Bordeaux is getting hotter. and 2003 was the hottest yet. So how will warmer, dryer summers affect the wines of the future? To find out, says David Peppercorn MW, we must look to the heatwave vintages of the past

The phenomenal summer of 2003 highlighted once more that hot years can create problems. And if the forecasts are correct, these are problems which winemakers in Bordeaux may have to get used to.

In the past, those problems centred on high temperatures during the harvest itself, which made it difficult to control the temperature of the fermentation. When temperatures rise above 34°C, fermentation can stick, and then acetic bacteria can take over. If you’re lucky, you get away with high volatile acidity, commonplace in a year such as 1947. If you’re unlucky, the whole vat can turn to vinegar. In 1928, Lafite was pasteurised to prevent disaster, while in 1945, half the crop at Cheval Blanc suffered the same fate.

I remember lunching with my father at Château Cantemerle, in 1953, on my very first visit to Bordeaux. The last great resident proprietor, Pierre Dubos, showed me his weather records and, by way of illustrating the varying challenges of different vintages, picked out the years 1920 and 1921. The former was very straightforward, but the pages related to the latter were covered in red ink, indicating corrective action. Dubos remembered having to get up during the night to check temperatures and even sleeping in the cuvier on one occasion.

I was reminded of all this recently when drinking a wonderful bottle of Haut-Brion 1921. Jean Delmas’ notes say that in 1921, because of the high temperatures, there were few successful clarets apart from Cheval Blanc and Haut-Brion. (The great sweet wines were another story, both because the small cask-fermentations of very rich musts were not susceptible to such high temperatures, and because the harvesting was usually later.)

In the past, the possibilities for controlling fermentation were very limited. In essence, they consisted of pumping over and using ice. Blocks of ice, wrapped in sacking and lowered into vats, were still being used as late as 1961. More usually, one saw devices resembling milk coolers. The must was pumped through the coil while the cold water ran over the outside. This was why stainless steel vats were introduced during the 1960s, with Haut-Brion leading the way and the new British owners at Latour soon following. However, well before this, less wealthy proprietors had discovered that concrete vats were much more resistant to temperature fluctuations. Indeed, these continue to be widely used across the region. Château Pétrus has long had them and there is a battery of them at Cheval Blanc.

Of course, the 1980s arrival of heat exchangers has meant temperature concerns during fermentation are now a thing of the past, as coils can be fitted in traditional wooden vats just as easily as in stainless steel ones. Today, the main concerns in hot years are managing the vineyard to obtain both phenolic ripeness and high sugar content (see 1989) and, with the experience of 2003, trying to prevent grapes from being grilled on the vine.

2003

Whether this will prove to be a freak or a new type of ‘global warming vintage’ remains to be seen. The main feature was a record number days of extreme heat (over 30°C and even 35°C) during June, July and August. Fortunately, September saw cooler weather and some showers, which favoured later-ripening varieties, such as Cabernet Sauvignon, and later harvesting areas.

This looks like being a year of paradoxes. In January, after tasting ‘more wines at this stage than I have ever done before’, Christian Moueix commented that this was a year when ‘those who did nothing in their vineyards have done better than those who carried out all the work now fashionable’. He added that it was also a year when, ‘one can taste wines from two adjoining vineyards, one of them a disaster, the other magnificent!’

In March, Jacques Thienpont took the tough decision to declassify Le Pin because it simply wasn’t good enough. Yet on the other hand, his cousins at Puyguéraud in the Côtes de Francs had made something spectacular. This is an area, though, that is usually up to two weeks behind Pomerol in ripeness and where they have 40% Cabernet Franc and Cabernet Sauvignon.

The problem with this year was the sheer heat of the sun, meaning the grapes were often grilled on the vine, and instead of taking off leaves to obtain full ripeness, the grapes needed the shade to protect them. One can therefore understand that the Right Bank, with its preponderance of early-ripening Merlots and its fashion-conscious vignerons practising all the latest new-wave manoeuvres, would be especially vulnerable to such climatic conditions.

After tasting a wide range of wines from all over the region at the end of March, one can confirm that the Cabernet-based wines did best. Graves – with a few notable exceptions – is not up to the Médoc, while some later-ripening areas on the Right Bank, such as Lalande de Pomerol, have done better than their more prestigious neighbours. Equally, there are some stunning cru bourgeois wines. St-Estèphe is the star commune.

1990 & 1989

Despite the chronological proximity of these two years, since inception, the wines from them have grown further and further apart. 1989 is important

as an unusually early vintage (the earliest since 1893, until the record was broken in 2003), as well as a hot one, and it marked an important learning curve for Bordeaux vignerons.

Until 1989, the normal practice was to determine when picking should start by testing the rising sugar levels against the falling acidity counts. However, for the first time, there were instances of high sugars combined with a lack of ripe phenolics, the acids found in the skins. This meant you could find wines with 13.5% alcohol, but a certain ‘greenness’ in the taste.

