In this article I set out to account for the varying quality of the wines in the years 1999, 2000 and 2001. The earliest year was a great vintage and the later years did not come up to the same standard.
In his authoritative work on climate as it affects the vine John Gladstones sets out the optimal ripening conditions for the production of various styles of wine.

The temperatures under which Cabernet Sauvignon matured in recent years in the Three Hills vineyard were:

• 1997    Average  17.7C  Highest Maximum  29.9C
• 1998     Average 17.4 C  Highest Maximum  32.9C
• 1999     Average 18.7C  Average Maximum 25.1  Highest Maximum 33C
• 2000     Average 16.9C  Highest Maximum  29.9C
• 2001     Average 16.62C Highest Maximum 30.7C
• Average 1997-2001 17.46C  Average Highest Maximum 31.28

The 1999 maturation period for vines exhibited the highest mean and maximum temperature. The temperatures for all years were well within Gladstoneís recommendations lying generally between the light fresh and full bodied categories. Given the lower temperatures in 2000 and 2001 we might expect the wines to lie more at the light fresh end of the spectrum than at the full bodied end and I would expect that grape flavour would be well retained.

I am revising this article in October 2004. Last week, I came across a piece of research emanating from Bordeaux in France that I regard as one of the most practical, most thorough and most helpful contributions to an understanding of what happens in vineyards during the maturation cycle. The article is in the public domain and can be accessed on the net. It is entitled RESEARCH ON 2-METHOXY-3-ISOBUTYLPYRAZINE IN GRAPES AND WINES. Dominique ROUJOU DE BOUBEE (School of Oenology, University of Bordeaux II)

Briefly, the authors established the level of methoxypyrazine that is perceptible in Bordeaux reds, traced the origin of the compound to the leaves, establish that it was the oldest leaves closest to the bunches that have the highest concentration, that the compound can move from leaves to fruit, that the generation of the compound in the leaves is related to high moisture levels in the soil, rainfall events and high soil moisture status, discovered that the greatest period of migration of the compound from leaves to fruit occurs between flowering and verasion and that it could be controlled by removing the leaves that surround the bunches. The authors recommend that this procedure be carried out pre verasion. Brilliant. Of course peasant growers have been doing this for generations in parts of Europe. I saw it being done in Bulgaria and I am told that it is done in Chateau Margaux. Some people do it for their tomatoes.

Independently I have come to a similar understanding of a slightly different means to contain pyrazine and green flavours in grapes simply by observing the difference between the fruit in my own vineyard, which is not watered and that of other vineyards that are watered. Over the last couple of years I have paid a lot more attention to the colour of the seeds than in the past and have come around to the point of view that brown seeds from fully mature fruit is associated with vines that mature fruit quickly and that there is a clear association between fruit maturity and leaf colour. Green chlorophyll rich leaves are associated with green flavours in the grape. For several years we have been excluding those parts of the rows that are too green from the harvest when our aim is to make great red wine.

The French researchers did discover that when leaves are removed from around the fruit zone the extra light falling on the berries can degrade the pyrazine to a form that has no organoleptic impact. Even without removing the leaves we have high light levels on the fruit because of our split canopy and low shoot densities. In dry years the leaves from around the fruit frequently dry up and fall off.

Recently I visited Virginia in the US for a conference of the American Society of Viticulture and Enology. I presented a paper on the relationship between temperature and flavour in the grape. Temperature is important. Under excessive maturation temperatures flavours are truncated. However, temperature is not the end of the story. I told them about my observations in relation to leaf colour.

Great was my surprise and delight when Kees Van Leeuwen Professor of Viticulture from Bordeaux presented a paper in which he clearly demonstrated the relationship between soil moisture levels in late season and the ratings of Bordeaux vintages. Dry soils are associated with the greatest years. Kees went on in a second presentation to demonstrate the association between wine quality and soil type. The best wine with the most complete and mouth filling range of flavour and the minimum of interference from green flavours came from soils that shed water away from the vines or allowed it to drain freely. Presto.

