Subseasonal Weather Predictions in Northern Hemisphere Winter

Subseasonal-to-Seasonal (S2S), or extended-range, weather predictions cover lead times of up to two months, bridging the gap between medium-range and seasonal forecasts. Accurate predictions at this time-range are important in many socio-economic sectors, including agriculture, energy, and transport. The progress in S2S forecasting has been moderate due to the complexities of atmospheric processes and limitations in numerical models. Unlike medium-range weather forecasting, which predicts weather conditions up to 15 days ahead, S2S predictions do not leverage predictability from initial conditions. Nor do they fully exploit the predictability from boundary conditions, which play a significant role in seasonal outlooks. However, forecast quality can be improved under specific atmospheric conditions. The overall objective of this study is to identify new sources of predictability at the S2S timescale which can increase forecast accuracy. This thesis contributes to subseasonal predictability research, analysing potential windows of opportunity for more accurate forecasting and enhancing existing knowledge in the field. The thesis mainly focuses on the variability of the stratospheric polar vortex and stratosphere-troposphere coupling as a source of additional predictability during the Northern Hemisphere winter. The most prominent phenomenon of the stratospheric variability is Sudden Stratospheric Warming (SSW), which often has a significant impact on tropospheric weather. Paper I of the thesis presents a case study of the 2018 SSW event, which led to a severe cold spell over Northern Eurasia. The study investigates the processes responsible for triggering the stratospheric warming and evaluates the predictability of the event. It was found that accurately capturing the chain of events, which began with anticyclonic Rossby wave breaking, followed by blocking and a record high eddy heat flux, was crucial for the early prediction of the SSW. Furthermore, precisely reproducing both the timing and intensity of the event was essential for reliable forecasts of the subsequent cold spell. Winter-time cold spells, which represent extreme weather events, were further analysed to explore potential sources of enhanced predictability. Papers II and III address this topic by providing insights into the contribution of remote drivers to predictability. The results show that the weakening of the stratospheric polar vortex can extend the predictability of cold spells by three to five days. However, the stratospheric signal can sometimes be misleading, resulting in false alarms of cold spells. This occurs when the downward propagation signal is weak or absent, allowing inner tropospheric processes to shape surface weather. Overall, this thesis contributes to our understanding of stratospheric dynamics and stratosphere-troposphere coupling that are useful in developing forecasting tools and improving subseasonal forecasting.

peerReviewed

ei saavutettava