Yves-Marie Maquet : Weather Forecaster & Publisher

Are you originally from Chamonix? If not, where did you grow up and what brought you here?

I’m from Brittany, I’m a sailor and I became interested in weather forecasting mainly because of sailing; I was yacht racing and so I knew quite a lot about meteorology. I was always aware of wind shift for example, if you can not see how the wind changes at sea you can find yourself in a lot of trouble!

I’ve been living permanently in Chamonix since 1994; I became responsible for an art publishing house in Geneva and I soon realised that it was better to live in Chamonix and only be in Geneva for work! At the time I decided to run my own publishing house here in Chamonix but when I started to work with the tourist office on Chamonix.com I had so much to do with understanding the beginnings of the internet that I decided to end the publishing business.

Have you been studying the weather a long time?

I have been dealing with the weather in Chamonix since 1998. I started the Chamonix.com website for the tourist office and the first webpage we published was a simple question “what would you like to see on Chamonix.com?” and we got a lot of replies, with more than 90% of them asking for a weather forecast. So we asked Meteo France if we could publish their daily weather forecast on our website and at first they said OK but after some years they felt that they were losing money because people were visiting our website rather than theirs! They were still using Minitel, a prehistoric version of the internet, so for some time we had a deal where I would provide them with an English translation that they could use in exchange for their weather data but eventually this came to an end.

When the mayor of Chamonix changed from Michel Charlet to Eric Fournier in 2008, we decided to create Chamonix-meteo.com as a more specialised service, dealing solely with the weather forecasts. By that time there were a lot of changes on the web; the US weather data was open and free to use, and more and more was being posted on the web by the Germans, the Swiss… although it is still quite difficult to get the original data from Meteo France!

How long does it take you to make a weather forecast?

At least an hour and a half per report but difficult ones can often take a few hours and I often make two per day. It’s not always obvious which weather conditions will be the hardest to predict. Sometimes steady weather can be harder to explain; this summer in fact I spent more time studying the weather during July and August than at any other time of the year. Low pressure systems that bring bad weather are often very obvious, but the causes that bring about good weather can be harder to explain.

What processes do you use to predict the weather?

There are a number of prediction models that I use, huge computers that analyse data. When I started the only model that we were able to access was the American GFS and it was on a very large grid, and not very useful for a small area like the Chamonix valley; Since then many more models have become available, updating their information as much as four times a day, there are also many universities all around the world that are also analysing various models and interpreting the data in different ways  - I think that now we almost have too much access to too many models! It is obviously better than to have too little but it changes our way of dealing with the forecasting world - you have to decide which model might be the best for the current synopsis and sometimes they can be quite different and give you contradictory information.

So how do you decide which one to use?

(Yves-Marie chuckles and taps his nose!) It is true that at a certain point you have to decide without anything else that experience is sometimes so valuable. You have also to be very focussed; I feel that to tell my story I have to watch very carefully and we have more and more ways to know what is happening particularly with webcams. I watch a lot of webcams, not only in the Mont Blanc area but also in Switzerland, Italy, towards Lyon… I make timelapses to see how it goes, for instance there is a beautiful webcam in Talloires on Lake Annecy which shows me exactly how the wind shifts south-west of Chamonix, so I see what happens there before it happens here, I can track rain and tell if it will come to the Alps.

To tell the story you have to follow what makes the story. Science is when people invent mathematical equations that they feed the computer with and they are able to predict the flow of the air around the earth. After that it is only technique, so you have to know a bit of the science but experience and practical knowledge are more important for forecasting because what comes out of the computer is precise, it’s incredible, you can see everything, but you have to decide if it will happen or not. You have to decide if there is a discrepancy between model A and model B and which one to believe.

That’s another thing which is quite interesting in the world of weather forecasting; it started dealing with cause and effect throughout the 17th century, when atmospheric pressure was discovered by Torricelli, Descartes and other genius people. The astronomers of England and France then used this new information about air pressure to forecast weather that would keep their war ships and trade ships safe - they understood the relationship between high pressure with fair weather and low pressure with bad weather. But the problem was that they could not see into the middle of the Atlantic, as their ships were not equipped with radio equipment. By the 18th century a lot of people all around Europe were becoming really interested in weather forecasting, meteorology and the science of the atmosphere. They were aware of weather forecasts really soon and of the idea that you can be wrong. I’ve experienced this strange relationship where people understand that it is a science and so they request better information and more accurate forecasting, however this deep historical trend in society means that they also trust in someone’s intuition more than in science.

How do you interpret and share the data that you study?

With some years of experience I think more and more that it is always important to look west, north, south and east of this area, even if I am focussed on a very small area I am at my best if I have the whole picture and the whole story, because in space and time, there is always a story. In some very complicated criminal stories, like those of Raymond Chandler, it can be very complicated and tricky to understand but it is a story and it can be told and that’s why I use words and not icons; even if the sentences are very short and structured, you can say so much more with words than with icons. I've been told so many times (not by people who are using the forecast, they seem to like the way it is, but by people who want to sell advertisements) that I should use icons but I will continue the way it is!

