After a benign start to December 2009, a sudden change in the meteorological set-up occurred in the second half of the month which propelled not only Gibraltar but also its Iberian neighbour, Andalucía, into an exceptionally wet period. From December 2009 to February 2010 the total rainfall recorded at RAF Gibraltar was 1379mm which equates to 382% of the winter long-term average (LTA) for Gibraltar, far exceeding the previous record of 931.6mm set in 1962/1963. This, however, takes account only of the data gathered at the present observing site at North Front, established in 1947. Wheeler (2007) examined the full record for Gibraltar that goes back as a monthly series to the early nineteenth century (Figure 1). The previous winter record for this longer-term series was 1200mm set in 1855/1856. It is therefore clear that the winter's rainfall total was the highest for at least two centuries.
The synoptic background
The onset of this wet period coincided with the southerly displacement of the main polar front jet stream (PFJ) as shown in Figure 2 and its persistence at lower latitudes over the whole of the winter period. While its displacement is reflected in the notably negative North Atlantic Oscillation (NAO) index value at the time (Osborn, 2011), which was itself part of a very strongly negative Arctic Oscillation (Hadley Centre, pers. comm.), much investigation remains to be undertaken to account for this behaviour. One possible factor is that the strength of the El Niño in the run-up to winter may have been an influence. The persistence of the jet's displacement from the climatological mean will also have been helped by the intensification of the cold over higher latitudes: for example, the UK experienced its coldest winter for 31 years – since 1978/1979 (Prior and Kendon, 2011). It may not be a coincidence that the very wet winter of 1962/1963 in Gibraltar also coincided with the (much more exceptional) long and hard winter in the UK.
Displacement south of the PFJ is often linked with a weak or misplaced Azores High, and over this winter period the Canary Islands area became a prominent source region for a succession of low-pressure systems to develop and subsequently sweep northeast towards the Gulf of Cádiz. The positive sea-surface temperature (SST) anomalies in this source region during the winter period were of the order of 1–2 degC (Figure 3). This significantly large area of warmer-than-average water, in association with the enhanced upper-air forcing provided by the presence of the displaced PFJ, aided development of intense, moisture-laden storms, such as that which caused devastating floods over the Canaries and Madeira in early February 2010. This same storm went on to affect Gibraltar overnight into the early hours of 4 February. In late February, a further storm ravaged western Iberia and France with an associated tidal surge resulting in at least 50 deaths (Rowley, 2010).
Local causes and effects
In addition to the broader factors already outlined, Gibraltar's location was a significant additional factor contributing to the record-breaking rainfall. Its climate is typical of its Mediterranean setting, but with some variation due to its locality within the confines of the Strait of Gibraltar. It experiences generally hot dry summers and mild wet winters, with the wettest months normally extending from November to January. On the local scale, an important feature of some depressions that enter the Strait area is the convergence of contrasting winds. At the surface, the boundary layer flow into the eastern Strait becomes initially easterly induced by high pressure over Iberia. Funnelling of this more stable easterly (Levanter) wind through the Strait and its subsequent convergence with warm southwesterly (Vendeval) winds provides the wind shear and extra dynamical forcing that may result in embedded thunderstorm cells and in some instances slows down progression of the fronts, thus enhancing rainfall totals. Orographic uplift across the Rif and Atlas Mountains to the south over Morocco is also a notable contributing factor.
These factors are illustrated in the rainfall of 25/26 December 2009, which stands out as the wettest 24-hour period of winter 2009/2010 with 128mm recorded from 0900–0900 utc. This was brought about by a low-pressure system originating just south of the Azores that moved east to become centred just to the southwest of Iberia at 0000 utc on 26 December with central pressure 999mbar (Figure 4). Ahead of this system on the 25th, a freshening easterly wind was established with mean winds in Gibraltar peaking at 18kn (with gusts to 27kn) between 1400 and 1500utc, which incidentally coincided with the highest hourly rainfall recorded (31mm) over the 24-hour period. Winds aloft were already southwesterly at this point (Figure 5), providing a further mechanism for embedded cumulonimbus cells to form. The upper-air ascent shown for 1200 utc on 25 December 2009 clearly shows the considerable depth of instability over the Gibraltar region at that time, with cumulonimbus bases around 950mbar (600m) and tops around 200mbar (approx 12 000m).
Although the depression was not particularly deep, its warm source resulted in several active troughs and lines of embedded convection. For much of Christmas Day the temperature hovered around 15–16 °C; it rose to 17.5 °C by 2100 utc and continued to rise overnight to a maximum of 19.1°C at 0000 utc on 26 December. Winds overnight eventually switched to a strong southwesterly, with a maximum gust of 53kn recorded between 0600 and 0700utc on the 26th. Over this 24-hour period, a thunderstorm was observed at the station on 14 of the 24 observations.
Winter rainfall distribution
Although the current rain-gauge location is at the North Front site with records dating back to 1947, prior to this the gauge was located at a number of different sites across Gibraltar, making up what is now the Met Office series with records dating to 1790 (initially the rainfall data was only collated annually in what was documented as a ‘rain year’, normally running from August to July: the full monthly series is available from 1813). These sites, whilst all located within the small confine of Gibraltar, did vary to some degree in their exposure and altitude (Wheeler, 2007). Although more investigation might be needed, analysis of the interannual data does not suggest that any one site may have been significantly wetter or drier in any way that might weaken the conclusion that winter 2009/2010 was the wettest on record.
