• No results found

Extreme Storms in the North Atlantic and Europe

N/A
N/A
Protected

Academic year: 2022

Share "Extreme Storms in the North Atlantic and Europe"

Copied!
44
0
0

Loading.... (view fulltext now)

Full text

(1)

Extreme Storms in the North Atlantic and Europe –

John Hanley

(2)
(3)

Extreme Storms in the North Atlantic and Europe

John Hanley

(4)

This thesis is dedicated to my parents.

Cover image: A cyclone situated close to Iceland.

Photo from: http://en.wikipedia.org/wiki/File:Low_pressure_system_over_Iceland.jpg.

ISBN 978-91-7447-630-9 John Hanley, Stockholm 2013.c

Printed in Sweden by US-AB, Stockholm 2013.

Distributor: Department of Meteorology, Stockholm University.

(5)

Abstract

A study of the most extreme cyclones affecting the North Atlantic and Europe is presented with particular focus on extreme windstorms over the densely pop- ulated area of Western Europe, whose associated high surface wind speeds are capable of causing extensive structural damage and occasionally a loss of life.

A novel cyclone identification and tracking algorithm is presented which explicitly recognises ‘multi-centre cyclones’ (MCCs), defined as cyclonic sys- tems which contain two or three sea-level pressure minima. The method also recognises cyclone merging and splitting events and reduces the number of tracks which would have been spuriously split at some point in their life-cycle.

MCC frequency is shown to increase with storm intensity, with approximately 60% of the top 30% of cyclones constituting MCCs at some point in their life-cycle.

The first findings of the IMILAST (Inter-comparison of MId-LAtitude STorm diagnostics) project, an intercomparison study of 15 cyclone identi- fication and tracking algorithms, are presented. Each method was applied to a 20 year period of the ERA-Interim dataset and results for cyclone frequency, intensity, life-cycle and track location were compared across the methods.

The relationship between the evolution of the most intense wind storms affecting Western Europe (Britain and Ireland, Scandinavia, and Western Con- tinental Europe) and the large-scale atmospheric flow is investigated using an automated cyclone tracking algorithm and an objective measure of cyclone de- structiveness applied to ERA40 and ERA-Interim reanalysis data as well as EC-Earth model output data at two different spatial resolutions. Composite analyses reveal a clear connection between the precise location of upper-level anti-cyclonic wave breaking and cold air intrusion from the north and the po- sition and orientation of an intense jet; this, in turn, plays a crucial role in determining into which region a developing extreme storm will be steered.

(6)
(7)

List of Papers

This thesis consists of an introduction and the following papers:

I Hanley, J. and Caballero, R. (2012). Objective identification and tracking of multi-centre cyclones in the ERA-Interim reanalysis dataset, Quart. J.

Roy. Meteorol. Soc., 138, 612–625.

II Neu, U. and Akperov, M.G. and Bellenbaum, N. and Benestad, R. and Blender, R. and Caballero, R. and Cocozza, A. and Dacre, H.F. and Feng, Y. and Fraedrich, K. and Grieger, J. and Gulev, S. and Hanley, J. and Hewson, T. and Inatsu, M. and Keay, K. and Kew, S.F. and Kindem, I.

and Leckebusch, G.C. and Liberato, M.L.R. and Lionello, P. and Mokhov, I.I. and Pinto, J.G. and Raible, C.C. and Reale, M. and Rudeva, I. and Schuster, M. and Simmonds, I. and Sinclair, M. and Sprenger, M. and Tilinina, N.D. and Trigo, I.F. and Ulbrich, S. and Ulbrich, U. and Wang, X.L. and Wernli, H. (2012). IMILAST – a community effort to intercom- pare extratropical cyclone detection and tracking algorithms: assessing method-related uncertainties, Bull. Am. Met. Soc. doi 10.1175/BAMS-D- 11-00154.1

III Hanley, J. and Caballero, R. (2012). The role of large-scale atmospheric flow and Rossby wave breaking in the evolution of extreme windstorms over Europe, Geo. Phys. Res. Lett., 39, L21708, doi:10.1029/2012GL053408 IV Hanley, J. and Caballero, R. (2013). Rossby wave breaking and extreme

windstorms over Western Europe, Manuscript.

For Papers I, III and IV, I was the main contributor in terms of analysis and writing, with the help of R. Caballero. For Paper II, I contributed analysis from the cyclone identification and tracking algorithm developed in Paper I;

lead authorship of the paper was provided by C.C. Raible, S. Gulev, J.G. Pinto, G.C. Leckebusch and X.L. Wang, with additional input from the rest of the members of the collaboration. The original ideas for Papers I and III came from discussions between R. Caballero and myself; Paper IV was a natural

(8)

continuation of Paper III. The original idea for Paper II was provided by U.

Neu. Reprints were made with permission from the publishers.

The following paper was not included in this thesis:

Caballero, R. and Hanley, J. (2012). Midlatitude Eddies, Storm-Track Diffusivity and Poleward Moisture Transport in Warm Climates. Jour. Atmos.

Sci., 69, 3237–3250, doi:10.1175/JAS-D-12-035.1

(9)

Contents

Abstract v

1 Introduction 11

2 Midlatitude Cyclones 13

2.1 Conceptual models . . . 13

2.2 Baroclinic Instability . . . 15

2.3 Cyclone life-cycles . . . 17

2.4 Cyclone development and planetary-scale flow . . . 18

2.5 Latent heating . . . 18

3 The relationship between cyclones and the large-scale flow 23 3.1 Overview . . . 23

3.2 NAO . . . 23

3.3 Upper-level Rossby wave breaking . . . 24

3.4 Recent statistical findings . . . 24

4 Cyclone identification and tracking 27 4.1 Overview . . . 27

4.2 Historical evolution . . . 27

4.3 Modern analysis . . . 28

5 Summary of papers 31 5.1 Paper I . . . 31

5.2 Paper II . . . 32

5.3 Paper III . . . 32

5.4 Paper IV . . . 33

6 Outlook 35

Bibliography 39

(10)
(11)

1. Introduction

Weather patterns in the midlatitudes are largely dominated by cyclonic sys- tems and their associated winds and precipitation. Through their transport of cold polar air equatorwards and warm sub-tropical air poleward, these cyclonic systems play a crucial role in the transport of energy and moisture across these latitudes. While most of these cyclones produce benign day-to-day weather conditions, a very small subset achieve extremely high surface winds capable of causing extensive structural damage, storm-surge flooding and occasion- ally a loss of life. Scientific and societal interest in these storms has grown in the wake of a number of extremely high-impact events over the past 30 years, such as the windstorms ‘Daria’ (January 1990), ‘Lothar’ (December 1999) and

‘Kyrill’ (January 2007). ‘Daria’ and ‘Lothar’, in particular, were each respon- sible for over 100 deaths and inflation-adjusted economic losses in excess of US$ 10 billion, comparable to those caused by a major US hurricane (Berz, 2005).

This thesis focuses on such extreme, high-impact windstorms, particularly over the densely populated area of Western Europe. The thesis consists of four papers, which address two main topics of research:

• Identifying and tracking midlatitude cyclones in either reanalysis data or climate model output data.

• Studying the large-scale atmospheric features associated with the de- velopment and intensification of the most destructive windstorms over Western Europe.

