1. Introduction
The European winter of 2005/06 was notable in that it bucked the prevailing decadal trend toward warmer winter conditions, and indeed there were sustained anomalously cold temperatures and high snow accumulation over most of the continent (Pinto et al. 2007). Features of the larger-scale flow that influenced the occurrence of anomalous European winter conditions include 1) the prevailing pattern of interannual climate variability in the Euro–Atlantic sector, 2) the day-to-day sequence of weather systems in the sector, and 3) the interplay between these climate and weather features.
In relation to climate variability, a range of atmospheric analysis fields have pointed to the existence of statistically preferred spatial patterns in the Euro–Atlantic sector (e.g., Barnston and Livezey 1987; Rogers 1990). The most dominant pattern is the so-called North Atlantic Oscillation (NAO) with its two centers of action located, respectively, over the Icelandic and the Azores regions. Downstream over Europe the two phases of the NAO are often accompanied by markedly different winter conditions. Large positive (negative) values of an NAO index equate strong (weak) latitudinal pressure gradient between the two centers of action, connote to stronger (weaker) time-mean westerly flow at the surface toward the European mainland, and tend to be associated with relatively milder (colder) conditions over central Europe (see e.g., Hurrell 1995; Hurrell et al. 2001, 2003). These studies have been supplemented by considerations, for example, of the linkage to and coupling with the ocean state (Hurrell et al. 2006), the attendant finer details of the linkage to European weather and precipitation (Trigo et al. 2002; Beranova and Hadan 2008), and the possible impact of increased greenhouse gases upon the NAO’s link to European weather (e.g., Scaife et al. 2008; Kuzmina et al. 2005). The “NAO–European weather” linkage has been deemed sufficiently strong for many statistical-empirical seasonal prediction procedures to be underpinned by forecasts of the NAO itself (e.g., Hurrell et al. 2003; Folland et al. 2006; Saunders and Lea 2006).
Other statistically based patterns of climate variability for the Euro–Atlantic sector during winter account for significantly less of the interannual variance, but do posses centers of action over or in the immediate vicinity of the continent. One particular pattern has a primary center located in the neighborhood of the United Kingdom (Barnston and Livezey 1987; Rogers 1990) and it has been related to blocking (Pavan et al. 2000; Scherrer et al. 2006) and to both upper-tropospheric and precipitation anomalies over Europe (Massacand and Davies 2001b).
In terms of the influence of weather systems upon European winter conditions, a key factor is the nature of the prevailing weather type(s). One significant type corresponds to the occurrence of a series of transient low pressure systems traveling in rapid succession onto the European mainland—the so-called cyclonic storm bands—this type is associated with comparatively milder, wetter conditions. A counter type is the occurrence of a few longer-lasting, quasi-stationary blocking events over or in the immediate vicinity of the mainland. Such events, in addition to being relatively long lasting (several days up to weeks), are anticyclonic in sign, equivalent barotropic in structure, and intense enough to disrupt the prevailing extratropical westerly flow. A block located over or immediately upstream of Europe in winter is almost invariably accompanied by persistent dry conditions at its core and by strong cold-air advection on its eastern fringe (e.g., Trigo et al. 2004). Thus, a temporal succession of collocated blocks can imprint markedly upon a single winter’s mean climate. Note however that, depending on the blocking index used, it is the central North Atlantic Ocean (e.g., Dole and Gordon 1983; Sausen et al. 1995; Croci-Maspoli et al. 2007b) as opposed to western Europe (e.g., Lejenäs and Økland 1983; Tibaldi and Molteni 1990; Pelly and Hoskins 2003) that is the most preferred location for blocking in the Euro–Atlantic sector.
There is some limited evidence linking the prevailing patterns of climate variability with these types of weather systems. The NAO’s positive (negative) phase is accompanied by enhanced (reduced) values of the time-mean baroclinicity in the mid-Atlantic that is conducive to/inhibits cyclogenesis farther downstream, and hence has been traditionally linked to the storm band (blocking) weather type. This inference is supported by the fact that both the NAO’s negative phase (Massacand and Davies 2001a) and blocks (Schwierz et al. 2004) are associated with deep equivalent-barotropic anticyclonic structures and negative potential vorticity (PV) anomalies at tropopause levels. Moreover, there is a notable increase in block occurrence over the western and central North Atlantic during the negative NAO phase (e.g., Shabbar et al. 2001; Barriopedro et al. 2006; Croci-Maspoli et al. 2007a). A caveat to the foregoing is that a correlation between NAO and blocking in the central Atlantic does not necessarily translate to distinctive weather over Europe. On the contrary, patterns of variability with center(s) near or over Europe could link more directly with the continent’s climate and the two phases of the aforementioned variability pattern with its major center in the neighborhood of the United Kingdom are consistent, respectively, with the occurrence of “storm band” and “blocking” types of synoptic systems over the same region (e.g., Massacand and Davies 2001b).
In this study the focus is on the nature, dynamics, distinctiveness, and interrelationships of the large-scale patterns of climate variability and the dominant weather systems that prevailed during the anomalous European winter (December–February) of 2005/06. The paper is structured as follows. Section 2 sets out the data and methodology deployed in the study. In section 3, a conventional principal component analysis is used to identify the dominant statistical patterns of climate variability during the 2005/06 winter, and consideration given to some of the salient winter-mean features. In section 4, the focus is on the nature and instigation of the winter’s major synoptic weather systems. Thereafter in section 5, a hypothesis is proposed that is prompted by the previous analyses and designed to explore the linkage between the time-mean climate conditions and the major day-to-day weather systems. Its validity is examined using a series of numerical simulations conducted with a state-of-the-art limited-area model. In section 6, the study’s results are overviewed and general comments made on their relation to patterns of climate variability, seasonal prediction, and trends in interannual climate variability.