The main change in weather pattern in Gibraltar began in the second half of December. The total monthly rainfall recorded was 508.2mm (Figure 6(a)) amounting to 339% of the LTA, making it the third wettest December on record at the North Front site, despite the fact that the first two weeks of December were predominately dry. According to the Spanish Meteorological Service (AEMET, 2010) statistics, December rainfall across much of neighbouring Andalucía also exceeded 300% of the 1971–2000 average with Granada, Jaén and Almería setting new December records: Granada Airport recorded 184.4mm, 348% of the LTA.
January 2010 continued in the same vein with 386.2mm recorded in Gibraltar (Figure 6(b)), 299% of the LTA, again taking its place as the third wettest January in the North Front record. In Andalucía, many places recorded over 200% of the normal. However, Gibraltar's February rainfall was even more excessive: 484.6mm, 551% of the LTA (Figure 6(c)). This was the wettest February not only according to North Front statistics but also those for the previous rain-gauge locations contributing to the composite record dating back two centuries. Across most of Andalucía, February rainfall anomalies again exceeded 300% with many sites setting new records. One of these was Jerez de la Frontera, Cádiz, with 269.8mm, 450% of the February LTA. Also of note was the Rota Naval site in Cádiz, where the total was almost twice that of the previous wettest February (in records dating back to 1958).
Although winter 2009/2010 was clearly the wettest in Gibraltar since monthly records began in 1813, comparison of the interannual ‘rain year’ data from 1790, shows that that of 1855/1856 will go down as the wettest (Figure 7) in the official series. In the 12-month period, a total of 1969.9mm was recorded, as against 1817.8mm for 2009/2010.
The impact that this wet period has had on Gibraltar is not obvious. Gibraltar extends over approximately 2.6 square miles, the majority of which is taken up by the iconic feature of the ‘Rock’, made of limestone and extending to a height of 426m. Since it is surrounded by water on all but its northern limit, it is not particularly vulnerable to flooding and most of the rainfall, particularly intense rainfall, would have escaped as run-off. Apart from the occasional ingress of water to both commercial and residential properties, perhaps the more permanent impact has been to the Rock itself. During this winter Gibraltar experienced several rock falls on its steepest and most unstable east face. One of these occurred early evening on 7 January after a night of very heavy rain. Part of the cliff face weighing approximately 125 000kg (see photograph on page 14) is thought to have been dislodged from a height of around 150m.
Across the border in Andalucía, the impact of the wet weather was much more marked with the cost of the damage running into millions of euro. Farmers lost entire crops while villages had to cope with severe flooding on a frequent basis as rivers broke their banks and reservoirs were over-topped. Whole sections of motorways collapsed as embankments gave way beneath them, while the Costas had to mount a rapid clear up of beaches in the run-up to the Easter holidays (see photograph on page 14). In late February the provinces of Jaén, Córdoba and Seville were put on high alert as the River Guadalquivir broke its banks. The Guadalquivir is the longest river in Andalucía with its source in the Sierra Cazorla, Jaén: it runs through Córdoba and Seville before exiting into the Gulf of Cádiz. Following two months of above-average rainfall over the river's catchment area, February's estimated 209mm of rain (344% of the average) led to the river in Córdoba rising to such a height that, on the 24th, it covered all the pillars of the Puente Romano (an old Roman bridge) with some water ingress into nearby city streets. Incidentally, the last major flood of the Guadalquivir to affect the city of Córdoba was in 1963, drawing further parallels with the previous wet winter of 1962/1963.
All this is in stark contrast to the hydrological conditions of only a few years ago, when parts of Spain, including Andalucía, experienced severe drought. According to AEMET statistics, an average of 411mm fell across Spain in the ‘rain year’ of September 2004 to August 2005, making this the driest 12 months since 1947, resulting in one of the worst droughts since 1887. Later, in 2008, Spain again registered severe drought conditions with water restrictions set in place across parts of Andalucía and other provinces. From October 2007 to March 2008, Spain experienced its driest such six-month period in a composite record that began in 1947. The rain year as a whole though did not quite match the drought of 2004/2005 as the arrival of some very wet weather in spring 2008 did enough to reduce the deficit allowing some drought restrictions to be lifted.
It is clear that three main factors contributed to Gibraltar's record-breaking rainfall during winter 2009/2010: its location, the displacement of the PFJ, and the persistence of the warm-water anomaly in the Canary Islands region. A review of the selected European rainfall data shown on page iv of this journal's Weather logs for December 2009 to February 2010 clearly illustrates that whilst Gibraltar's rainfall total was historically significant, it was also noteworthy on a European scale.
The author would like to acknowledge with gratitude the many observers, both past and present, responsible for collating the series of rainfall data that made this article possible, and thank Dennis Wheeler (University of Sunderland) for the provision of his Gibraltar rainfall statistics and information on earlier rain-gauge locations. Thanks are also due to Met Office colleagues Adrian Semple and Martin Willett for provision of Figure 2 and an anonymous reviewer for much helpful advice in improving this article. The author would also like to thank the Spanish Met service (AEMET) for the rainfall and drought statistics made available via their website and also the editors of the Gibraltar Chronicle and La Opinión de Málaga for the photographs provided.