RAINFALL AND DEW

The years 2000 and 2001 were very much wetter over the maturation period than 1999. The year 2000 in particular was a wet year in the context of the other vintages.  In 2000 there was enough rain to produce the unusual circumstance of lots of green grass in the vineyard at picking time.  In retrospect, it appears that higher rainfall is the main factor accounting for the vintage downgrade. In this case, although temperatures were entirely favourable, soil moisture conditions were clearly not.

Adrian Jones and Hobart Brown check out the lifting potential of a stiff South Easterly sea breeze at Cape Leeuwin.

The graph below summarises the monthly heat load accumulated at weekly intervals in March and April across the years 1997, 1998, 1999, 2000 and 2001.   The earliest varieties Pinot and Chardonnay are usually picked in the second week of March and the latest red varieties including Merlot, Malbec, Cabernet Franc, Cabernet Sauvignon and Shiraz in mid to late April.  Study of the thermal conditions in classic wine producing regions in Europe and elsewhere reveal that, in the month prior to vintage low heat loads (less than 500degree hours in excess of 22C over a four week period) are required for full flavour accumulation regardless of variety.  This ideally means that weekly heat loads less than 125 degree hours should be experienced for full flavour outcomes.  The graph below indicates that 1997, 1998 and to a lesser extent 1999 (because the heat came early enough not to inflict damage on the late varieties) were likely to be the best years.  In the years 2000 and 2001 a marked heat event occurred on the 14th March which impinged on the ripening cycle of the latest varieties. 

Weighing it all up I would suggest that the thermal conditions are normally excellent but the rain can be the fly in the ointment. It certainly was in 2000. In the years when the heavens open up in January, February or March we need to look carefully at our leaves. In all years we need to be very careful when we water.

In mid March tropical cyclone activity introduced warm air, heavy cloud cover and persistent humidity over a period of about 10 days.  Minimum temperatures were uncommonly high.  However, peak daytime temperatures were low except for one day in the middle of the period when the temperature ran to just over 30C.  Only 6mm of rain fell for the entire month.  Humidity fell as low as 40% on four days.  The close association between daily temperature and humidity is clearly seen.  Overnight, humidity will recover to 100% and ground chilling will produce dew on the vines and fruit.  The leaves directly absorb some of this moisture.  Regular dews can promote disease if the berries are damaged in any way.

As the last grapes were picked in mid April 1999 temperatures warmed and humidity stayed high with the incursion of more tropical air.    No rain fell in Margaret River even though places further to the east were deluged.  Uncharacteristically a very warm period was experienced in mid April as the late varieties were picked.

Summing up, it appears from this rough analysis that maturation conditions in Karridale in 1999 were classically favourable to flavour production, conservation and retention. Early, midseason and late maturing grape varieties experienced relatively mild ripening conditions.  The period was dry throughout.

Getting serious:
While the analysis above appears to cover the ground we can do better. Using a spreadsheet to lay out hourly temperatures we can compute the degree to which heat was experienced in the range where flavour damage occurs.  Obviously, temperatures in the thirties are more damaging than the twenties.  A simple way to compare sites and years is to look at hour degrees accumulated above a given temperature

The graph below summarises the monthly heat load accumulated at weekly intervals in March and April across the years 1997, 1998, 1999, 2000 and 2001.   As mentioned earlier Pinot and Chardonnay are usually picked in the second week of March and the latest red varieties including Merlot, Malbec, Cabernet Franc, Cabernet Sauvignon and Shiraz in mid April.  Study of the thermal conditions in classic wine producing regions in Europe and elsewhere reveal that, in the month prior to vintage low heat loads (less than 500degree hours in excess of 22C) are required for full flavour accumulation regardless of variety.  This ideally means that weekly heat loads less than 250 degree hours, and preferably less than 125 degree hours should be experienced for full flavour outcomes.  The graph below indicates that 1997, 1998 and to a lesser extent 1999, because the heat came early enough not to inflict damage on the late varieties were likely to be the best years.  In the years 2000 and 2001 a marked heat event occurred on the 14th March which impinged on the ripening cycle of the latest varieties.  Mercifully it was early enough to do little damage.  The years 1997 and 1998 should have been capable of producing excellent wines across all the varieties, early to late.