2. Data and methodology
a. Data
The diagnostic analyses and numerical computations for the winter 2005/06 fields [December–February (DJF)] are performed using the analysis fields of the European Centre for Medium-Range Weather Forecasts (ECMWF). For the time period 1958–2002 the 40-yr ECMWF Re-Analysis (ERA-40) dataset (Uppala et al. 2005) was used for anomaly calculations. The Centre’s spectrally based fields are interpolated horizontally onto a 1° × 1° grid at the full 6-hourly temporal resolution. In addition to the atmospheric variables we also extract the ECMWF’s 2-m temperature (T2M) and the sea surface temperature (SST). Note that the 1° × 1° grid resolution is retained throughout (i.e., for both the ERA-40 reanalysis dataset pertaining to the time period from 1958 to 2002 and for the winter 2005/06), although the spectral resolution of the ECMWF was increased from T511L60 to T799L91 on 1 February 2006.
In addition Meteosat infrared (IR) satellite images for the Euro–Atlantic region are used to characterize a single blocking event, and use is also made of the daily NAO index (standardized by one standard deviation) of the Climate Prediction Center (CPC) to characterize the temporal evolution of the NAO during the winter.
b. Climate variability and climate state
The first of this study’s three methodological components is the consideration of the patterns of climate variability and climate state during the 2005/06 winter. Attention is directed to the mid- and upper-tropospheric flow and surface temperature. A computation of the statistical patterns of 500-hPa variability is performed for each of the three winter months using a standard principal component (PC) analysis. The aim is to examine which pattern(s) prevailed or dominated during this particular winter. To link the pattern of interannual variability and the prevailing weather systems consideration is given to the mean patterns and anomalies of the upper-tropospheric PV and the surface temperature during the 2005/06 winter.
c. Blocking events
The study’s second methodological component relates to block occurrence and the PV dynamics of blocking events during the 2005/06 winter. Herein blocks are defined as negative anomalies of PV in the upper troposphere that persist for longer than 5 days. They are identified using an index (Schwierz et al. 2004) based upon the departure of the two-dimensional field of the vertically averaged PV (VAPV) in the 150–500-hPa layer from its climatological mean (hereinafter referred to as the VAPV index). The dataset is again the ECMWF analysis fields and the block’s location, spatial extent, and track are determined at 6-hourly intervals [see Croci-Maspoli et al. (2007b) for further details of the scheme].
From a PV perspective there are two features of a block that are particularly pertinent to our study. First, a block’s occurrence requires that its core be a region of anomalously low PV, and our identification scheme is predicated upon this feature. Second, the origin of this low PV within the block can be attributed to either its adiabatic advection on isentropic surfaces (Nakamura 1994; Nakamura et al. 1997; De Pondeca et al. 1998; Michelangeli and Vautard 1998), and/or its cloud-diabatic generation and across-isentropic advection associated with deep ascent of warm moist air ahead of a surface cold front (Wernli and Davies 1997; Schwierz 2001). During block inception the PV advection and diabatic generation usually occur within a large-scale flow pattern characterized on the basis of theoretical (Nakamura et al. 1997; De Pondeca et al. 1998), empirical (Pelly and Hoskins 2003), and climatological (Altenhoff et al. 2008) considerations as a breaking Rossby wave.
Herein a two-step procedure is adopted to ascertain the existence and assess the contribution of moist ascent to the development of a block’s core of low PV air. Infrared Meteorological Satellite (Meteosat) images are inspected to detect the presence of cloud activity at the incipient and maturing block’s periphery. Thereafter the “earlier-in-time” location of air parcels within a block’s initial core of low PV in the upper troposphere (500–150-hPa geopotential height layer) is examined by computing 96-h backward trajectories (following Wernli and Davies 1997).
d. Model simulations
The third methodological component relates to the diagnosis and quantification of the aforementioned influence upon block formation of the deep ascent of prefrontal warm and moist low-level air. Such ascent would be favored by suitable land and sea surface temperature configurations, comparatively weak stratification aloft, and attendant surface-frontal and upper-level troughlike structures. Kung et al. (1990) pointed to the need for adequate representation of SST to successfully simulate of blocking events in a forecast, and in a related context Cassou et al. (2004) investigated the role of SST anomalies in differing atmospheric regimes.
To explore the sensitivity of the blocks during the 2005/06 winter to the thermodynamic state of the low-level air a series of numerical model simulations are conducted for the realized blocking events. The simulations fall into four generic types: 1) a control run with the ECMWF analyses fields with the analyzed initial and the time-dependent lateral boundary fields; and further integrations that incorporate, respectively, 2) reduced and enhanced values of the North Atlantic SST for the model levels below 500 hPa, 3) reduced and enhanced values of the initial and lateral boundary low-level relative humidity, and 4) the combination of both the SST and the humidity changes.
The simulations are undertaken with the hydrostatic Climate High Resolution Model (CHRM). It is an extension of the European Model (EM) of the Deutscher Wetterdienst (DWD) that was originally developed for numerical weather prediction (Majewski 1991). It operates with a regular horizontal grid and a hybrid sigma level coordinate system, and has been tested and used extensively at various horizontal resolutions (14–50 km). Its usefulness is well attested for both climate studies (e.g., Vidale et al. 2003) and short-range mesoscale simulations (e.g., Gheusi and Davies 2004).
Here the CHRM is deployed with a nonrotated grid at 0.5° × 0.5° horizontal resolution and with 40 vertical levels. The model domain is chosen to extend from 25°–85°N, 120°W–50°E, and hence it encompasses the Euro–Atlantic sector. The initial conditions and time-dependent lateral boundary conditions are taken from ECMWF analysis data (interpolated to the CHRM grid resolution). For each blocking event the integration period is 120 h (5 days) and the simulation is initiated some days preceding the block onset.