Paper I introduces a novel method for identifying and tracking multi-centre cyclones in reanalysis data which also allows for the recognition of cyclone merging and splitting events. Paper II is an intercomparison study of the method-related uncertainties associated with extratropical cyclone identifica- tion and tracking algorithms, with most of the main methods of the past 30 years included (including the algorithm presented in Paper I). Paper III is a study of the relationship between upper-level Rossby wave breaking and the evolution of the most destructive windstorms affecting the most densely popu- lated area of Continental Europe. Paper IV is an extension of Paper III, focus- ing on a wider area of Western Europe as well as investigating how well such

(12)

storms, and their associated large-scale features, are captured in a state-of-the- art global circulation model.

Chapters 2–4 of this thesis provide a brief introduction to the topics cov- ered in Papers I-IV, Chapter 5 presents a summary of the main findings of the papers and Chapter 6 gives a outline of possible future work.

(13)

2. Midlatitude Cyclones

In this section, we will first give a conceptual description of the life-cycle of a typical midlatitude cyclone before proceeding to briefly outline the theory of midlatitude cyclogenesis, a process primarily driven by baroclinic instability (Charney, 1947; Eady, 1949). Sections 2.2 and 2.3 closely follow chapters 6 and 8 of Holton (2004).

2.1 Conceptual models

The first widely adopted conceptual model of cyclone development was de- veloped by Bjerkness & Solberg (1922) at the Bergen school of Meteorology and was known as the Norwegian cyclone model. This conceptual model orig- inated from detailed analyses of synoptic weather charts, from which the con- cept of cyclogenesis as a disturbance on a stationary polar front (the boundary where a cold air mass from the north meets a warm air mass from the south) was developed. As this initial disturbance grows, the polar front is deformed in such a way to produce characteristic warm and cold fronts.

Figure 2.1 (a) shows the four stages of development in the Norwegian cy- clone model. Stage I shows the cyclone beginning as an initial small pertur- bation on a stationary front. In stage II, a clearly defined cyclonic circulation advects cold air equatorward and warm air poleward of the cyclone centre, thereby forming cold and warm fronts respectively. In stages III and IV, an occluded front is formed as the faster moving cold front catches up with the warm front (Schultz et al., 1998).

The Norwegian cyclone model served as the primary conceptual model for cyclone development for almost 70 years before being updated by the Shapiro- Keyser model, which was developed based on both observations and numerical simulations (Shapiro & Keyser, 1990). Figure 2.1 (b) shows the four stages of development for this model. If we compare these stages with the Norwegian cyclone model, we see a similar evolution in stages I and II, with the exception of a slightly different angle forming between the cold and warm fronts in stage II; in the Shapiro-Keyser model, the cold front is orientated approximately perpendicular to the warm front. This feature is clearer in stage III, where a

‘frontal T-bone’ has formed (Shapiro & Keyser, 1990). Unlike the Norwegian

(14)

Figure 2.1: From Schultz et al. (1998) [their Figures 2 and 12 combined]. This figure shows (a) the Norwegian conceptual model of cyclone development and (b) the Shapiro-Keyser model. The lower panels show lower-tropospheric po- tential temperature while the top panels show lower-tropospheric geopotential heights and associated cyclone fronts.

cyclone model, a narrowing warm tongue does not develop in the Shapiro- Keyser model and therefore there is no catch-up; instead, a bent-back front develops with warm-core seclusion (Schultz et al., 1998).

Conceptually, the three dimensional structure of a cyclone can be described in terms of three main airflows or ‘conveyer belts’; the warm conveyer belt (WCB), the cold conveyer belt (CCB) and a dry intrusion (Carlson, 1980).

The WCB is a band of warm, moist air which originates at low levels within the warm sector and moves poleward ahead of the cold front, ascending over the warm front where it can often split into two branches; one branch turns cyclonically, forming part of the upper-level cloud head, while the other turns anti-cyclonically joining the westerly jet flow (Browning, 2004; Browning &

Roberts, 1994; Carlson, 1980). The WCB can ascend to heights of greater than 7 km, transporting warm, moist air which then forms clouds and precipitation (Browning, 1985).

The CCB is a cold, moist band of air originating at low levels ahead of the warm front and flowing underneath the warm frontal zone (Carlson, 1980).

Previous work suggests that after passing underneath the warm frontal zone, the CCB can take one (or both) of two paths: studies prior to 1980 (Schultz, 2001) depict the CCB turning cyclonically around low centre, remaining at low levels behind the cold front; Carlson (1980) demonstrated that the CCB can turn anti-cyclonically and ascend along the cold side of the warm front; while

(15)

ascending, it receives moisture from the WCB, leading to the development of a cloud head.

The final airflow is the dry intrusion, an air stream which originates in the upper troposphere or lower stratosphere. This airflow descends on the western side of the upper-level trough before fanning out and ascending over the bent- back front and descending behind the surface cold front (Schultz, 2001). Due to its origins in the upper atmosphere, this airflow usually contains very little moisture and can often be identified in satellite imagery as a dark, cloud-free region behind the cold front.

2.2 Baroclinic Instability

The sphericity of the Earth results in a greater amount of solar radiation inci- dent on those latitudes located close to the equator in comparison with those nearer the pole. This uneven heating creates a meridional temperature gradi- ent which produces a vertical shear of the geostrophic wind; the relationship between a horizontal temperature gradient and an increase in the geostrophic wind with height is described by the thermal wind relationship, which in vector form is given by:

VT =∂Vg

∂ p = ˆk

f × ∇∂ Φ

∂ p

where VT is the thermal wind vector, Vg is the geostrophic wind vector, p is the pressure, f is the Coriolis parameter, Φ is the geopotential height and ˆk is the vertical unit vector.

In the presence of a horizontal temperature gradient on a pressure surface and the associated vertical wind shear, the flow is baroclinic and there exists available potential energy which may be extracted by baroclinically growing disturbances and converted into kinetic energy.

The process by which cyclones grow from an initial baroclinic instability was mathematically developed independently by Charney (1947) and Eady (1949). Although both theories describe the main features of cyclogenesis, the focus here will be on the Eady model. Analysis of the Eady model shows that the growth rate of a cyclone is inversely proportional to the static stability (N). For higher static stabilities, a low-level circulation cannot extend as far vertically through the tropopause; therefore, static stability effectively acts to dampen the growth of an initial disturbance. The Eady model also reveals that long waves are baroclinically more unstable than short waves and will grow exponentially with time. The typical cyclone wavelength predicted from the model is of the order of 1000 km with a growth rate of approximately 1 day;

(16)

both of these values are reasonable estimates for the growth of cyclones in the midlatitudes.

A useful description of the mechanism of baroclinic instability is in terms of potential vorticity (PV), a concept first used by Rossby (1940) and Ertel (1942). PV is defined as:

PV = 1 ρζ .∇θ

where ρ is the density of the air, ζ is the absolute vorticity and θ the potential temperature.

Rossby (1940, 1939) first identified large-scale planetary waves (later known as Rossby waves) as transverse waves with a gradient of potential vorticity serving as their restoring force. These waves can be excited through interaction of the mean flow with orography, through surface forcing due to a land–ocean temperature contrast and in the presence of large-scale baroclinicity.

A description of baroclinic instability in terms of potential vorticity was first developed by Bretherton (1966a,b) and later added to by Hoskins et al.

(1985). It describes the instability in terms of the interaction of two counter- propagating Rossby waves. A qualitative overview of this process follows.