However if high heat loads are to be experienced in early March in three years out of five one would expect that early and mid season varieties are likely to be adversely affected in most years.  In seeking to get a longer perspective temperature data was obtained from the lighthouse at Cape Leeuwin some 14 Km to the South West.

The figures for Cape Leeuwin Lighthouse demonstrate quite clearly the unpredictability of the incidence of continental heat on this otherwise warm but very maritime corner of the extreme south west of Australia.  Weekly heat loads greater than 50 degree hours greater than 22C can occur at any time between the last week in January and late April.

Twenty two degrees is a very comfortable temperature for plant and animal.  The vine year is considered to begin at higher latitudes where winters are much cooler when temperatures rise above 10C.  In the South West mean temperatures do not fall to 10C and leafing out of the vine corresponds more with a rise in the temperature above 15C.  With that perspective in mind the graph below presents a heat load accumulation above 15C rather than 22C.  The apparent variations in heat loads are less spectacular than when they are summed off a 22c base.  This statistic is related more to vine growth whereas the former relates to flavour accumulation late in the cycle of the vine.

Summary
Generalising in what may appear to be rather a cavalier fashion……...for it is impossible without going into the matter rather more than has been achieved in this session…....If we were to consider maturation period heat loads in excess of 125 hour degrees a week or 500  degree hours a month unfavourable to the production of the highest quality wine and were to avoid all places that ever experienced such conditions there would be few viable choices to be had in either hemisphere.  In the Australian context Karridale fares well in this respect across a wide range of varietal choices, but very reliably so for late ripeners that mature in mid April.  These include Shiraz, Merlot, Cabernet Sauvignon, Nebbiolo, Sangiovese, Grenache, Mataro, Graciano, Tempranillo, Tinta Cao, Petit Verdot and Carignan.   We expect to do very well with these varieties in most years.  In addition there will be years where we make intense and interesting wines from Pinot Noir, Chardonnay, Semillon, Viognier, Sauvignon Blanc, Chenin Blanc Muscat a Petit Grains and Verdelho.  Taken together this represents a wealth of possibility.

Are there warm and cool years?  Is the planet warming?
The aggregated heat load for the summer period or the year gives a good indication of whether the year was, on balance ‘cool’ or ‘warm’.   It demonstrates that of the five years 1998 was the coolest and 1999 the warmest.  Although the critical heat load of the warmest year is roughly twice that of the coolest year the ripening heat loads for both years were very favourable.  There is, on balance, a picture of relative consistency from the fruit production viewpoint, as one would expect in a remarkably maritime climate. Vintage 2001 was up to two weeks early due it appears to a warm and dry spring rather than a warm pre vintage period. Winter conditions following the 2000 vintage are exceptionally dry.  July in Perth was the driest since 1914.  As I write in August 2001 rainfall in the southwest is running 50% below average.  However, this variation is well within the range experienced in the past.  There is little reason to begin to expect the worst.

Note that the accumulated Hour degrees above 22C at Cape Leeuwin (790 degree Hours > 22 C) is rather less than that experienced at Three Hills(1200 Degree Hours > 22C) in the months February to April.  The vineyard is a short distance inland 10Km from the sea to the west and South.  The lighthouse is on a promontory, a windswept rock projecting into the sea.  It’s temperature regime is therefore quite unlike that of the hinterland.  Such is the influence of the sea.  Fortunately, it is the wind off the sea that prevails for all but a tiny fraction of the time.  It is that wind that makes our wine what it is.  You will now understand why it has the character that you see.