3. Characteristics of the European winter 2005/06 climate
It was noted earlier that the European winter of 2005/06 bucked the recent decadal trend for warmer winters with anomalously cold temperatures across most of mainland Europe. In particular the pre-Alpine regions of Germany and Austria experienced an extended period of almost uninterrupted subzero temperatures and a series of heavy snowfalls. This in turn led to very high snow accumulation depths in late winter and early March, and subsequent severe flooding following the onset of warm temperatures in mid-March (Pinto et al. 2007). In this section we set out the character of the patterns of climate variability and the mean winter climatic conditions for the 2005/06 European winter.
a. Patterns of climate variability
Figure 1 displays the three leading EOF patterns over the Euro–Atlantic region (20°–80°N, 80°W–60°E) for the 500-hPa field (Z500) derived using the individual monthly (DJF) mean fields from the ERA-40 analysis plus the three months of the 2005/06 winter. The patterns correspond to those mentioned in the introduction and replicated patterns computed in earlier studies using different time periods and a variety of fields (e.g., SLP, 700 and 500 hPa, and PV on tropopause bisecting surfaces). The three patterns displayed here correspond respectively to the NAO, a Euro–Atlantic wave train, and a major center located over the United Kingdom and embedded within a signal of the opposite sign.
Figures 2a,b show the time trace of the three PCs and the corresponding standardized PC density distribution, respectively. The amplitude of the signal associated with the two leading EOFs was not exceptional during the 2005/06 winter. The weak amplitude for the NAO-like component in January and February contrasts with the customary positive correlation between the NAO index and the European temperature anomaly (e.g., Hurrell 1996; Hurrell et al. 2001; Casty et al. 2005). Thus, the nature and extent of the NAO’s influence upon the character of the 2005/06 European winter is not transparent. In contrast Fig. 2 also indicates that the third “blocklike” mode possessed large negative values for each individual winter month. Indeed the December and January values were larger than one standard deviation, and the January signal was the second largest of the entire time period. This hints at a possibly seminal role of blocking during the 2005/06 winter and concomitantly prompts questions related to the nature of the mode.
b. Salient winter-mean fields
Here we address two questions. First, to what extent is the preeminence of the blocklike EOF evident in the time-mean fields and their anomalies; and second, are there other distinctive features of the time-mean surface temperature fields upstream of the cold anomaly over Europe?
It was noted earlier that both the NAO and blocking events are associated with significant signatures of the PV pattern at tropopause elevations and it is therefore instructive to examine the winter-mean PV distribution and accompanying anomaly pattern on a tropopause-level isentropic surface. Figure 3 shows these fields on the 315-K surface in the Euro–Atlantic sector. For the 2005/06 winter the standard climatological pattern of a strong latitudinal gradient of PV with a major trough (i.e., high PV values) over the eastern Atlantic has in addition two superimposed features. First, there are modest negative (positive) PV anomalies located to the north (south), and this connotes a reduction in the strength of the time-mean jet and less positive values for the NAO index. Second, in the eastern Atlantic there is a pronounced southwest–northeast-aligned anomalous ridge with accompanying strong negative PV anomalies between Greenland and Scandinavia and a counterpart positive PV anomaly extending over central Europe and North Africa. This is consistent with a blocked flow regime prevailing off the continent and advection of cold northerly air onto the European mainland. Comparison with the spatial patterns of climate variability (see Fig. 1) suggests that the overall PV pattern with negative (positive) anomalies located to the north (south) bears comparison with an NAO positive anomaly signature, whereas the salient spatially refined and large-amplitude signatures over the eastern Atlantic and western Europe conform closely to those of the blocklike EOF.
Now consider the pattern of the surface temperature anomalies. The winter-mean daytime 2-m temperature anomaly pattern for the Euro–Atlantic sector (Fig. 4a) shows that most of continental Europe, excluding Scandinavia and Scotland, experienced negative temperature anomalies of between 0.5° and 3°C. In addition parts of Russia (not shown) registered anomalies in excess of −10°C. In contrast, the North American continent and the polar region experienced pronounced positive temperature anomalies of more than 10°C. Indeed the warm surface temperatures over the conterminous United States for the year of 2006 set a record and has been attributed primarily to human influence (Hoerling et al. 2007).
The corresponding SST anomaly pattern (Fig. 4b) serves to emphasize that almost the entire North Atlantic experienced positive temperature anomalies, and in the western margins the values amount to 3°–4°C. This essentially monopolar pattern is distinct from the SST anomaly patterns that have been shown to correlate with the NAO signature (see, e.g., Bresch and Davies 2000) and European weather (see, e.g., Losada et al. 2007).
4. Characteristics of the European winter 2005/06 weather systems
a. Temporal evolution
To help to forge a link between the realized patterns of climate variability and the occurrence weather systems during the 2005/06 winter, we display in Fig. 5 the daily time traces of 1) the NAO index (the CPC standardized values), 2) a measure of the VAPV anomaly in the upper troposphere within a domain (50°–70°N, 30°W–30°E) encompassing both to the center of the blocklike mode of climate variability and the region of negative mean PV anomaly values on the 315-K isentropic surface (see previous section), 3) the surface 2-m temperatures over central Europe (38°–53°N, 5°W–25°E), and 4) the timing and duration of blocks that persisted for more than 5 days.
The NAO trace (upper curve) attains significantly negative values only during the first and last few weeks of the “3 month” winter period. Comparison of the NAO trace with the corresponding trace of the European temperature anomalies (bottom curve) indicates that the conventional linkage of a negative NAO phase with negative European temperature anomalies is not discernible for most of the winter. Indeed, during January a strong positive NAO index (monthly mean of +0.65) was accompanied by significant negative temperature anomalies (monthly mean of −1.26°C) over central Europe, and Fig. 6 shows that the standardized difference between these two variables amounted to one of its largest values during the last 40 years.
The VAPV trace (second curve in Fig. 5) shows the prevalence of distinctly negative values during December and January. Moreover it shows a notable correspondence with the trace for the European surface temperature for most of the three-month winter period. Likewise comparison of the VAPV trace with the bars signifying the timing and duration of the block (bottom of figure) indicates that VAPV values of less than about −0.25 coincide with the imminent onset or continued existence of blocks.