PV has the useful property that it is conserved for adiabatic, frictionless flow, i.e. D(PV )Dt = 0, a fact which allows for the movement of air masses to be traced. Generally speaking, PV in the troposphere is small and positive whereas PV in the stratosphere is much higher due to stronger vertical potential temperature gradients; PV can therefore be used to identify air originating from the stratosphere which has descended into the troposphere.

Another useful property of PV is the ‘invertibility principle’, first intro- duced by Hoskins et al. (1985); it states that under a suitable balance condition, such as geostrophic balance, if the full three-dimensional global distribution is known, the wind and temperature fields can be obtained by inverting this PV distribution. This principle leads to the concept of ‘action at a distance’, which describes how a local PV anomaly can induce a circulation at a distance.

Figure 2.2 shows a schematic picture of how PV anomalies lead to cyclone formation via baroclinic instability. Due to the increase of PV with height and latitude, a small downwards or southwards perturbation leads to the creation of a positive PV anomaly. Relative to its surroundings, this high PV air has both higher relative vorticity (ζ ) and higher static stability (N2=θg∂ θ∂ z). This positive upper-level PV anomaly has associated cyclonic vorticity which, via the principle of ‘action at a distance’, creates a weak cyclonic vorticity at the surface.

This weak surface circulation acts on the surface temperature gradient, cre- ating a cold temperature anomaly to the west of the upper-level PV anomaly

(17)

Figure 2.2: A schematic picture of cyclogenesis associated with an upper-level positive vorticity perturbation over a lower-level baroclinic region [taken from Holton, Fig 8.1, fourth edition].

and a warm temperature anomaly to the east. Again, via the principle of ‘ac- tion at a distance’, the lower-level warm temperature anomaly itself induces an upper-level cyclonic circulation which will then act to reinforce the origi- nal upper-level anomaly. The resulting upper and lower level disturbances are slightly out of phase, with the lower-level cyclone ahead of the upper-level cy- clone. The two features are now locked together, mutually intensifying each other and forming a baroclinic wave; this process of mutual intensification is known as baroclinic instability and it is the primary mechanism driving the creation of cyclones in the midlatitudes.

2.3 Cyclone life-cycles

The Eady model briefly outlined in Section 2.2 is a linear theory; therefore, strictly speaking, it applies only to small perturbations. Since the atmosphere is a non-linear system, extending the theory into the non-linear regime has been the focus of research over recent decades. In particular, Thorncroft et al.

(1993) outlined two paradigms of non-linear cyclone development. The first paradigm was denoted LC1, which consists of a normal mode perturbation on the basic state westerly jet. At the surface, a strong cold front dominates a much weaker warm front while at upper levels the associated tropopause PV wave ‘breaks’ anticyclonically at the moment of occlusion – this results in the cyclone decaying to cyclolysis. The second paradigm, denoted LC2, forms with the addition of a barotropic contribution to the zonal flow. At the

(18)

surface, this results in a low pressure system dominated by a strong warm front and weak cold front, while at upper levels the associated tropopause PV wave

‘breaks’ cyclonically at the moment of occlusion – this results in a cut-off cyclone which does not decay.

2.4 Cyclone development and planetary-scale flow

The theoretical models of baroclinic instability outlined in Section 2.2 repre- sent midlatitude cyclones as disturbances superimposed on a zonal flow that varies only with latitude and height. While this viewpoint is useful in un- derstanding the underlying mechanisms driving cyclogenesis, it is only a first approximation to the reality of synoptic-scale disturbances in the atmosphere.

Part of the additional complexity is due to the fact that developing cyclonic systems in the midlatitudes are embedded in a large-scale flow which is it- self baroclinic. This planetary-scale flow, with wavelengths of the order of 1,000 – 10,000 km (compared with approximately 100 - 1,000 km at the syn- optic scale), is itself highly influenced by highly longitudinal-dependent fac- tors such as the presence of orography and continent-ocean heating contrasts.

These planetary-scale waves are relevant to Papers III and IV of this thesis.

Planetary waves in the atmosphere are identifiable as large-scale meanders in the upper-level flow. This leads to the creation of upper-level ridges and troughs which accelerate and decelerate the upper-level flow respectively; thus, the poleward-flowing sections of the upper-level flow will form jet streams with lifetimes of the order of a few days. Accelerating flow (the passing of the flow from a trough to a ridge) is associated with divergence which results in convergence at the surface and falling pressure. Thus, locations downstream of an upper-level trough are associated with areas favourable for cyclogenesis.

The time-averaged zonal wind component for December, January and Febru- ary on the 200-hPa surface in the Northern Hemisphere is shown in Figure 2.4 and gives a guide to the typical zonal flow in which synoptic-scale eddies are embedded. It shows strong zonal wind maxima (jets) just off the east coasts of North America and Asia and north of the Arabian peninsula. As outlined, synoptic-scale eddies tend to develop preferentially in regions of maximum zonal winds associated with these jets and propagate downstream, approxi- mately following the jet axes, to produce the Atlantic and Pacific storm tracks.

2.5 Latent heating

Recent work has shown that latent heat release can play an important role in the rapid intensification of a cyclone (Moore & Montgomery, 2005; Parker &

(19)

Figure 2.3: Mean zonal wind at the 200-hPa level for December-February aver- aged for years 1958–1997 [taken from Holton, Fig 6.2, fourth edition]. Contour interval 10 m s−1(heavy contour, 20 m s−1). (Based on NCEP/NCAR reanalysis after Wallace, 2003.)

Thorpe, 1995; Wernli et al., 2002). As discussed in Section 2.1, the structure of a cyclone can be described in terms of three main airflows; the warm con- veyer belt, the cold conveyer belt and a dry intrusion. The warm conveyer belt consists of a band of warm, moist air which cools as it ascends ahead of the low pressure centre. As the air cools, the moisture within it condenses out to form clouds and precipitation. As the moisture changes phase from gas to liquid, it releases latent heat. This local heating results in a positive PV anomaly be- ing created below, which then enhances the circulation of the surface cyclone (Martin, 2006).

Recent modeling studies have studied the affect this latent heating has on the rapid intensification of some historically well-known storms; for example,

(20)

Wernli et al. (2002) performed a sensitivity study using mesoscale model sim- ulations of the December 1999 extreme windstorm ‘Lothar’. Simulations of

‘Lothar’ were made with diabatic components in the model turned on and off and these simulations were compared with ECMWF analysis and observations.

Figure 2.4 shows the time evolution of the minimum core sea-level pressure of

‘Lothar’ from these four sources. Clearly, latent heating appears to be playing a crucial role in ‘Lothar’s rapid intensification.

Figure 2.4: Time evolution of the minimum sea-level pressure in the core of the cyclone, from observations (German Weather Service DWD), ECMWF analyses, and HRM mesoscale hindcast simulations (moist and dry). [taken from Wernli et al.(2002), Figure 6].

More recently, Fink et al. (2012) applied a modified form of the pressure tendency equation (PTE) to ERA-Interim reanalysis data to quantify the con- tribution of latent heating to the deepening of five well-known recent histor- ical windstorms, namely; ‘Lothar’ and ‘Martin’ in December 1999, ‘Kyrill’

in January 2007, ‘Klaus’ in January 2009 and ‘Xynthia’ in February 2010.

