability with other climate system components increases considerably. Causes for the nonstationarity are still discussed, with possible influences from the Quasi-Biennial Oscillation (QBO), El Niño-Southern Oscillation (ENSO) or simply snow cover variance (Peings et al, 2017;O'Reilly et al, 2017;Tyrrell et al, 2018;Wegmann et al, 2020;Weisheimer et al, 2020). Disentangling co-variability is further challenged by the co-occurrence of increased Eurasian snow cover and increased Ural blocking frequency, questioning the lead-lag relationship between snow cover and blocking (Peings 2019;Song and Wu, 2019;Santolaria-Otín et al, 2021).…”

“…To disentangle the issues of non-stationarity (found in observations) and causality (found in models), we base our investigation on a 110-year long (Weisheimer et al, 2017). This 51-member ensemble experiment with four start dates per year and a forecast length of 4 months has been used in several studies on the predictability of the NAO and other climate patterns (e.g., O'Reilly et al, 2017;Parker et al, 2019;Weisheimer et al, 2019;Weisheimer et al, 2020;O'Reilly et al, 2020). To investigate the influence of land surface conditions, in this case snow cover, on the evolution of the atmospheric state throughout the season, we use a novel, 21-member twin set of the ASF-20C forecasts with perturbed initial land conditions.…”