The five blocks recorded in Fig. 5 formed over the North Atlantic, propagated slowly eastward toward the Icelandic–Scandinavian region, and in the mean were located farther northeastward than their customary climatologically favored location over the central North Atlantic (Croci-Maspoli et al. 2007b). Block occurrence corresponded closely with the interruption of the flow of warmer oceanic air onto the European mainland and its replacement, on the block’s eastern flank, with the sustained transport of very cold continental air from the polar region or Siberia. Heavy snowfalls occurred over the pre-Alpine region of central Europe in the second half of December, 2–3 January, 17–18 January, and 7–9 February (Pinto et al. 2007) coinciding with periods when blocking prevailed to the west.
Intercomparison of the traces in Fig. 5 points to a mismatch between the NAO’s signature and the realized temperature conditions over Europe for the 2005/06 winter. In effect the winter’s sequence of weather systems (i.e., blocks) were not primarily associated with the phase of the NAO, but rather associated with the repeated occurrence of upper-level anomalies of low PV and the blocklike pattern of climate variability.
b. Key factor for block instigation: A case study
One of the winter’s major blocking events had its inception on 24 December 2005 over the North Atlantic. It intensified rapidly, propagated slowly toward Scandinavia, and on the basis of the VAPV index ceased to register as a block on 3 January 2006.
Figure 7 displays aspects of the block’s evolution at the upper level. Each panel comprises a superimposition of the Meteosat’s IR image upon a display of the PV contours on the 315-K isentropic surface. Note that the 2-PVU contour is indicated in bold and that the dark blue shading of the image signifies warmer temperatures (no clouds) whereas the white–blue shading is representative of cold temperatures (high clouds). The figure illustrates the sequence of stages in the block’s life cycle. Initially (Fig. 7a) the incipient block is merely a small poleward undulation of the 2-PVU isentrope in the mid-Atlantic, but there is nevertheless evidence of strong cloud activity on its western flank (see the western edge of the IR image). In the subsequent 24 h the weak undulation has evolved to a notable block characterized by a large region of low PV (Fig. 7b), and again a meridionally elongated cloud band is evident on its upstream flank. The latter feature remains present as the block reaches maturity over the following 48 h (Figs. 7c,d).
As noted earlier low PV in the upper-tropospheric core of a deep block can be attributed to either isentropic advection of low PV air into the region and/or diabatically related injection of air with low PV from the lower troposphere. To examine the source of the low PV air present at the block’s inception its Lagrangian time history is computed. Figure 8 shows 696 backward trajectories emanating from the block’s core (between 500 and 150 hPa) at 1200 UTC 24 December 2005 (here inception is defined as the first day of the block detection based upon the VAPV index). On noting that the color scheme signifies the height of the air parcels at the corresponding time/location, it is evident that
the block ingests air from two distinct source locations—one from over the North American continent (N) and the other from the east Pacific (P);
the two airstreams (N and P) originate from distinct different vertical regions. Airstream N comes from very low levels (>750 hPa), but ascends rapidly to higher levels (<500 hPa) as it approaches the location of the block, whereas airstream P occupies a near-tropopause layer (400–150 hPa) throughout.
It is also pertinent to note that airstream N first tracks over the relatively warm North American continent (cf. Fig. 4a) and then over the anomalous warm western North Atlantic (cf. Fig. 4b). Its subsequent path encroaches upon and enters an active prefrontal zone within which it ascends and experiences the effects of diabatic heating.
c. Key factor for block instigation: All winter 2005/06 blocks
Inspection of the IR images of each of the five 2005/06 winter blocks indicates that each one was accompanied by deep convection on its upstream side (not shown). Here an indication of the diabatic contribution to block formation is provided by computing for each block of air parcels the 5-day backward trajectories of air parcels located at the block’s core at inception, recording the accompanying absolute humidity and diabatic heating rate, and compositing the results (Fig. 9).
The ensemble Lagrangian time trace for the absolute humidity (Fig. 9a) indicates a slight moistening in the 2–5 days prior to onset and thereafter a rapid drying out as the onset day approaches. This development is mirrored by the trace of the diabatic heating rate culminating (Fig. 9b) in a significant diabatic heating rate [order of 2°C (6 h)−1] in the drying-out phase. This composite result is in harmony with that discussed earlier for the single event, and indicates that diabatic heating is an integral feature of block formation.
5. Numerical simulation studies
a. Simulation strategy
This section is motivated by and predicated upon the tentative hypothesis that during the 2005/06 winter a causal chain links the weather over central Europe with the successive occurrence of blocks slightly upstream, the contribution of diabatic heating to the initiation of those blocks, and the warm surface anomalies even farther upstream over North America and the North Atlantic. This causal chain is a counterpart of the occurrence of distinctive PV streamers over western Europe (Massacand et al. 2001).
The rudimentary rationale prompting the hypothesis is that the warm surface conditions upstream can serve to enhance the heat and moisture content of the overlying atmosphere, and this in turn increases both the diabatic heating effects associated with fronts and cyclones in the mid-Atlantic and the concomitant injection of low PV air into the upper troposphere. The latter is then viewed as integral to the formation or enhancement of blocks in the eastern Atlantic.
To explore this hypothesis, a series of simulations are undertaken for each of the five blocking events using the CHRM in the configuration described earlier. For each blocking event a multiple set of 5-day simulations are performed each commencing 1.5 days prior to the observed block onset. Along with the 1) control run conducted with the initial and time-dependent lateral boundary conditions taken from the ECMWF analysis fields, integrations are undertaken with the following settings: 2) modification of the entire North Atlantic SST by a specified amount in the range (from −3° to +3°C) while keeping all other specified fields unchanged, 3) modifications in the initial and time dependent lateral boundary relative humidity by a specified percentage in the range (10%–80%) while keeping all other specified fields unchanged, and 4) some combinations of the two previous simulations.
The spatial extent of the domain and the use of the analyzed (flow and temperature) time-dependent boundary flow conditions provide the limited-area modeling system with a degree of pseudopredictability. They ensure that the gross features of the realized synoptic development are reproduced in the simulations while allowing the system to capture the net effects of the imposed changes upon the finer features of that development. The runs conducted with differing stipulated changes provide an indication of the sensitivity of the flow response to the SST and humidity changes.