They found diabatic processes played an important role in the intensification of ‘Xynthia’, ‘Klaus’ and ‘Lothar’ but they had a less prominent role in the intensification of ‘Kyrill’ and ‘Martin’ – these storms appeared to be primarily

(21)

driven baroclinically.

(22)
(23)

3. The relationship between

cyclones and the large-scale flow

3.1 Overview

As outlined in Section 2.4, planetary-scale flow can influence the growth and development of synoptic-scale disturbances. Papers 3 and 4 in this thesis ex- plore the role of this large-scale flow in the evolution of the most destructive windstorms over Western Europe, with particular focus on their interaction with the dominant large-scale mode of variability in the region, the North At- lantic Oscillation (NAO, Walker (1924)). This chapter gives a brief overview of the NAO before considering planetary-scale flow in general. Some recent statistical findings will also be presented.

3.2 NAO

The NAO is the dominant mode of variability in the North Atlantic and refers to the redistribution of atmospheric mass between the Arctic and the subtropical Atlantic (Hurrell, 1995; Hurrell et al., 2003). Swings from one NAO phase to another correspond to changes in the strength and position of a semi-permanent dipole over the North Atlantic, characterised by a low pressure system located over Iceland (the Icelandic Low) and a high pressure system located over the Azores (the Azores High), which in turn produce large changes in the mean wind speed and direction over the Atlantic. These changes in wind speed and direction influence the heat and moisture transport into the North Atlantic and Europe, as well as the frequency and intensity of midlatitude cyclones (Hurrell et al., 2003).

A positive phase of the NAO corresponds to a deeper Icelandic low and stronger Azores high; the increased pressure gradient results in an increase in the frequency and intensity of midlatitude cyclones. It also shifts the storm track to a more northerly orientation, resulting in warm, wet, stormy weather over Northern Europe and cool, dry weather over Southern Europe. A negative phase of the NAO corresponds to a shallower Icelandic low and weaker Azores high, which results in a decreased pressure gradient over the North Atlantic

(24)

and Europe. This reduced gradient leads to a decrease in the frequency and intensity of midlatitude cyclones in the region and a shift in the storm track to a more zonal direction, which reverses the main airflows over Europe compared with the positive phase of the NAO.

3.3 Upper-level Rossby wave breaking

Recent work has revealed that the NAO can be dynamically interpreted in terms of upper-level Rossby wave breaking (Benedict et al., 2004; Kunz et al., 2009; Rivière & Orlanski, 2007; Woollings et al., 2008) and that the onset of midlatitude blocking episodes can also be associated with these type of wave breaking events (Pelly & Hoskins, 2003).

Figure 3.1 shows a composite of potential temperature over 10 positive NAO events. The dark shading highlights strong anti-cyclonic wave breaking (defined as a reversal in the north–south potential temperature gradient) in the days preceding, and during, peak positive NAO events. The lighter shading shows a simultaneous strong intrusion of cold polar air originating over the North American continent. This intrusion of cold air collides with the warm air intrusion associated with the strong anti-cyclonic wave breaking, creating an extremely intense, focused upper-level jet and a region of high baroclinicity in the mid-Atlantic.

3.4 Recent statistical findings

The relationship between the phase of the NAO and the frequency and in- tensity of extreme windstorms over the North Atlantic and Europe has been documented in previous studies (Serreze et al., 1997). More recently, Pinto et al.(2009) demonstrated that extreme cyclones in the North Atlantic occur more (less) frequently during strong positive (negative) NAO phases while the occurrence of extreme European storms is associated with moderately-positive phases of the NAO. Raible (2007) also demonstrated a correlation between an positive NAO-like pattern in the 500 hPa geopotential height and the occur- rence of extreme cyclones over Northern Europe.

The association of large-scale atmospheric flow in general with windstorms over Europe has been extensively examined in previous studies, a summary of which can be found in Donat et al. (2010). Recently, Leckebusch et al. (2008) performed a ‘k-means’ clustering algorithm on 3-day MSLP development and found that four primary storm clusters were associated with 72% of histori- cal extreme storm events over Continental Europe. Using an objective scheme for daily flow classification based on Jones et al. (1993), Donat et al. (2010)

(25)

Figure 3.1: A composite of potential temperature fields over 10 positive NAO events [taken from Benedict et al. (2004), Figure 3]. The contour interval is 5 K, with the 320–K line emboldened to emphasize the low–over–high structure, the anticyclonic wave breaking, and the SW–NE title. Darker (lighter) shading denotes positive (negative) t values that exceed the 98% confidence level. The corresponding lag days are shown to the right of each panel. The wind vectors are denoted by arrows.

showed that approximately 80% of storm days in Continental Europe, and most of the severe historical storms (e.g. ‘Daria’ (January 25/26, 1990), ‘Weibke’

(February 28, 1990) and ‘Anatol’ (December 3, 1999; cf. Ulbrich et al. (2001)) are associated with a westerly flow. A corresponding MSLP pattern showed that this flow was associated with an anomalous low pressure pattern centred over Scandinavia.

Rivière & Joly (2006) investigated the tendency of midlatitude cyclones to explosively deepen in specific specific areas relative to the large-scale flow, such as the jet-exit regions. They show that the strongest deepening rates of midlatitude cyclones developing in the northeast Atlantic tend to occur down-

(26)

stream of the baroclinicity maximum, in the jet-exit region. This suggests that large baroclinicity is a necessary condition for their development but does not give any information on the location of explosive growth. Rivière & Joly (2006) demonstrate that this explosive deepening is the result of the interaction between the developing cyclone and the upper-level jet as the cyclone ‘crosses the jet’.

(27)

4. Cyclone identification and tracking

4.1 Overview

Extratropical cyclones exhibit a wide range of physical characteristics: their scale can vary by an order of magnitude from small-scale initial disturbances of less than 100 km in diameter to large-scale mature systems of over 1000 km in diameter; their three-dimensional structure can vary from those cyclones confined to the lower troposphere to those which extend throughout the tropo- sphere; they can remain stationary for long periods of time or they can have translational velocities of greater than 100 km/hr; and they can merge, split and re-merge during their lifecycle. As a result, identifying and tracking these features, a seemingly straight-forward task, is extremely challenging. A tes- tament to the complexity involved is the fact that no single commonly agreed definition of what exactly constitutes a midlatitude cyclone exists within the scientific community (Neu et al., 2012).

4.2 Historical evolution

The identification and tracking of midlatitude cyclones began in earnest in the mid 19th Century with the first simultaneous surface observations, made pos- sible by the arrival of the electric telegraph. These observations allowed for the first synoptic weather charts to be drawn, from which features such as sur- face cyclones could be manually detected. The first studies of midlatitude cy- clones which utilized these new synoptic charts included Köppen in 1881 and VanBebber in 1891 who recognized the key role of cyclones in synoptic-scale variability (Raible et al., 2008). Cyclone identification and tracking remained a laborious and time-consuming task for over 100 years, before the develop- ment of increasingly power computers allowed for the process to be eventually automated.

(28)

4.3 Modern analysis

Modern analysis of synoptic-scale variability from reanalysis data has typically taken the form of an Eulerian or Lagrangian measure. One common Eulerian measure is the so-called dynamic storm track, which is defined as a region of enhanced standard deviation of the bandpass filtered 500-hPa geopotential height, e.g. Blackmon (1976); Lau (1988). A more common approach over the past 30 years has been to take a Lagrangian approach by considering individual cyclone tracks. The automated methods developed over this period consist of two steps: cyclone identification and cyclone tracking.