As a consequence, phenolics were closely monitored, with the surest and quickest check found to be tasting the grapes. During the current vintage at Lafite, Charles Chevallier and three or four colleagues go out into the vineyard every morning and taste the grapes.

Looking at all the statistics produced by the Faculty of Oenology only tends to confirm the fact that statistics can be misleading. In general terms, the 1989s are concentrated and intense, with powerful tannins and thick textures – a classic hot year vintage character. Yet 1990s are generally more flattering and supple, more elegant, even with a hint of dilution in some cases. The wines are very classic for the Cabernet-based Médoc and Graves, but can be rather ungainly for the Merlot-based Right Bank wines. However, if one studies the figures, one finds that, while the sum of temperatures between April and September is almost identical for the two vintages, there were 38 days of great heat (over 30°C) in 1990 compared with 35 days in 1989, and 364mm of rain in 1989 compared with 319mm in 1990. Looking at individual months, July was hotter in 1989, but August was hotter in 1990, as was September. So while they both stand out as historically hot years, there is nothing in the bare figures which points to the different character of the vintages (although one can certainly say that proprietors benefited from their experiences in 1989 when tackling 1990).

A footnote here on the great sweet wines. Essentially, the quality and character of Sauternes is dependent on getting humidity at the right moment. So, while 1989s are rich and superbly concentrated as botrytis was delayed by dryness, in 1990 showers at the beginning of September resulted in an explosion of botrytis among already very ripe grapes, which led to unusually high sugar levels. Here, 1989s are more elegant and classic than the 1990s.

1982

While not a classic hot year, 1982 was an important milestone in Bordeaux, because very warm weather during the vintage made proprietors realise that many châteaux were not as well equipped to cope with such conditions as they had thought. The result was that they coped well with similar conditions in 1983 and could deal with 1989 when it came.

1976

The figures for the hours of sunshine and days over 30°C in 1976 were very close to 1989, but the weather broke during the vintage and there was just too much rain too late. The result was that, while the very thick skins gave powerful tannins, the juice was diluted and the development of the wines was compromised, so that there was often a certain separation of the fruit, which tended to fade, and the tannins, which became dry.

By the mid-1980s, my favourites were Lafite and Ausone, as both had a harmony rarely found in 1976. But when I compared three wines blind with Jean-Hubert Delon in 2001, Lafite, while supple and attractive, did not really have first-growth quality and was beaten into third place by a most impressive Mouton. This had wonderful middle fruit, richness and harmony, and was still youthful.

There was also a superb Léoville Las Cases – Michel Delon’s first vintage – which vied with the Mouton for concentration and ripe middle fruit. Shortly after this, a bottle of Ausone was still lovely.

1961

This is not a year one immediately thinks of as hot, but the figures clearly show it was. However, the factors which give the character of this superb year have more to do with the very small crop – due to coulure – and an exceptionally dry year.

What is remarkable about the year as a whole is that in spite of the concentration due to small yields, the tannins are so supple and elegant. This was a milestone on the way to controlled fermentations, in that over-extraction was avoided and everyone felt that the wines were more successful right across the region than might have been the case a few years before. It seemed that volatile acidity had been banished at last.

1959

Hailed before the harvest as ‘the vintage of the century’, the summer of 1959 was hot, but especially so during the harvest. The wines had a real rotie hot year character, with tougher, drier tannins, in marked contrast to 1961, and much less regular in quality than that year. Even so, wines like Lafite, Mouton and Haut-Brion remain superb. The Right Bank was still suffering from the effects of the disastrous 1956 frost.

1949

A year of exceptional heat, not far short of what was experienced in 1947. There were 43 days with temperatures over 30°C, compared with 38 in 1990. The wines were hard to vinify, due to the heat during the harvest, and volatile acidity was a problem. Lafite and Mouton were both exceptional and well-cellared examples will surely still be very fine. Haut-Brion and La Mission have both aged better than their 1947s. Pétrus and Cheval Blanc were memorable, not far behind their legendary 1947s, while Ausone is better than its 1947. However, many 1949s have now succumbed to volatile acidity and each bottle must be tackled on its merits.

1947

Apart from 1976, this was the hottest year in half a century. It was so dry that water was rationed in the Gironde and swarms of locusts came to Bordeaux. The most famous wine of the vintage was Cheval Blanc, where exceptional sweet richness and 14.1° covered up the volatile acidity that ruined many others. Pétrus, though relatively unknown then, was also very fine. The top Médocs were Margaux and Mouton. Sadly, most wines today have been all but consumed by the volatile acidity caused by the problems of vinifying wines high in sugar in very warm conditions.

1945

A year of extremes. It snowed on 2 May, causing a freeze which destroyed 80% of the crop. Then a very dry, hot summer produced wines of great richness, high in tannins and very slow to develop. In the end, Mouton must take the palm for its majestic power and richness and Haut-Brion for its astonishing beauty and harmony. But volatile acidity is a problem today for most wines, so perfect storage is an issue. There are also fakes around, so caveat emptor!

David Peppercorn MW is a world-leading expert on Bordeaux wines.

Written by David Peppercorn MW