To summarize the results of the multiple simulations we proceed as follows: first select one set of three simulations; second discuss the results of these selected simulations for both one particular blocking event and for the composite of the five blocking events; and third, note briefly the relative character of the results from the other sets of simulations.
The selected set of simulations discussed in this study corresponds to 1) the control run (CTRL), 2) an SST reduction of −3°C (referred to as SSTM3), 3) a humidity reduction of 10% (referred to as HUM10), and 4) the combination of these two simulations (referred to SSTM3HUM10). This selection is influenced by the anomalous conditions that prevailed in the western Atlantic and the contiguous continent during the 2005/06 winter, and hence they are simulations that correspond more closely in terms of western Atlantic SST to the multiyear winter-mean climatological conditions.
b. Simulation of the block of late December 2005
Figure 10 shows the PV field on the 315-K isentropic surface at 0600 UTC 26 December 2005 for a simulation initiated 196 h earlier at 0000 UTC 21 December 2005. The panels show, respectively, the fields for each of the four mentioned settings (CTRL, SSTM3, HUM10, and SSTM3HUM10) of the different CHRM simulation scenarios.
The CTRL simulation (Fig. 10a) replicates the observed block’s location and intensity as captured in the ECMWF analysis (not shown). The SSTM3 simulation (Fig. 10b) shows a block with reduced spatial extent and weaker in amplitude (i.e., higher PV values), while for the HUM10 simulation (Fig. 10c) the weakening of the block is not so prominent. Finally in the combined SSTM3HUM10 simulation (Fig. 10d) no block is present and the blocking region demarked in the CTRL run (PV values of around 0–2 PVU) is replaced with values generally greater than 2 PVU. In effect this simulation suggests that the combination of the SST and relative humidity distribution plays a potent role in the formation (or nonformation) of the block.
c. Composite of simulations
To overview the influence of the SST and the relative humidity for the five blocks of the 2005/06 winter we present a composite of the PV values within the domain of the CTRL during its evolution. In effect, the mean PV distribution for the CTRL simulation is calculated within the blocking area for each event and then averaged over the five events, and thereafter the calculation is repeated for the three different simulation settings.
The results are shown in Fig. 11 and display the time evolution on the 315-K surface of the composite mean PV within the block for 3.5 days from block onset. In the CTRL simulation a significant drop (order of 2 PVU) is observed in the first day after initiation and this is in line with the steady intensification of the blocks during the onset phase, and after two days there is a more gradual weakening associated in part with the different life times of the individual blocking events.
In contrast the simulations with a reduced SST or a reduced humidity yield less of a PV reduction particularly during the onset phase of the CTRL block. For the composite it is the humidity, as opposed to the SST, reduction that exhibits the largest impact (i.e., highest values of the PV field). As for the single case it is the combined scenario that shows the strongest effect upon the mean PV distribution in comparison with the CTRL run. For the combined run the PV values remain in the range 1.5–2 PVU.
d. Overview comments
The simulations for the selected settings of a 3°C reduction in SST, a reduction to 10% of the relative humidity, and their co-reduction suggest that block formation during the 2005/06 winter was sensitively dependent upon the values of the low-level thermodynamic fields of the SST and humidity. This assertion is further supported by simulations conducted with SST modifications in the range (from −3° to +3°C), humidity changes in the range (10%–80%), plus simulations with covariations of these externally specified fields. The response to the amplitude of the imposed modifications is essentially incremental in nature. For example, a modification of only −1°C in SST has a slightly discernible effect in comparison with the result recorded earlier, while a modification of +3°C in SST results in a larger scale to the resulting block. Our limited integration time, however, does not allow us to draw inferences regarding possible changes in block duration.
In effect, the simulations can provide an overall indication of the cumulative impact of the SST and relative humidity changes upon the diabatic processes and block generation. Limitations to the simulation strategy adopted are that the SST reduction within the model’s limited domain is imposed uniformly across the oceanic basin, and this forestalls consideration of a richer spatial structure to the SST anomaly in the extratropical North Atlantic and the possible impact of SST forcing from outside the model domain (e.g., the tropical North Atlantic or an ENSO-like SST signature from the tropical Pacific).
6. Further remarks
In this study we explored the nature of the linkage between the patterns of climate variability and the type of weather systems that prevailed during the anomalous 2005/06 winter in Europe. It was shown that the most significant pattern of variability during this particular season was not the NAO, but rather one that possessed a “blocklike” spatial structure. The dominance of this pattern was paralleled by the occurrence during the winter of five significant blocking events over the eastern Atlantic/western Europe. Diagnostic analysis suggests that cloud-diabatic effects on the western flank of the blocks played a role in block development. In addition, it was adduced, based on a suite of numerical model simulations, that block formation during the 2005/06 winter might well be sensitively dependent upon the upstream SST and low-level humidity over the western Atlantic.
The foregoing results relate to this particular winter and therefore are not necessarily generalizable. Further examination of other anomalous winters using the ERA-40 reanalysis dataset would provide one way of exploring their more general validity. Nevertheless the present study does invite consideration of: 1) the nature of the blocklike pattern of climate variability and its relationship (or lack of) to both the NAO and realized blocking events, 2) the possible implications of the results for seasonal prediction, and 3) the extent of the linkage or nonlinkage of the 2005/06 winter to trends in interannual climate variability. Here we comment briefly on each of these issues.