The choice of field used for cyclone identification varies according to the features of interest one wishes to capture; the high-frequency vorticity field is more closely related to the wind field, while the central pressure is linked to the mass field and captures the low-frequency scale better (Hodges et al., 2003; Neu et al., 2012). Those methods using sea-level pressure or geopo- tential height typically identify cyclone centres by searching for local min- ima in these fields (Blender et al., 1997; Hanley & Caballero, 2012; Wernli

& C., 2006), while those methods using the relative vorticity field search for local maxima (Sinclair, 1994). A number of methods combine both fields at the identification stage, by first searching for local vorticity maxima and then searching for an associated ‘open’ or ‘closed’ pressure depression (the terms ‘open’ and ‘closed’ refer to whether pressure depressions have regions of closed sea-level pressure isobars or not) (Hodges, 1994; Murray & Sim- monds, 1991; Pinto et al., 2005).

Aside from such a wide range of cyclone identification possibilities, the methods employed in cyclone tracking also vary considerably (Raible et al., 2008). Cyclone tracking is further complicated by the issue of cyclone merging and splitting; the choice of what constitutes the most appropriate track in the case of the generation of a new cyclone centre close to an existing one is a particularly difficult problem (Hanley & Caballero, 2012; Neu et al., 2012).

The issue of objectively comparing the large number of existing cyclone identification and tracking methods and their strengths and weaknesses has recently begun to be addressed though an inter-comparison project named IM- ILAST1, which was founded in 2009 with the intention of providing ‘a quan- titative comprehensive assessment of all types of uncertainties inherent in the mid-latitudinal storm tracking by comparing different methodologies’. The first set of results to emerge from this inter-comparison project have recently been published (Neu et al., 2012) and this paper constitutes the second paper in this thesis.

1http://www.proclim.ch/imilast/index.html

(29)

Figure 4.1: This figure shows the cyclone centres (mean sea-level pressure min- ima) identified by the Hanley & Caballero (2012) algorithm from a mean sea- level pressure snapshot over the North Atlantic region. The cyclone centres are denoted with red dots while mean sea-level contours are plotted at intervals of 2 hPa.

(30)
(31)

5. Summary of papers

5.1 Paper I

Paper I introduces a novel cyclone identification and tracking algorithm based on mean sea-level pressure (MSLP) which recognises ‘multi-centre cyclones’

(MCCs); a ‘multi-centre cyclone’ is defined as a cyclonic system which con- tains two or three MSLP minima within its outermost MSLP contour. This method was developed as a solution to the problem of choosing between a cy- clone tracking algorithm which applies spatial smoothing to the MSLP field prior to cyclone identification, but which may remove small-scale features of interest as a result, and an algorithm which does not apply such spatial smooth- ing, but which may have difficulty in correctly tracking the increased number of identified MSLP minima (particularly MSLP minima in close proximity to each other). By recognising such ‘multi-centre cyclones’, the method al- lows for a more consistent evolution of cyclone size, while also reducing the number of cyclone tracks which would have been spuriously split as a result of neighbouring, short-lived MSLP minima. The method also recognises cy- clone merging and splitting events, allowing statistics on these phenomena to be gathered.

The relationship between the frequency of MCC occurrence and cyclone intensity is investigated. MCC frequency is shown to increase with storm in- tensity, with approximately 60% of the top 30% of cyclones constituting MCCs at some point in their life-cycle. This result is of particular importance for Papers III and IV which focus on extreme cyclones. The frequency of track reconnection (those tracks which would have been spuriously split using a con- ventional single-centre method) is also shown to increase with storm intensity, with approximately 40% of the top 30% of cyclones requiring reconnection.

The error rate of this track reconnection is estimated to be approximately 20%;

this high error rate is considered a reasonable price to pay for the improve- ment in tracking associated with the remaining 80% of re-connected tracks.

Statistics on the frequency of cyclone merging and splitting events are also presented, along with track density plots of cyclones at various stages of de- velopment.

(32)

5.2 Paper II

Paper II presents the first findings of the community project IMILAST (In- tercomparison of MId-LAtitude STorm diagnostics), an intercomparison study of 15 cyclone identification and tracking methods (including the method pre- sented in Paper I). These 15 methods are compared using the same input data (20 years of ERA-Interim data with a spatial resolution of 1.5 degrees and a temporal resolution of 6 hours) to assess their similarities and differences. All methods were run in their ‘standard form’, with the only standardization ap- plied the condition of a 24 hour minimum cyclone lifetime. The methods were applied to both the Northern and Southern Hemispheres.

Differences between the various methods are explored by comparing re- sults for cyclone frequency, intensity, life-cycle and track location. Compar- isons of the geographical distribution of cyclone centre densities for the win- ter season reveals a good qualitative agreement amongst the methods in both hemispheres. Despite this agreement, there are relatively large differences in the total number of cyclones identified; in the Northern Hemisphere, the total number of winter cyclones varies from 6,000 to 21,000 while in the South- ern Hemisphere, the total number of summer cyclones varies from 5,000 to 28,000.

Analysis of life-cycle characteristics reveal a reasonable agreement across the methods while there is good agreement with regards to cyclone interannual variability, the distribution shape for many life cycle characteristics, and geo- graphical patterns of strong trends. Consistency amongst the methods is gen- erally higher for deep (or strong) cyclones compared with shallow (or weak) ones. Aside from differences in total cyclone numbers, large disparities also exist with respect to the detection of weak cyclones and cyclone distribution over some densely-populated regions. Two case studies of intense cyclones, one in each hemisphere, reveal a close agreement in the identification of the most intense part of the cyclone’s life cycle but with considerable differences with regards to the development and dissolution phases.

5.3 Paper III

Paper III investigates the interaction between large-scale planetary waves and the most extreme wind storms affecting Western Continental Europe (WCE).

Applying the tracking algorithm presented in Paper I to the ERA-40 43-year reanalysis dataset, a climatology of cyclones is generated over the Northern Hemisphere, from which the most destructive WCE wind storms are selected using a destructiveness measure. A ranking of the top 100 most destructive wind storms in the region shows an exponential-like distribution, with the top

(33)

25 storms dominating; these 25 storms are then selected, with the goal of as- sessing what characteristics, if any, they share. This subset also contains most of the major named European windstorms of the period (1958–2001), lend- ing confidence that both the tracking algorithm and the method for measuring destructiveness are performing well.

Detailed analysis of each one of these 25 storms reveals that 22 of them can be considered as having a similar trajectory and evolution. These 22 storms are subsequently grouped together, with the term ‘embedded storms’ intro- duced to identify them. Temporal composites reveal that these storms typi- cally occur during persistent, moderately-positive North Atlantic Oscillation (NAO) events which peak approximately 2 days before maximum destructive- ness. The evolution of spatial composites of mean sea-level pressure reveal an eastward shift of the NAO dipole pattern which results in a large-scale low positioned over Scandinavia at the moment of maximum destructiveness. This creates a strong MSLP gradient over WCE, on top of which the storms super- impose their own MSLP gradient, resulting in extremely destructive winds in the region.

