The Annual Red Bordeaux Report

I traveled to Bordeaux this spring with the expectation that most 2003 clarets, from the region's hottest summer and earliest harvest in more than a century, would show a pronounced roasted quality and forever be marked by the character of the year. What I found was something far more complex, and nothing less than a miracle of Nature: in the Northern Medoc in particular, but in other parts of the sprawling Bordeaux vignoble as well, there are vibrant, balanced, even floral wines, albeit with extra density and tannic richness, that completely belie the extreme heat of the year. In the Southern Medoc and across much of the Right Bank, however, there was more evidence of fruit cooked or stunted by the sun: wines that are green, hollow, lacking flavor interest, flat, or dry on the finish. Vintage 2003 is in fact a wildly uneven year, and for this reason it may never be considered a great vintage. But I found myself more excited about the best wines than when I was tasting the 2002s and 2001s at a similar stage of their evolution.

The 2003 growing season. Thanks to a very wet autumn of 2002 and early winter, the soils of the region built up good water reserves, which would prove to be critical in the heat of the 2003 summer. A couple of cold snaps in late January and February had the beneficial effect of killing off bugs in the soil and forcing the sap deep down into the vines. But then a very warm March led to bud-burst a week or more ahead of normal. Dry weather limited the number of embryo grape bunches, however. Temperatures then descended to the freezing mark on several consecutive nights in early April, but the region was barely touched by frost. I've never had more consistently perfect tasting weather for my early spring trip to Bordeaux than I did at the end of March and beginning of April in 2003, what with high pressure, brilliant blue skies and a constant chill breeze from the northeast, but this same very cool weather slowed down the vegetation. The rest of April brought more normal temperatures and relatively little precipitation, which allowed the vegetation to catch up again.

The flowering began in mid-May, about two weeks ahead of the long-term norm, but some rain and cold weather in the middle of the merlot bloom led to an average set and small berry size. The cabernets, which flowered a bit later, enjoyed better conditions. From late May through the end of August, the Bordeaux region (and most of Europe) experienced an extended heat wave without precedent. Hot weather in June accelerated the development of the vines, but by July the vegetation began to suffer in the heat, especially on drier gravelly and sandy soils. The heat wave culminated in 11 consecutive days (August 3 through 13) in which afternoon temperatures reached or surpassed 95 degrees. Many vineyard owners told me that they routinely measured temperatures of 110 degrees or more in the vines during the scalding afternoons. In vineyards where the plants' roots could not find water, the vines' leaves turned yellow, withered and fell, and the vines shut down completely to protect themselves from the heat, but others had just enough moisture in the soils to somehow make it through. One chateau director told me that the stomata in the leaves literally closed up, to protect the plants from further dehydration. The vines, he went on, were thus virtually in a state of suspended animation, and remained preserved until showers could invigorate the vines and jump-start the process of photosynthesis. According to this theory, that's why so many wines avoid tasting obviously cooked. As a rule, the grapes remained small.

In general, vines on cooler chalk and moisture-retentive clay soils made it through the worst of the heat relatively unscathed, and were in the best position to take maximum advantage of showers that fell during the second half of August and near-ideal weather during the last two-thirds of September. Many proprietors in the Northern Medoc (in St. Estephe especially, where many top properties are largely on clay-based soils) told me that although 2003 was a summer of extreme heat, it was not a drought year. (Drought here was more of an issue in 1989 and 1990, years in which the water table was already low going into spring and summer.) More than one Bordeaux insider told me that you could take a helicopter tour of the vines in late August and virtually predict the future quality of the wines then. In some spots the foliage was limp and pitiful, while in others the vegetation remained heroically green. It was clear that St. Estephe and portions of Pauillac and St. Julien were the least adversely affected by the extreme heat. These appellations produced the lion's share of the year's top successes (including most of the wines that are likely to enjoy the longest evolution in bottle). But cooler argilo-calcaire (clay and limestone) soils on the St. Emilion cotes, especially on the escarpment to the east of the town, allowed for healthier merlot and late-picked cabernet franc brought in at record levels of ripeness. Clearly the most difficult important sector in 2003 was the sandy and gravelly Pomerol plateau and its extension into St. Emilion. Many properties here saw their vines shut down completely in August, then were forced to pick in early September. Too many of these wines are green, flavor-deficient or dominated by dry tannins, and do not rate a mention in this report.