There is a recognized statistical relationship between the NAO index and atmospheric blocking frequency in the mid-Atlantic (e.g., Pavan et al. 2000; Shabbar et al. 2001; Croci-Maspoli et al. 2007a) with enhanced blocking frequency accompanying the negative NAO phase. Such a linkage was not evident during the 2005/06 winter. In contrast the peak in the blocking activity shifted eastward relative to its climatologically favored location, and the dominant prevailing pattern of climate variability acquired a blocklike configuration at a location commensurate with the blocking activity. The foregoing betokens either a major modification of the conventional NAO phenomenon or another pattern of climate variability (e.g., EOF3 in Fig. 1). In this context we note that the latter pattern is statistically robust in so far as it is evident at different levels and is reproduced in one-point correlation maps based on the pattern’s primary center (not shown), and physically relevant through its association with blocking in the eastern Atlantic. Likewise, we note that, whereas a temporal interdependence of the EOFs themselves cannot be ruled out (cf. Cassou et al. 2004), the projection (not shown) of time snapshots of the evolving 500-hPa geopotential field that accompany each of the 2005/06 winter blocks upon the three EOF patterns of Fig. 1 points to the centrality of the EOF3 component over most of a block’s lifetime.
Two aspects of the present study relate to seasonal prediction. Many statistical-forecasting procedures are predicated upon relating the future evolution of the NAO index to European winter conditions. For the 2005/06 winter the mean value of the NAO index as displayed in Figs. 3 and 6 was comparatively modest, and moreover Saunders and Lea (2006) show that for three different measures of the NAO’s index the winter values were within one standard deviation of their multiwinter mean. [Notwithstanding empirical NAO-based predictions for the winter 2005/06 matched the actual outcome with 2/3 probability—see Graham et al. (2006)]. In contrast we demonstrated here the anomalous PC value of the blocklike pattern of variability. An inference would be that the prediction of the latter pattern would have been pertinent and useful for the 2005/06 winter. Again the results of the diagnostic trajectory analysis presented herein suggests that a key ingredient of a successful forecast of a block (or series of blocks) would be the adequate representation of diabatic effects associated with the injection of low PV air into the upper troposphere. In effect, the inference would be that capturing the net effect of the flow dynamics and cloud thermodynamics in a numerical weather prediction model is important.
In relation to the third issue raised above, we note that the 2005/06 winter in Europe poses a conundrum with regard to recent trends in interannual climate variability. The observed low temperatures and the exceptional values for the accumulated snow depth ran counter to the prevailing trend for warm European winters. However, in contradistinction we note here that 1) the occurrence of several significant blocking events in the eastern Atlantic during the 2005/06 winter is consistent with the decadal trend in regional blocking (Croci-Maspoli et al. 2007b), which exhibits a significant reduction of mid-Atlantic blocks and an increase of blocks farther to the east; 2) the monopolar pattern (Fig. 4b) for the SST anomaly during the 2005/06 winter closely resembles the observed decadal trend (Solomon et al. 2007) in the wintertime SST for the North Atlantic. A continuation of the aforementioned blocking and SST trends would imply that a more frequent occurrence of winters akin to that of 2005/06 cannot be ruled out.
Acknowledgments
The authors thank MeteoSwiss for access to, and the ECMWF for the quality of, the ERA-40 dataset. In addition, we thank Daniel Lüthi for the support with the CHRM model. This study was funded in part by the Swiss NCCR Climate Programme.
REFERENCES
Altenhoff, A. M., O. Martius, M. Croci-Maspoli, C. Schwierz, and H. C. Davies, 2008: Linkage of atmospheric blocks and synoptic-scale Rossby waves: A climatological analysis. Tellus, 60 , 1053–1063.
Barnston, A. G., and R. E. Livezey, 1987: Classification, seasonality, and persistence of low-frequency atmospheric circulation patterns. Mon. Wea. Rev., 115 , 1083–1126.
Barriopedro, D., R. Garcia-Herrera, A. R. Lupo, and E. Hernandez, 2006: A climatology of Northern Hemispheric blocking. J. Climate, 19 , 1042–1063.
Beranova, R., and R. Hadan, 2008: Time variations of the effects of circulation variability modes on European temperature and precipitation in winter. Int. J. Climatol., 28 , 139–158.
Bresch, D. N., and H. C. Davies, 2000: Covariation of the mid-tropospheric flow and the sea surface temperature of the North Atlantic: A statistical analysis. Theor. Appl. Climatol., 65 , 197–214.
Cassou, C., L. Terray, J. Hurrell, and C. Deser, 2004: North Atlantic winter climate regimes: Spatial asymmetry, stationarity with time, and oceanic forcing. J. Climate, 17 , 1055–1068.
Casty, C., J. Luterbacher, H. Wanner, J. Esper, and R. Böhm, 2005: Temperature and precipitation variability in the European Alps since 1500. Int. J. Climatol., 25 , 1855–1880.
Croci-Maspoli, M., C. Schwierz, and H. C. Davies, 2007a: Atmospheric blocking: Space–time links to the NAO and PNA. Climate Dyn., 29 , 713–725. doi:10.1007/s00382-007-0259-4.
Croci-Maspoli, M., C. Schwierz, and H. C. Davies, 2007b: A multifaceted climatology of atmospheric blocking and its recent linear trend. J. Climate, 20 , 633–649.
De Pondeca, M. S. F. V., A. Barcilon, and X. Zou, 1998: The role of wave breaking, linear instability, and PV transports in model block onset. J. Atmos. Sci., 55 , 2852–2873.
Dole, R. M., and N. D. Gordon, 1983: Persistent anomalies of the extratropical Northern Hemisphere wintertime circulation—Geographical distribution and regional persistence characteristics. Mon. Wea. Rev., 111 , 1567–1586.
Folland, C. K., D. E. Parker, A. A. Scaife, J. J. Kennedy, A. W. Colman, A. Brookshaw, S. Cusack, and M. R. Huddleston, 2006: The 2005/06 winter in Europe and the United Kingdom. Part 2: Prediction techniques and their assessment against observations. Weather, 61 , 337–346.
Gheusi, F., and H. C. Davies, 2004: Autumnal precipitation distribution on the southern flank of the Alps: A numerical-model study of the mechanisms. Quart. J. Roy. Meteor. Soc., 130 , 2125–2152.
Graham, R. J., and Coauthors, 2006: The 2005/06 winter in Europe and the United Kingdom. Part 1: How the Met Office forecast was produced and communicated. Weather, 61 , 327–336.