For a dynamically oriented perspective on the evolution of these ‘embed- ded storms’, temporal composites of θPV2(potential temperature on the nom- inal tropopause, the 2-PVU surface) are computed. They reveal that the east- ward NAO dipole shift observed at the surface can be associated with concur- rent cyclonic and anti-cyclonic wave breaking in the mid-Atlantic penetrating much further eastward compared with climatological high NAO events. These contemporaneous wave breaking events serve to intensify and elongate an ex- tremely strong, zonally-orientated upper-level jet which plays an important role in the explosive intensification of pre-existing cyclones while also steer- ing these cyclones into WCE; this is in contrast to climatological high NAO events where the upper-level jet steers cyclones on a path close to Iceland.

5.4 Paper IV

Paper IV is a natural continuation of the research contained in Paper III. Here, the regions of Scandinavia and Britain and Ireland are also included along with WCE. Analyses are performed on the ERA-40 and ERA-Interim reanalysis data sets (at resolutions of T159 and T255, respectively) as well as EC-Earth model output (run in AMIP mode) at two different spatial resolutions (T159 and T511). A similar approach to Paper III is taken, where the top 25 most destructive storms in each region in each dataset are selected from a Northern Hemisphere cyclone climatology.

Before considering the evolution of these destructive storms, an assess- ment of how well both model resolutions capture the extreme winds in each

(34)

region compared with the reanalysis data sets is performed. The EC-Earth model at T511 resolution shows a very similar distribution to that obtained us- ing ERA-Interim data, while the T159 resolution model exhibits large biases, particularly along coastal regions. Ranking the top 25 most destructive storms in each region in each dataset reveals an exponential-like distribution, similar to that obtained in Paper III.

Temporal composites of the NAO index reveal that these storms typically occur during persistent, moderately-positive NAO events which peak approx- imately 2 days before maximum destructiveness in the case of WCE and be- tween 1–3 days before maximum destructiveness in the case of Britain and Ireland. The most extreme storms over Scandinavia typically occur during static, neutral NAO events.

MSLP composites reveal a similar large-scale pattern observed in Paper III for WCE, with an eastward shift of the NAO dipole pattern resulting in a large-scale low pressure system located over Scandinavia at the moment of maximum destructiveness; ERA-Interim shows a similar evolution to ERA40 while both model resolutions also broadly agree with the reanalysis data, al- though the T511 resolution is in closer agreement. For the most extreme storms over Britain and Ireland, a similar evolution is observed, but without the same eastward shift of the NAO dipole. In Scandinavia, a large-scale, persistent blocking over Europe dominates and deflects eastward propagating cyclones northwards into Scandinavia.

θPV2 composites reproduce the results of Paper III for the WCE region.

In Britain and Ireland, we see similar concurrent wave breaking in the mid Atlantic compared with WCE, but without the same eastward penetration – this orientates the upper-level jet in a northeastward direction while also posi- tioning it further north, which results in eastward propagating cyclones being steered into Britain and Ireland. In Scandinavia, anti-cyclonic wave break- ing is observed much further north compared with WCE, helping to reinforce the strong blocking observed in the surface composites – this orientates the upper-level jet to the west of Scandinavia, helping to steer cyclones into this region. Overall, a clear connection is drawn between the precise location of anti-cyclonic wave breaking and cold air intrusion from the north and the po- sition and orientation of an intense upper-level jet; this, in turn, plays a cru- cial role in determining into which region a developing extreme storm will be steered. θPV2composites computed using model output data show a good, qualitative agreement with the reanalysis data, with the exception of a positive bias of between 10–15 K and 5–10 K most likely due to a lower tropopause height in the model.

(35)

6. Outlook

This thesis has focused on the identification and tracking of midlatitude cy- clones as well as the role planetary-scale waves play in the evolution of ex- treme windstorms over Western Europe.

As outlined in Section 4.1 of the thesis and in Papers I and II, the identi- fication and tracking of cyclones in reanalysis and model output data remains an extremely challenging topic. A number of approaches could be explored to improve the method introduced in Paper I, in particular to address the is- sue of the smooth tracking of a cyclone as it transitions from a single-centre system to a multi-centre one (and vice-versa). One approach to this issue is to consider the centre of a multi-centre cyclone as being the geographical point which minimises the distance to the cyclone’s boundary; this has the potential drawback, however, of locating the instantaneous location of the cyclone at a position unrelated to minima or maxima in the local MSLP and vorticity fields (areas typically considered as the cyclone’s centre). Another possible approach is to extend the algorithm to include the identification of vorticity maxima and to denote the centre of a cyclone as the nearest identified vorticity maximum in situations where there is more than one MSLP minimum (i.e., an MCC).

Paper II constitutes a first step in an ongoing collaborative effort to measure the similarities and differences of commonly used automated cyclone tracking algorithms. The next steps of this project include a comparative study of the variability in results across the different methods in identifying and tracking a set of selected historically notable windstorms over Europe from the past 30 years, and a number of studies focusing on particular themes such as cyclone merging and splitting.

Papers III and IV established a clear link between the evolution of the most extreme windstorms over Western Europe and planetary-scale waves; the dy- namical mechanisms underlying the observed wave breaking behaviour were not explored, however, and remain a topic for future research. One possible mechanism is the propagation of exceptionally high-amplitude Rossby waves from the Pacific to the Atlantic (Feldstein, 2003; Franzke et al., 2004; Strong

& Magnusdottir, 2008). Another possibility is the upscale cascade of energy from high-frequency synoptic scale disturbances through increased merging in the mid-Atlantic; the ability of the method presented in Paper I to identify cy- clone merging events could be used to generate statistics on merging frequency

(36)

in the build-up to the observed wave breaking events.

As outlined in Paper III, some of the most destructive storms of the past 30 years over Europe have had the remarkable feature of being separated by as little as 2–3 days. Our results may help to explain this by demonstrating that a particular set of large-scale conditions persist for longer-than-synoptic timescales, creating an environment which amplifies several pre-existing, smaller- scale perturbations that might have otherwise developed into more benign sys- tems. It is also possible that the very strong baroclinicity generated by the wave breaking events spawns its own small-scale instabilities. This remains a topic for future research.

(37)

Acknowledgements

I am eternally indebted to all the people who have helped to make my incredible five year PhD journey from Dublin to Stockholm possible.

First and foremost, I wish to express my deepest gratitude to my supervisor, Ro- drigo Caballero. Thank you for giving me the opportunity to study such a fascinating topic and for all your assistance throughout my PhD. It has been a privilege to work with you during this time and to witness your deep passion for research. Thank you for sharing so much of your time with me and for your patience in communicating your knowledge and insight.

In Ireland, I wish to thank Peter Lynch at University College Dublin for first spark- ing my interest in Meteorology. Your vast knowledge of the subject coupled with your exuberant teaching style made lectures a real joy. I want to thank Science Founda- tion Ireland for their financial support while I was in Ireland. I also want to thank all my CASL friends for the endless lunch discussions, Wednesday morning cake experi- ments and, of course, Thirsty Thursdays in O’Sheas. I want to particularly thank Colm and Will for providing so many moments of light relief and for the eventful experience that was EGU 2008. I also owe an enormous debt of gratitude to Jen, for all her help and support throughout my PhD.

The Swedish leg of my journey has been an unforgettable experience, all the more so for being so unexpected. I want to thank the Bert Bolin Climate Change Centre for funding me throughout this period, Gunilla Svensson for facilitating my transfer to MISU and Jonas Nycander for all this help during the transition.

