Emma B posted a question to my last post to which I replied and I wanted to draw attention to it. My answer was as follows…..
“I agree that applying any viticultural index or rules-of-thumb to the UK should only be done with caution. These tools, whilst useful, are crude and are typically developed by calibrating input to output data (e.g. temperature summation to yield) in specific locations which can be different to the UK in many other ways. I also agree that it makes sense to use more detailed data for temperature summation (e.g. degree hours) since it makes intuitive sense that such a measure is closer to what’s really going on in the vines. Of course it’s still just a calibrated metric which should be treated with caution. My hope was to be able to analyse historic yield data from many UK producers and look at correlation of historic vintage variation to weather. By doing this it would be possible to create viticultural weather-indices specific to the UK. Clearly such index creation would be vulnerable to the complexities of the real world muddying the picture, but I would hope that with enough data this might even out, making the indices valid. Sadly it’s been tricky to get hold of quality data, and many producers are also protective of such numbers despite my protestations of confidentiality. However, I do have 10 years of data for one group of vineyards and have been able to cook up a very simple climate/yield model with a correlation over 80%. This is clearly a spectacular number, but one which should be treated with enormous scepticism until many more producers can be analysed and until any such model is “blind” tested on real world data as we go through time. It’s important to remain sceptical on this since although the model has a very intuitively appealing parameterisation, the details were chosen by me with an eye on getting a high weather/yield correlation, hence the model is “fitted” to describe history and its ability to predict the future is utterly untested. For you information my climate/yield model is currently saying that the 2011 harvest will be +12% on long term average, after taking into account increases and decreases for maturing and ageing vines. And yes, the model is based only on monthly data, the drawbacks of which you are clearly aware.
Anyway, a simpler answer is to say that the maps I created give a good picture of the relative climate in different areas of the country, and for that I find them of interest.”
I should add to that that the value of any such index might be limited due to the domination of meso-climatic factors such as sheltering by surrounding hills, aspect, proximity of large bodies of water and so on.