Hoerling, M., J. Eischied, X. Quan, and T. Xu, 2007: Explaining the record US warmth of 2006. Geophys. Res. Lett., 34 , L17704. doi:10.1029/2007GL030643.
Hurrell, J. W., 1995: Decadal trends in the North-Atlantic Oscillation—Regional temperatures and precipitation. Science, 269 , 676–679.
Hurrell, J. W., 1996: Influence of variations in extratropical wintertime teleconnections on Northern Hemisphere temperature. Geophys. Res. Lett., 23 , 665–668.
Hurrell, J. W., Y. Kushnir, and M. Visbeck, 2001: Climate—The North Atlantic Oscillation. Science, 291 , 603–605.
Hurrell, J. W., Y. Kushnir, G. Otterson, and M. Visbeck, 2003: The North Atlantic Oscillation: Climatic Significance and Environmental Impact. Geophys. Monogr., Vol. 134, Amer. Geophys. Union, 279 pp.
Hurrell, J. W., and Coauthors, 2006: Atlantic climate variability and predictability: A CLIVAR perspective. J. Climate, 19 , 5100–5121.
Kung, E. C., C. C. Dacamara, and W. E. Baker, 1990: Simulations of winter blocking episodes using observed sea-surface temperatures. Quart. J. Roy. Meteor. Soc., 116 , 1053–1070.
Kuzmina, S. I., L. Bengtsson, O. M. Johannessen, H. Drange, L. P. Bobylev, and M. W. Miles, 2005: The North Atlantic Oscillation and greenhouse-gas forcing. Geophys. Res. Lett., 32 , L04703. doi:10.1029/2004GL021064.
Lejenäs, H., and H. Økland, 1983: Characteristics of Northern Hemisphere blocking as determined from a long-time series of observational data. Tellus, 35 , 350–362.
Losada, T., B. Rodriguez-Fonseca, C. R. Mechoso, and H-Y. Ma, 2007: Impacts of SST anomalies on the North Atlantic atmospheric circulation: A case study for the northern winter 1995/1996. Climate Dyn., 29 , 807–819.
Majewski, D., 1991: The European-Model of the Deutscher Wetterdienst. Proc. ECMWF Workshop on Numerical Methods in Atmospheric Models, Vol. 2, Reading, United Kingdom, ECMWF, 147–191.
Massacand, A. C., and H. C. Davies, 2001a: Interannual variability of the extratropical Northern Hemisphere and the potential vorticity wave guide. Atmos. Res. Lett., 2 , 61–71. doi:10.1006/asle.2001.0027.
Massacand, A. C., and H. C. Davies, 2001b: Interannual variability of European weather: The potential vorticity insight. Atmos. Res. Lett., 2 , 52–60. doi:10.1006/asle.2001.0026.
Massacand, A. C., H. Wernli, and H. C. Davies, 2001: Influence of upstream diabatic heating upon an Alpine event of heavy precipitation. Mon. Wea. Rev., 129 , 2822–2828.
Michelangeli, P. A., and R. Vautard, 1998: The dynamics of Euro-Atlantic blocking onsets. Quart. J. Roy. Meteor. Soc., 124 , 1045–1070.
Nakamura, H., 1994: Rotational evolution of potential vorticity associated with a strong blocking flow configuration over Europe. Geophys. Res. Lett., 21 , 2003–2006.
Nakamura, H., M. Nakamura, and J. L. Anderson, 1997: The role of high- and low-frequency dynamics in blocking formation. Mon. Wea. Rev., 125 , 2074–2093.
Pavan, V., S. Tibaldi, and C. Brankovic, 2000: Seasonal prediction of blocking frequency: Results from winter ensemble experiments. Quart. J. Roy. Meteor. Soc., 126 , 2125–2142.
Pelly, J. L., and B. J. Hoskins, 2003: A new perspective on blocking. J. Atmos. Sci., 60 , 743–755.
Pinto, J. G., T. Brücher, A. H. Fink, and A. Krüger, 2007: Extraordinary snow accumulations over parts of central Europe during the winter of 2005/06 and weather related hazards. Weather, 62 , 16–21.
Rogers, J. C., 1990: Patterns of low-frequency monthly sea level pressure variability (1899–1986) and associated wave cyclone formation. J. Climate, 3 , 1364–1379.
Saunders, M. A., and A. S. Lea, 2006: The 2005/06 UK and European winter: The UCL forecast and its assessment against observations. Weather, 61 , 347–352.
Sausen, R., W. König, and F. Sielmann, 1995: Analysis of blocking events from observations and ECHAM model simulations. Tellus, 47 , 421–438.
Scaife, A. A., C. K. Folland, L. V. Alexander, A. Moberg, and J. F. Knight, 2008: European climate extremes and the North Atlantic Oscillation. J. Climate, 21 , 72–83.
Scherrer, S. C., M. Croci-Maspoli, C. Schwierz, and C. Appenzeller, 2006: Two-dimensional indices of atmospheric blocking and their statistical relationship with winter climate patterns in the Euro-Atlantic region. Int. J. Climatol., 26 , 233–250.
Schwierz, C., 2001: Interactions of Greenlands-scale orography and extra-tropical synoptic-scale flow. Ph.D. thesis, Swiss Federal Institute of Technology ETH, Dissertation ETH 14356, 222 pp.
Schwierz, C., M. Croci-Maspoli, and H. C. Davies, 2004: Perspicacious indicators of atmospheric blocking. Geophys. Res. Lett., 31 , L06125. doi:10.1029/2003GL019341.
Shabbar, A., J. P. Huang, and K. Higuchi, 2001: The relationship between the wintertime North Atlantic Oscillation and blocking episodes in the North Atlantic. Int. J. Climatol., 21 , 355–369.
Solomon, S., D. Qin, M. Manning, M. Marquis, K. Averyt, M. M. B. Tignor, H. L. Miller Jr., and Z. Chen, Eds. 2007: Climate Change 2007: The Physical Sciences Basis. Cambridge University Press, 996 pp.
