| dc.description |
Extratropical cyclones and their associated extreme precipitation and wind cause large
damage and loss of life over Europe. When such extremes occur at the same time and same
place they can cause even greater risks. Knowing more about the frequency and spatial
pattern of these extremes over Europe and their relationship to cyclones is important for
forecasting, the (re)insurance industry and emergency planners. This thesis takes both an
Eulerian and a Lagrangian perspective to investigate compound precipitation and wind
extremes over Europe and within cyclones. Observational data including ERA5 as well as
a high resolution climate model, GloSea5, is used to do this.
Over Europe high co-occurrence is found over western coasts and low co-occurrence is
found over eastern coasts. These results are found to be robust over different timescales and
datasets. Cyclones are found to be strongly associated with compound extremes, given an
extreme co-occurring event the chance of a cyclone being nearby is more than 70% for much
of Europe. Using an objective feature tracking method, insight into the frequency, spatial
pattern and footprint size of extremes within cyclones from a Lagrangian perspective is
given. Wind extreme footprints have almost double the area of precipitation and are 5
times larger than compound. The spatial pattern of extremes does not vary much during
a cyclones lifecycle, however, the frequency does. The frequency of extreme wind peaks at
maximum intensity, within the cold conveyor belt to behind the cold front. The frequency
of precipitation and compound extremes peak 6 hours before maximum intensity, ahead
of the cyclone centre, in the warm conveyor belt and warm front. The spatial structure
of extremes within cyclones are very consistent and vary little with seasonality, intensity
and speed. The frequency of extremes and footprint size however varies largely.
GloSea5 is evaluated against ERA5, to find out how well it can represent extremes from
both an Eulerian and Lagrangian perspective and to be used to investigate unprecedented
extremes using the UNSEEN method. Substantial differences in magnitude are found between precipitation and wind extremes between the datasets, however, compound extremes
compare well spatially over Europe. GloSea5 represents the general structure of extremes
within cyclones well. However, extremes at the cyclone centre tend to be overestimated and
regions with the largest frequencies of extremes are underestimated, particularly within
cyclones at maximum intensity. Despite biases in GloSea5 it is shown to have a reasonable
representation of the storm tracks over Europe and the North Atlantic and can simulate
the variability in cyclone frequency well. Although GloSea5 underestimates the number
of storms over Europe there are still winters in GloSea5 with unprecedented number of
storms for some locations. Extreme storm seasons over Europe and over particular locations arise from a variety of winter mean circulation types. GloSea5 has shown that it can replicate circulation patterns that appear feasible in and similar to the real world.
Stormy seasons in GloSea5 are found that produce circulation patterns that are unrealised
in the observational record. These large scale dynamics drive large numbers of storms over
Central and Southern Europe. GloSea5 represents the size of extreme wind, precipitation
and compound footprints very well. It is possible in the current climate to get storms with
footprints of extremes almost double the size of that seen historically. |
|