| Summary: | This dissertation investigates the prediction and predictability of extreme and severe weather on the sub-seasonal to seasonal (S2S) timescale while exploring physical mechanisms that result in successful statistical prediction of atmospheric blocking, extreme temperatures, and severe weather activity. Successful prediction of extremes on the S2S timescale is becoming more desired for several socio-economic sectors. The studies that are presented in this dissertation have provided prediction products that may be used and adapted in real-time and may be useful tools for centers like the Climate Prediction Center (CPC). The mechanisms of blocking onset are also investigated, as atmospheric blocking is a significant producer of extreme weather. This work is the first that presents such a thorough investigation into blocking onset over four sectors. Part I of this dissertation presents the S2S prediction of extreme and severe weather. First, a new statistical model was developed for the prediction of the winter seasonal blocking frequency over Eurasia one month in advance using sea surface temperature, geopotential height at 70-hPa, and sea ice concentration as predictors, and the model captures more than 65% of the interannual variance. Furthermore, we applied the same predictors used for blocking prediction to predict the seasonal occurrence of winter extreme hot and cold days, and the skillful prediction was achieved over Greenland and large portions of Eurasia. Next, an investigation of the relationship between large-scale weather regimes and tornado occurrence in boreal spring is presented. Results show that weather regimes strongly modulate the probability of tornado occurrence in the United States due to changes in shear and convective available potential energy, and that persisting weather regimes (lasting ≥3 days) contribute to greater than 70% of outbreak days (days with ≥10 tornadoes) . A hybrid model based on the weather regime frequency predicted by a numerical model is developed to predict above/below normal weekly tornado activity and has skill better than climatology out to week 3. Lastly, A simple statistical model is developed to predict the weekly frequency of extreme warm days and 14-day standardized precipitation index (SPI) during boreal summer in the United States (US). A leading principal component of US soil moisture and an index based on the North Pacific sea surface temperature are used as predictors. The model outperforms the CFSv2 at weeks 3-4 in the eastern US. Part II explores the mechanisms for atmospheric blocking onset and addresses how blocking impacts the prediction skill of the GEFSv12 over four sectors around the globe. Here, we objectively separate blocking into four regions and present how the blocking onset mechanisms vary from one region to another and relate these mechanisms to blocking predictability. Atlantic blocks are associated with strong low-frequency components of the flow, which resembles the negative phase of the North Atlantic Oscillation. Europe blocks are influenced by a traveling wave across the Atlantic Ocean and develops rapidly, mainly attributed to strong anticyclonic Rossby wave breaking. Asian blocks are fixated within a stationary wave train that spans upstream to the western Atlantic Ocean and contains strong low- and intermediate-frequency variability. The Pacific blocks contain a low-frequency component resembling the Pacific-North American pattern, but are largely influenced by a retrograding wave train within the intermediate-frequency components of the flow. Backward trajectory analysis was also performed, and a large percentage of parcels initialized within the Atlantic, Europe, and Pacific blocking anticyclones experience heating and ascent, while more parcels experience isentropic lift prior to Asian blocks than the other sectors. The impact of atmospheric blocking on the 500-hPa geopotential height prediction skill was examined using the GEFSv12 reforecasts. In general, 7-day prediction skill tends to decrease prior to onset and increase past onset. Our analysis aids in further understanding of the blocking predictability and identifying the sources of predictability over different regions.
|
|---|