As a rule, the best wines were made at estates where the fruit weathered the heat and was able to benefit from more moderate September conditions and additional hang time, which would bring about more thorough ripening of the skins. In general, this meant the later-ripening cabernets and petit verdot, and fruit on "colder" (i.e., later-ripening), moister soils. Warm but moderate afternoon temperatures along with cooler nights in September allowed healthy vines to gain in phenolic ripeness at a faster rate than they lost acidity.

Christian Moueix, whose firm J. P. Moueix makes and markets wines from all over the Right Bank, described the difficulties of the summer as follows: "On gravel the leaves were drying out and falling in August, as in 1990. The grapes were then exposed to direct hot sun and some berries began to desiccate. There was no chance of positive progress on pure gravelly soils so we had to pick beginning literally on the first day of September. But the clay soils kept more moisture and more leaves. Clay and chalk were the most resistant to drought and heat; in most of these vineyards we could wait two more weeks. Basically the quality of the fruit across much of St. Emilion and Pomerol depended on the amount and condition of the leaves that were left on the vines in early September."

Although rainfall during the summer of 2003 was much lower than average, sporadic showers were generally well-timed. (There were also two violent storms in the region, the first during VinExpo on June 24, which caused extensive hail damage in the Entre-Deux-Mers, the second on July 15, which hit hardest in Moulis and Blaye, but also clipped some parcels at Latour, Leoville-Las Cases and Ducru-Beaucaillou.) It is also worth noting that the prevailing breezes through most of the summer heat were from the west, while much of inland France experienced considerably drier winds from the east.

The 2003 harvest. The harvest was the earliest since 1893. But with the exception of Pessac-Leognan, where Haut-Brion picked its white grapes quickly beginning on August 13 and began its merlot on the 25th, most merlot across the Bordeaux region was brought in during the first half of September. Some parcels of merlot in the Medoc and on cooler clay and chalk soils on the Right Bank were harvested during the week of September 15, a period that featured very warm daytime temperatures but reasonably cool nights. Much merlot on the Pomerol plateau and in adjacent St. Emilion, where the grapes were especially fragile after having been weakened by the extreme conditions in early August, began to show signs of incipient rot following showery weather between September 6 and 10. Much of this fruit had to be picked quickly, before it was properly ripe (in many instances, sugars were already high, but acidity levels were dangerously low and seed and skin tannins were still green). Most cabernet sauvignon and cabernet franc was picked during the second half of September, with a few estates harvesting into early October.

The 2003 vinifications and wines. Most winemakers told me they carried out cooler fermentations, gentler extraction and shorter total cuvaisons (the total time the wine spends on its skins) to extract gently and to avoid getting excessive skin or tougher seed tannins. Many said they attempted to extract very early in the fermentation process, before alcohol levels increased, in order to minimize extraction of tannins. The high skin-to-juice ratio due to the small grapes and thick skins resulted in very high IPT readings (indice des polyphenols totaux), especially in the cabernets that were able to take advantage of longer hang time under good conditions in late September. A couple of chateau proprietors told me they did a "positive bleeding" of the juice, or a "reverse saignee," using some young-vines to supplement the highly concentrated older-vines juice, as there was so little juice in the grapes. Negociant Bill Blatch, the owner of Vintex, noted that this was a year in which quality was often inversely proportional to quantity: some of the best wines were made from normal yields (as in St. Estephe), while unusually low crop levels frequently signified problems with extreme heat and vine shutdown.

The harvest featured record sugar levels across much of the region, frequently 13+% for merlot, but routinely over 12% for cabernet sauvignon. Some late-picked cabernets on the Right Bank reached a freakish 14%. (This was a year in which some satellite Right Bank appellations where fruit normally struggles to ripen produced unusually successful wines with tannins less rustic than usual, such as limestone-rich Canon-Fronsac and the Cotes de Castillon.) Despite the grapes' low levels of acidity, most of Bordeaux's smartest winemakers did not acidify, often against the advice of their enologists. Time and time again I was told that acidity levels actually rose during fermentations. The acid levels measured at the outset in the juice were low, but these figures often proved to be deceptive, as the thick skins released significant quantities of acidity during fermentation. And with very low levels of malic acidity, there was little acidity to lose during the secondary fermentations.

