The tropospheric response to Sudden Stratospheric Warmings (SSWs) is associated with an equatorward shift in the midlatitude jet and associated storm tracks, while Strong Polar Vortex (SPV) events elicit a contrasting response. Recent analyses of the North Atlantic jet using probability density functions of a jet latitude index have identified three preferred jet latitudes, raising the question of whether the tropospheric response to SSWs and SPVs results from a change in relative frequencies of these preferred jet regimes rather than a systematic jet shift. We explore this question using atmospheric reanalysis data from 1979 to 2018 (26 SSWs and 33 SPVs), and a 202‐years integration of the Whole Atmosphere Community Climate Model (92 SSWs and 68 SPVs). Following SSWs, the northern jet regime becomes less common and the central and southern regimes become more common. These changes occur almost immediately following “split” vortex events, but are more delayed following “displacement” events. In contrast, the northern regime becomes more frequent and the southern regime less frequent following SPV events. Following SSWs, composites of 500‐hPa geopotential heights, surface air temperatures, and precipitation most closely resemble composites of the southern jet regime, and are generally opposite in sign to the composites of the northern jet regime. These comparisons are reversed following SPVs. Thus, one possible interpretation is that the two southernmost regimes appear to be favored following SSWs, while the southernmost regime becomes less common following SPVs.
Sudden stratospheric warmings (SSWs) are the most dramatic events in the wintertime stratosphere. Such extreme events are characterized by substantial disruption to the stratospheric polar vortex, which can be categorized into displacement and splitting types depending on the morphology of the disrupted vortex. Moreover, SSWs are usually followed by anomalous tropospheric circulation regimes that are important for subseasonal-to-seasonal prediction. Thus, monitoring the genesis and evolution of SSWs is crucial and deserves further advancement. Despite several analysis methods that have been used to study the evolution of SSWs, the ability of deep learning methods has not yet been explored, mainly due to the relative scarcity of observed events. To overcome the limited observational sample size, we use data from historical simulations of the Whole Atmosphere Community Climate Model version 6 to identify thousands of simulated SSWs, and use their spatial patterns to train the deep learning model. We utilize a convolutional neural network combined with a variational auto-encoder (VAE)—a generative deep learning model—to construct a phase diagram that characterizes the SSW evolution. This approach not only allows us to create a latent space that encapsulates the essential features of the vortex structure during SSWs, but also offers new insights into its spatiotemporal evolution mapping onto the phase diagram. The constructed phase diagram depicts a continuous transition of the vortex pattern during SSWs. Notably, it provides a new perspective for discussing the evolutionary paths of SSWs: the VAE gives a better-reconstructed vortex morphology and more clearly organized vortex regimes for both displacement-type and split-type events than those obtained from principal component analysis. Our results provide an innovative phase diagram to portray the evolution of SSWs, in which particularly the splitting SSWs are better characterized. Our findings support the future use of deep learning techniques to study the underlying dynamics of extreme stratospheric vortex phenomena, and to establish a benchmark to evaluate model performance in simulating SSWs.
more » « less- NSF-PAR ID:
- 10500665
- Publisher / Repository:
- IOP Publishing
- Date Published:
- Journal Name:
- Environmental Research: Climate
- Volume:
- 3
- Issue:
- 2
- ISSN:
- 2752-5295
- Format(s):
- Medium: X Size: Article No. 025006
- Size(s):
- ["Article No. 025006"]
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract -
more » « less
Abstract Extreme cold events over North America such as the February 2021 cold wave have been suggested to be linked to stratospheric polar vortex stretching. However, it is not resolved how robustly and on which timescales the stratosphere contributes to the surface anomalies. Here we introduce a simple measure of stratospheric wave activity for reanalyses and model outputs. In contrast to the well-known surface influences of sudden stratospheric warmings (SSWs) that increase the intraseasonal persistence of weather regimes, we show that extreme stratospheric wave events are accompanied by intraseasonal fluctuations between warm and cold spells over North America in observations and climate models. Particularly, strong stratospheric wave events are followed by an increased risk of cold extremes over North America 5–25 days later. Idealized simulations in an atmospheric model with a well-resolved stratosphere corroborate that strong stratospheric wave activity precedes North American cold spells through vertical wave coupling. These findings potentially benefit the predictability of high-impact winter cold extremes over North America.