To all the people in MISU who made it such a fun and stimulating environment to work in, thank you. First of all, I would like to thank my officemates Wing, Marie and Cian for putting up with me, particularly towards the end. Thank you for all the welcome conversational distractions, the extremely useful language lessons and for tolerating my occasionally overly-energetic typing.

To the Goodfellas lunch mafia of Laurent, Saeed, Cian and Mondheur, thank you for the always entertaining lunch discussions, our weekly escapades into the perilous world that is Söder nightlife and for being all-round great friends. I particularly want to thank Saeed for all his help and support and for the many unique phrases he in- troduced to the English language.I want to thank Laurent for the use of his superior French technology in the form of his elegant interpolation method which saved me considerable time towards the end and for introducing me to the pleasure of exploring Stockholm on two wheels.

To Peggy and honorary MISU-ite Matthias, thank you for all the Feuerzangen- bowle and for the memorable whiskey-soaked Christmas cookie baking marathon.

(38)

Thanks to Raza, for always having a positive and upbeat thing to say even when things weren’t going so well and for offering the consoling words of “don’t worry man, old is gold” upon learning that it was my 30th birthday. To the MISU newbies of Friederike, Georgia, Filippa and Henrik – good luck with your research and I look forward to seeing your progress on visits back to MISU. I would also like to thank Joaqim, Leon, Jonas Mortin, Cecilia, Marcus, Anna, Johannes, Jenny Lindvall, Anders and my great colleague Per Pemberton for their friendship and the many educating and illuminating conversations on the nuances of Swedish culture.

Last, but by no means least, I would like to thank my family. To my sisters, thank you for the mental health breaks that were my visits to Edinburgh and Galway – they came at the perfect time and helped to maintain my sanity. To my parents, thank you for all your love and support, particularly at the end when home provided a welcome refuge when making the final push towards the finish.

(39)

Bibliography

BENEDICT, J., LEE, S. & FELDSTEIN, S. (2004). Synoptic view of the north atlantic osciallation. J. Atmos. Sci., 61, 121–144. 24, 25

BERZ, G. (2005). Windstorm and storm surges in europe: losstrends and possible counter-actions from the viewpoint of an international reinsurer. Phil. Trans. R.

Soc. A., 363, 1431–1440. 11

BJERKNESS, J. & SOLBERG, H. (1922). Life cycles of cyclones and the polar front theory of atmospheric circulation. Geofys. Publ., 3, 1–18. 13

BLACKMON, M.L. (1976). A climatological spectral study of the 500 mb geopotential height of the northern hemisphere. J. Atmos. Sci., 33, 1607–1623. 28

BLENDER, R., FRAEDRICK, K. & LUNKEIT, F. (1997). Identification of cyclone track regimes in the north atlantic. Quart. J. Roy. Meteor. Soc., 123, 727–741. 28 BRETHERTON, F. (1966a). Baroclinic instability and the short wavelength cut-off in

terms of potential vorticity. Quart. J. Roy. Met. Soc., 92, 335–345. 16

BRETHERTON, F. (1966b). Critical layer instability in baroclinic flows. Quart. J. Roy.

Met. Soc., 92, 325–334. 16

BROWNING, K.A. (1985). Conceptual models of precipitation systems. Meteor. Mag., 114, 293–319. 14

BROWNING, K.A. (2004). The sting at the end of the tail: Damaging winds associated with extratropical cyclones. Q. J. R. Meteorol. Soc., 130, 375–399. 14

BROWNING, K.A. & ROBERTS, N.M. (1994). Structure of a frontal cyclone. Q. J. R.

Meteorol. Soc., 120, 1535–1557. 14

CARLSON, T.N. (1980). Airflow through midlatitude cyclones and the comma cloud pattern. Mon. Wea. Rev., 108, 1498–1509. 14

CHARNEY, J.G. (1947). The dynamics of long waves in a baroclinic westerly current.

J. Meteor., 4, 135–162. 13, 15

DONAT, M., LECKEBUSCH, G., PINTO, J. & ULBRICH, U. (2010). Examination of wind storms over central europe with respect to circulation weather types and nao phases. Int. J. Climatol., 30, 1289–1300. 24

(40)

EADY, E.T. (1949). Long waves and cyclone waves. Tellus, 1, 33–52. 13, 15

ERTEL, H. (1942). Ein neuer hydrodynamischer wirbelsatz. Meteor. Z., 59, 277–281.

16

FELDSTEIN, S. (2003). The dynamics of nao teleconnection pattern growth and decay.

Quart. J. Roy. Met. Soc., 129, 901–924. 35

FINK, A., POHLE, S., PINTO, J. & KNIPPERTZ, P. (2012). Diagnosing the influence of diabatic processes on the explosive deepening of extratropical cyclones. Geo.

Phys. Res. Lett., 39, doi: 10.1029/2012GL051025. 20

FRANZKE, C., LEE, S. & FELDSTEIN, S. (2004). Is the north atlantic oscillation a breaking wave? J. Atmos. Sci., 61, 145–160. 35

HANLEY, J. & CABALLERO, R. (2012). Objective identification and tracking of multi-centre cyclones in the era-interim re-analysis dataset. Q. J. R. Meteorol. Soc., 138, 612–625. 28, 29

HODGES, K., HOSKINS, B.J., BOYLE, J. & THORNCROFT, C. (2003). A compari- son of recent reanalysis data sets using objective feature tracking: Storm tracks and tropical easterly waves. Mon. Wea. Rev., 131, 2012–2037. 28

HODGES, K.I. (1994). A general method for tracking analysis and its applications to meteorological data. Mon. Wea. Rev., 122, 2573–2586. 28

HOLTON, J.R. (2004). An Introduction to Dynamic Meteorology.. Academic Press, fourth edition. 13

HOSKINS, B., MCINTYRE, M.E. & ROBERTSON, A.W. (1985). On the use and significance of isentropic potential vorticity maps. Q. J. R. Meteorol. Soc., 111, 877–946. 16

HURRELL, J. (1995). Decadal trends in the north atlantic oscillation: regional tem- peratures and precipitation. Sci., 269, 676–679. 23

HURRELL, J., KUSHNIR, Y., OTTERSEN, G. & VISBECK, M. (2003). The north atlantic oscillation: Climate significance and environmental impact. Geophysical Monograph 134, American Geophysical Union. 23

JONES, P., HULME, M. & BRIFFA, K. (1993). A comparison of lamb circulation types with an objective classification scheme. Int. J. Clim., 13, 655–663. 24 KUNZ, T., FRAEDRICH, K. & LUNKEIT, F. (2009). Synoptic scale wave breaking

and its potential to drive nao-like circulation dipoles: A simplified gcm approach.

Q. J. R. Meteorol. Soc., 135, 1–19. 24

LAU, N.C. (1988). Variability of the observed midlatitude storm tracks in relation to low-frequency changes in the circulation pattern. J. Atmos. Sci., 45, 2718–2743. 28

(41)

LECKEBUSCH, G., WEIMAR, A., REYERS, M. & SPETH, P. (2008). Extreme wind storms over europe in present and future climate: a cluster analysis approach. Met.