Tibaldi, S., and F. Molteni, 1990: On the operational predictability of blocking. Tellus, 42A , 343–365.
Trigo, R. M., T. J. Osborn, and J. M. Corte-Real, 2002: The North Atlantic Oscillation influence on Europe: Climate impacts and associated physical mechanisms. Climate Res., 20 , 9–17.
Trigo, R. M., I. F. Trigo, C. C. DaCamara, and T. J. Osborn, 2004: Climate impact of the European winter blocking episodes from the NCEP/NCAR Reanalyses. Climate Dyn., 23 , 17–28.
Uppala, S. M., and Coauthors, 2005: The ERA-40 re-analysis. Quart. J. Roy. Meteor. Soc., 131 , 2961–3012.
Vidale, P. L., D. Lüthi, C. Frei, S. Seneviratne, and C. Schär, 2003: Predictability and uncertainty in a regional climate model. J. Geophys. Res., 108 , 4586. doi:10.1029/2002JD002810.
Wernli, H., and H. C. Davies, 1997: A Lagrangian-based analysis of extratropical cyclones. 1. The method and some applications. Quart. J. Roy. Meteor. Soc., 123 , 467–489.
Spatial (a) EOF1, (b) EOF2, and (c) EOF3 patterns of monthly DJF Z500 fields (ERA-40 + winter 2005/06) over the Euro–Atlantic region. Shown are the Z500 anomalies in geopotential meters associated with a principal component amplitude of one standard deviation. The variances explained by each of the three EOFs are shown in the lower-left corner of the corresponding figure parts. Contour interval is 10 m; the 0-m line is omitted.
Citation: Monthly Weather Review 137, 2; 10.1175/2008MWR2533.1
Standardized PC as (a) temporal evolution and (b) density distribution of the first three EOFs (EOF1 = bottom, EOF3 = top). The winter 2005/06 months are indicated in different colors (blue = December 2005, red = January 2006, green = February 2006). The thick straight lines represent the mean, and the dashed straight lines represent ±1 standard deviation of the standardized time series.
Citation: Monthly Weather Review 137, 2; 10.1175/2008MWR2533.1
Mean seasonal (December 2005–February 2006) Euro–Atlantic PV distribution (PVU) at 315 K in thick black contours. The color shading indicates negative PV anomalies in blue and positive anomalies in red with respect to the ERA-40 winter climatology.
Citation: Monthly Weather Review 137, 2; 10.1175/2008MWR2533.1
Mean seasonal (December 2005–February 2006) Euro–Atlantic (a) T2M and (b) SST anomalies (°C) with respect to the ERA-40 winter mean. Negative anomalies are indicated in blue shading, positive anomalies are shown in red shading, and the data are slightly filtered. Note the nonlinear scaling in (a).
Citation: Monthly Weather Review 137, 2; 10.1175/2008MWR2533.1
The bottom curve represents the temporal evolution of the winters 2005/06 daily 2-m temperature (5-day running mean) over central Europe. Light (dark) gray shading indicates negative (positive) anomalies in respect to the ERA-40 climatology. Dashed contours represent ±1 standard deviation of the daily climatological mean. The middle thick black line indicates the mean VAPV distribution over the northern European region, and the top thick black line indicates the daily NAO index with arbitrary scaling. The five horizontal bars at the bottom signify the duration of the major blocking events in the Euro–Atlantic region.
Citation: Monthly Weather Review 137, 2; 10.1175/2008MWR2533.1
Difference between the standardized monthly mean T2M over central Europe and the monthly mean NAO index for the entire winter ERA-40 period and for January 2006 (red). Note the biggest difference is February 1991.
Citation: Monthly Weather Review 137, 2; 10.1175/2008MWR2533.1
Infrared satellite image sequence from blocking formation to the mature blocking stage at 1200 UTC 23 Dec 2005–26 Dec 2005. Superimposed are PV contours at 315 K with the 2-PVU isoline in boldface (from 1 to 8 PVU with 1-PVU spacing). Note the cloud band on the upstream side of the block during formation.
Citation: Monthly Weather Review 137, 2; 10.1175/2008MWR2533.1
The 96-h backward trajectories initiated at the blocking center at 1200 UTC 24 Dec 2005. The color shading represents the height (hPa) of the air parcel at the corresponding location. The blue circles indicate the initiation location of the backward trajectory and the gray circles the corresponding 1-day steps. See text for the meaning of the symbols N and P.
Citation: Monthly Weather Review 137, 2; 10.1175/2008MWR2533.1
Temporal evolution of (a) the absolute humidity and (b) the diabatic heating rate during blocking development as composite of all winter 2005/06 blocking events. Displayed is the mean of all individual 5-day backward trajectories starting at blocking initiation and the four consecutive time steps (1 day in total). The gray shading indicates ±1 standard deviation, and the thick contour represents the composite mean.
Citation: Monthly Weather Review 137, 2; 10.1175/2008MWR2533.1
The PV at 315-K distributions at 0600 UTC 26 Dec 2005 for CHRM scenario runs initiated at 0000 UTC 21 Dec 2005: (a) control run (CTRL), (b) North Atlantic SST decrease by 3°C (SSTM3), (c) relative humidity decrease to 10% (HUM10), and (d) combination of SST and relative humidity decrease (SSTM3HUM10). The red thick contour in (a) signifies the location of the block.
Citation: Monthly Weather Review 137, 2; 10.1175/2008MWR2533.1
Temporal evolution during blocking formation of the mean PV at 315 K of the five winter 2005/06 blocks for different CHRM scenario runs. The black thick contour represents the control run, the red contour represents the scenario with reduced SST (SSTM3), the green contour represents the scenario with reduced relative humidity (HUM10), and in blue the combination of the former is shown (SSTM3HUM10). The gray shading represents one standard deviation of the control run of the five blocks.
Citation: Monthly Weather Review 137, 2; 10.1175/2008MWR2533.1