-
more » « less
Abstract Southern Hemisphere (SH) stratospheric sudden warmings (SSWs) result in smaller Antarctic ozone holes and are linked to extreme midlatitude weather on subseasonal to seasonal timescales. Therefore, it is of interest how often such events occur and whether we should expect more events in the future. Here, we use a pair of novel multimillennial simulations with a stratosphere‐resolving coupled ocean‐atmosphere climate model to show that the frequency of SSWs, such as observed 2002 and 2019, is about one in 22 years for 1990 conditions. In addition, we show that we should expect the frequency of SSWs, and that of more moderate vortex weakening events, to strongly decrease by the end of this century.
-
more » « less
Abstract Major sudden stratospheric warmings (SSWs), vortex formation, and final breakdown dates are key highlight points of the stratospheric polar vortex. These phenomena are relevant for stratosphere‐troposphere coupling, which explains the interest in understanding their future changes. However, up to now, there is not a clear consensus on which projected changes to the polar vortex are robust, particularly in the Northern Hemisphere, possibly due to short data record or relatively moderate CO2forcing. The new simulations performed under the Coupled Model Intercomparison Project, Phase 6, together with the long daily data requirements of the DynVarMIP project in preindustrial and quadrupled CO2(4xCO2) forcing simulations provide a new opportunity to revisit this topic by overcoming the limitations mentioned above. In this study, we analyze this new model output to document the change, if any, in the frequency of SSWs under 4xCO2forcing. Our analysis reveals a large disagreement across the models as to the sign of this change, even though most models show a statistically significant change. As for the near‐surface response to SSWs, the models, however, are in good agreement as to this signal over the North Atlantic: There is no indication of a change under 4xCO2forcing. Over the Pacific, however, the change is more uncertain, with some indication that there will be a larger mean response. Finally, the models show robust changes to the seasonal cycle in the stratosphere. Specifically, we find a longer duration of the stratospheric polar vortex and thus a longer season of stratosphere‐troposphere coupling.
-
more » « less
Abstract Deep generative learning cannot only be used for generating new data with statistical characteristics derived from input data but also for anomaly detection, by separating nominal and anomalous instances based on their reconstruction quality. In this paper, we explore the performance of three unsupervised deep generative models—variational autoencoders (VAEs) with Gaussian, Bernoulli, and Boltzmann priors—in detecting anomalies in multivariate time series of commercial-flight operations. We created two VAE models with discrete latent variables (DVAEs), one with a factorized Bernoulli prior and one with a restricted Boltzmann machine (RBM) with novel positive-phase architecture as prior, because of the demand for discrete-variable models in machine-learning applications and because the integration of quantum devices based on two-level quantum systems requires such models. To the best of our knowledge, our work is the first that applies DVAE models to anomaly-detection tasks in the aerospace field. The DVAE with RBM prior, using a relatively simple—and classically or quantum-mechanically enhanceable—sampling technique for the evolution of the RBM’s negative phase, performed better in detecting anomalies than the Bernoulli DVAE and on par with the Gaussian model, which has a continuous latent space. The transfer of a model to an unseen dataset with the same anomaly but without re-tuning of hyperparameters or re-training noticeably impaired anomaly-detection performance, but performance could be improved by post-training on the new dataset. The RBM model was robust to change of anomaly type and phase of flight during which the anomaly occurred. Our studies demonstrate the competitiveness of a discrete deep generative model with its Gaussian counterpart on anomaly-detection problems. Moreover, the DVAE model with RBM prior can be easily integrated with quantum sampling by outsourcing its generative process to measurements of quantum states obtained from a quantum annealer or gate-model device.