Young 2003s from favored sites may indeed be low in acidity by historical standards, but not meaningfully lower than in past outstanding warm years. Relatively few 2003s made from thoroughly ripe fruit seemed seriously unbalanced or cooked in the early going, and many of them show surprising aromatic complexity and inner-mouth vibrancy. Although numerous less-successful 2003s from St. Emilion, Pomerol and the Southern Medoc will be best consumed early, before their drier tannins overwhelm their relatively meager fruit, I suspect that the stars of the Medoc will be long-lived thanks to their extraordinary density and huge, ripe tannins. Wines like Latour, Margaux, Montrose, Ausone and Leoville-Las Cases should develop in bottle for 25 to 30 years, or more. Other top 2003s-the other two Leovilles, the two Pichons, Haut-Brion, Cos d'Estournel and Calon-Segur, to name just a few-should evolve positively for at least two decades. But many, many 2003s will be best suited for drinking within the first 10 to 12 years of their lives. As this is a style of vintage virtually without precedent, I believe that the projections of peak drinkability offered by other publications should be taken with a grain of salt (and perhaps washed down with a glass of young 2003).

My tastings in early spring. Although my main agenda in Bordeaux was to taste the young 2003s from barrel, I attempted to retaste the 2002s from barrel, as well as to sample the finished 2001s, at the 60 or so chateaux I visited. At several group tastings, I normally looked at the 2003s and the 2001s. This represents a change from past practice, when I would focus on the new vintage and retaste the previous one when it was still in barrel or tank. My reasoning this year was that the 2001s, which are just now arriving in the retail market, are of more immediate interest to most wine consumers than the 2002s. In addition, many 2002s were on their finings, or otherwise out of sorts, in early spring, and not in the best condition to taste.

My coverage is essentially limited to red wines, although I have included notes on a handful of top Graves whites (this is not, generally speaking, an exciting vintage for this category). My early look at 2003 Sauternes suggests that this is another excellent vintage for these wines. Most of the best fruit was harvested relatively efficiently, with good levels of botrytis, during the second half of September. As a rule, the wines are higher in sugar but lower in acidity than the more powerful '01s, but it was too early to make judgments on individual wines in early April, as some of the blends were still works in progress. (I expect to publish notes on the 2002s from barrel and the bottled 2001s in the next issue.)

The finished 2001 reds. This vintage is looking increasingly like a classic Bordeaux year whose greatest shortcoming is that it followed the crazy-making millennial vintage of 2000. The wines tend to have considerable purity and complexity of aromas, moderate weight and ripe tannins. There are highlights in virtually every major appellation. While the wines possess the balance and stuffing for at least mid-term aging, I was surprised by how pliant and attractive most 2001s were on my recent trip, especially in light of how austere many of these wines were in the early going. It's entirely possible that the 2001s will shut down over the next year or two, and, indeed, several chateau proprietors made this prediction, but very few wines show obvious rough edges today. I now find this vintage more consistent than 2002, although I very much like the firm acid/tannin structure and brightness of the best 2002s.

A word on Bordeaux pricing. The big boys had not yet opened when this issue went to press, but it is likely that the 2003 first growths will be pricey, possibly in the range of 100 euros a bottle, ex-cellar. (This compares to 60 to 65 euros for the 2002s last year, 85 for the 2001s, and 120 for the small first tranche of 2000.) A weak dollar will make the 2003s seem even more expensive. Still, chateau proprietors are anticipating strong demand for the top two dozen or so 2003s, and I also expect the best wines to sell briskly and escalate in price. But Bordeaux is increasingly a two-tier market, with large quantities of 2002s and 2001s remaining unsold and many chateaux struggling for survival. There will be literally hundreds of bargains among 2003s from lesser appellations, not to mention great values from normally solid but unspectacular Medoc chateaux that have produced unusually dense, ripe wines in this vintage. Many of these wines are already opening at prices just 5% to 10% above those of 2002-a level that seem quite reasonable. (In my Bordeaux coverage in this issue, price ranges shown for 2001s come from a dozen or so major retailers across the country who have these wines on their shelves.)