Year: 2025
Type: article
Abstract: It is widely agreed that subseasonal to seasonal (S2S) predictability arises from the atmospheric initial state during early lead times and from the land and ocean during long lead times. We test this hypothesis for the large-scale midlatitude atmosphere by training numerous extreme gradient boosting (XGBoost) models to predict weather regimes (WRs) over North America at 1–8-week lead times. Each model uses a different predictor from one Earth system component (atmosphere, ocean, or land) sourced from reanalysis. According to the models, the atmosphere provides more predictability during the first two forecast weeks, and the three components performed similarly afterward. However, the skill and sources of predictability are highly dependent on the season and target WR. Our results show greater WR predictability in fall and winter, particularly for the Pacific trough and Pacific ridge regimes, driven primarily by the ocean (e.g., El Niño–Southern Oscillation and sea ice). For the Pacific ridge in winter, the stratosphere also contributes significantly to predictability across most S2S lead times. Additionally, the initial large-scale tropospheric structure (encompassing the tropics and extratropics, e.g., Madden–Julian oscillation) and soil conditions play a relevant role—most notably for the Greenland high regime in winter. This study highlights the previously identified sources of predictability for the large-scale atmosphere and gives insight into new sources for future study. Given how closely linked WRs are to surface precipitation and temperature anomalies, storm tracks, and extreme events, the study results contribute to improving the S2S prediction of surface weather. Significance Statement Recurrent and persistent large-scale atmospheric patterns are referred to as weather regimes. Algorithms can be trained to predict weather regimes over North America up to 8 weeks into the future. Contributions to prediction skill can originate from different parts of the Earth system, such as stratospheric wind anomalies, subsurface ocean heat content, and soil water content. In this study, we found that contributions to the predictability of weather regimes can vary in importance depending on the lead time, season, and predicted weather regime. These results show that predictability contributions can vary more than previously documented and deepen our understanding of sources of predictability, which can help improve forecasts extending beyond 2 weeks for the benefit of society and decision-makers.
Institution: University of Maryland, College Park
Language: English
Subfield: Global and Planetary Change
Field: Environmental Science
Domain: Physical Sciences
SDG: Climate action
Open Access status: green
Funder: National Science Foundation
OpenAlex ID: W4411437154