Zeit., 17, 67–82. 24

MARTIN, J. (2006). Mid-Latitude Atmospheric Dynamics: A First Course.. John Wi- ley & Sons, Ltd. 19

MOORE, R.W. & MONTGOMERY, M.T. (2005). Analysis of an idealized, three- dimensional diabatic rossby vortex: A coherent structure of the most baroclinic atmosphere. J. Atmos. Sci., 62, 2703–2725. 18

MURRAY, R.J. & SIMMONDS, I. (1991). A numerical scheme for tracking cyclone centres from digital data, part 1: development and operation of the scheme. Aust.

Meteor. Mag., 39, 155–156. 28

NEU, U., AKPEROV, M., BELLENBAUM, N., BENESTAD, R., BLENDER, R., CABALLERO, R., COCOZZA, A., DACRE, H., FENG, Y., FRAEDRICH, K., GRIEGER, J., GULEV, S., HANLEY, J., HEWSON, T., INATSU, M., KEAY, K., KEW, S., KINDEM, I., LECKEBUSCH, G., LIBERATO, M., LIONELLO, P., MOKHOV, I., PINTO, J., RAIBLE, C., REALE, M., RUDEVA, I., SCHUSTER, M., SIMMONDS, I., SINCLAIR, M., SPRENGER, M., TILININA, N., TRIGO, I., UL-

BRICH, S., ULBRICH, U., WANG, X. & WERNLI, H. (2012). Imilast a commu- nity effort to intercompare extratropical cyclone detection and tracking algorithms:

assessing method-related uncertainties. Bull. Am. Met. Soc., doi 10.1175/BAMS–

D–11–00154.1 (in press). 27, 28

PARKER, D.J. & THORPE, A.J. (1995). Conditional convective heating in a baro- clinic atmosphere: a model of convective frontogenesis. J. Atmos. Sci., 52, 1699–

1711. 18

PELLY, J. & HOSKINS, B. (2003). A new perspective on blocking. J. Atmos. Sci., 60, 743–755. 24

PINTO, J.G., SPANGEHL, T., ULBRICH, U. & PETER, S. (2005). Sensivities of a cy- clone detection and tracking algorithm: individual tracks and climatology. Meteor.

Z., 14, 823–838. 28

PINTO, J.G., ZACHARIAS, S., FINK, A.H., LECKEBUSCH, G.C. & ULBRICH, U.

(2009). Factors contributing to the development of extreme north atlantic cyclones and their relationship with the nao. Clim. Dyn., 32, 711–737. 24

RAIBLE, C. (2007). On the relation between extremes of midlatitude cyclones and the atmospheric circulation using era40. Geo. Phys. Res. Lett., 34, L07703. 24 RAIBLE, C., DELLA-MARTA, P.M., SCHEWIERZ, C., WERNLI, H. & BLENDER,

R. (2008). Northern hemipshere extratropical cyclones: A comparison of detection and tracking methods and different reanalysis. Mon. Wea. Rev., 136, 880–897. 27, 28

(42)

RIVIÈRE, G. & JOLY, A. (2006). Role of the low-frequency deformation field on the explosive growth of extratropical cyclones at the jet exit. part ii: Baroclinic critical region. J. Atmos. Sci., 63, 1982–1995. 25, 26

RIVIÈRE, G. & ORLANSKI, I. (2007). Characteristics of the atlantic storm-track eddy activity and its relation with the north atlantic oscillation. J. Atmos. Sci., 64, 241–

266. 24

ROSSBY, C.G. (1940). Planetary flow patterns in the atmosphere. Q. J. R. Meteorol.

Soc., 66, 68–87. 16

ROSSBY, C.G.E.A. (1939). Relation between variations in the intensity of the zonal circulation of the atmosphere and the displacements of the semi-permanent centers of action. J. Marine. Res., 2, 38–55. 16

SCHULTZ, D.M. (2001). Reexamining the cold conveyor belt. Mon. Wea. Rev., 129, 2205–2225. 14, 15

SCHULTZ, D.M., KEYSER, D. & BOSART, L.F. (1998). The effect of large-scale flow on low-level frontal structure and evolution of midlatitude cyclones. Mon. Wea.

Rev., 126, 1767–1791. 13, 14

SERREZE, M., CARSE, F., BARRY, R. & ROGERS, J. (1997). Icelandic low cyclone activity: climatological features, linkages with the nao and relationships with recent changes elsewhere in the northern hemisphere circulation. Jour. Clim., 10, 453–464.

24

SHAPIRO, M.A. & KEYSER, D. (1990). Fronts, jet streams and the tropopause. Amer- ican Meteorological Society, Extratropical cyclones: the Eric Palmen memorial volume, pages 167–191. 13

SINCLAIR, M.R. (1994). An objective cyclone climatology for the southern hemi- sphere. Mon. Wea. Rev., 122, 2239–2256. 28

STRONG, C. & MAGNUSDOTTIR, G. (2008). How rossby wave breaking over the pacific forces the north atlantic oscillation. Geo. Phys. Res. Lett., 35, L10706, doi:10.1029/2008GL033578. 35

THORNCROFT, C.D., HOSKINS, B.J. & MCINTYRE, M.E. (1993). Two paradigms of baroclinic-wave life-cycle behaviour. Q. J. R. Meteorol. Soc., 119, 17–55. 17 ULBRICH, U., FINK, A., KLAWA, M. & PINTO, J. (2001). Three extreme storms

over europe in december 1999. Wea., 56, 70–80. 25

WALKER, G. (1924). Correlations in seasonal variations of weather ix. Memoirs of India Meteorological Department, 24, 275–332. 23

WERNLI, H. & C., S. (2006). Surface cyclones in the era40 data set (1958-2001). part 1: novel identification method and global climatology. J. Atmos. Sci., 63, 2486–

2507. 28

(43)

WERNLI, H., DIRREN, S., LINIGER, M. & ZILLIG, M. (2002). Dynamical aspects of the life-cycle of the winter storm ‘lothar’. Quart. J. Roy. Meteor. Soc., 128, 405–

429. 19, 20

WOOLLINGS, T., HOSKINS, B., BLACKBURN, M. & BERRISFORD, P. (2008). A new rossby wave-breaking interpretation of the north atlantic oscillation. J. Atmos.

Sci., 65, 609–626. 24

(44)

References

Related documents

46 Konkreta exempel skulle kunna vara främjandeinsatser för affärsänglar/affärsängelnätverk, skapa arenor där aktörer från utbuds- och efterfrågesidan kan mötas eller

This is the concluding international report of IPREG (The Innovative Policy Research for Economic Growth) The IPREG, project deals with two main issues: first the estimation of

I dag uppgår denna del av befolkningen till knappt 4 200 personer och år 2030 beräknas det finnas drygt 4 800 personer i Gällivare kommun som är 65 år eller äldre i

Detta projekt utvecklar policymixen för strategin Smart industri (Näringsdepartementet, 2016a). En av anledningarna till en stark avgränsning är att analysen bygger på djupa

Av 2012 års danska handlingsplan för Indien framgår att det finns en ambition att även ingå ett samförståndsavtal avseende högre utbildning vilket skulle främja utbildnings-,

Det är detta som Tyskland så effektivt lyckats med genom högnivåmöten där samarbeten inom forskning och innovation leder till förbättrade möjligheter för tyska företag i

As a consequence, shown Figure 28 below, some countries are specialized in Energy crops (Germany, Austria), others in agriculture waste (Luxembourg, Belgium, Ireland, Italy) or

Industrial Emissions Directive, supplemented by horizontal legislation (e.g., Framework Directives on Waste and Water, Emissions Trading System, etc) and guidance on operating