I think that people need to understand what is going on, even if it is complicated, and that there are causes and effects. We have to show people the relationship between the causes and the effects so that they can understand why things happen as they do - it’s really important. When I’m speaking of a very complicated story it doesn’t mean that there cannot be sudden or radical changes, but most of the time it’s steady and if there is a sudden change then it has a cause, even if the cause is not obvious, but if you try to explain the cause to the people they can better understand why it happened.

It is important because sometimes you have shifts in the timeline, so you are not wrong, but sometimes the phenomenon you are forecasting arrives two hours later or two hours before, and if people have an idea of what causes the dramatic change they can understand that maybe it will be a little bit later or a little bit earlier. The idea is to make people able to do their part of the job, the guy who is doing the forecasting cannot be at the Brevent or at the Grands Montets to tell you what decision to make! But if you have an idea of what is going on it can make you more aware of changes; if you must make a decision you must know why you do it.

How far in advance is it possible to accurately predict the weather?

I think it is an impossibility to forecast the weather for more than  ten days ahead. It’s a very difficult thing; computer people try to build models to go as far as 3 weeks, 4 weeks, months in advance but we know for sure that there is a limit. Exactly like the limit of the speed of light, you can not go faster than the speed of light, and that’s exactly the same for the weather forecast. What they call Chaos Theory forbids any possibility to give a forecast for more than three weeks or one month - a real forecast starts from the state of the atmosphere today and how it evolves hour after hour for up to three weeks. After that there are so many numbers to take into account that there is an impossibility to calculate them, from here it becomes statistics and statistics say that if we have certain conditions over a certain period then we have seen that there may be certain conditions in three months time but during this time there are a lot of things that can happen to change the outcome.

There are a lot of weather sources that attempt to give six month forecasts that cover the whole season - are they reliable?

Haha, it’s rubbish! There is an American model that tries to give a one month forecast and it is getting more and more precise, they are a little over 50% correct up to a time scale of 15 to 20 days. It is also a question of spatial scale - if you look at a large area like the whole of the Alps, it is easier to make a general forecast. You might be able to see that there will be a lot of rain in that area over the next three weeks but whether that rain will fall on Mont Blanc or not is a lot harder to say. Being based here and seeing the weather for myself I can sometimes tell whether it will be better to go skiing at Grands Montets or at Flegere on a certain day, sometimes it is really obvious but sometimes it is not. Localised weather is very hard to predict more than a few days ahead.

Have you seen many changes in weather patterns over the past 20 years?

Yes definitely, but it is difficult to describe as so many things are changing together so it is hard to see which causes bring about which changes. Also, we have more and more ways to monitor what is happening in the atmosphere.

There is a phenomenon called the Cold Drop that was barely acknowledged when I started in 1998, it was known but not well understood. The temperature in the atmosphere becomes lower when you are going up which is why it gets colder, but sometimes within that gradient you get a bubble of cold air which is surrounded by the normal air, this is the Cold Drop phenomenon. Because the cold air occupies less space than the warm air, there is low pressure building within the bubble, which can be responsible for unpredictable and turbulent weather.

On the other end of the spectrum is the income of warm air from North Africa and sub-tropical areas which settles as a kind of cover over the Alps preventing a change in air pressure, leading to the kind of hot dry weather that we had this summer. This is also what happened during the very hot summer of 2003. I have the impression that these phenomena occur more frequently these days than they used to… but I am not sure because I also know that we have better ways of observing and understanding it than we had 20 years ago.

Even if we knew that there were those kind of changes in history what is obvious is that each time somebody makes a forecast for the atmosphere warming up, the forecast is always well below the actual result. It is moving much faster than predicted; each time we make a comparison between the forecast and the outcome, we see that we have underestimated the outcome.

Another hot topic in Chamonix is air quality and pollution - have you seen significant changes in this in relation to the weather conditions? What do you think is the solution?

Yes, it’s obvious, particularly in the lower half of the valley. Last Christmas, for example, with the amount of cars we had with big engines making a lot of fumes. We had a temperature inversion and it made such a localised change. We are always a little bit late to react to it, we should be telling people in advance. It is too early to predict Christmas week just yet but if we know that there is the possibility of a temperature inversion on the way we should be telling people not to bring their cars but to use the public transport, otherwise it will only get worse.

The buses are much more efficient now in Chamonix than they used to be and the train is a wonderful way to travel around the valley; I used to drive to Les Houches to use the climbing wall but since the autumn I have been able to take the bus instead, I don’t need to use my car anymore because of the four evening buses that run as late as 11pm. It’s much better, you can come now from Argentiere to spend the evening in Chamonix and you don’t need to use your car anymore. For me it’s a major change.

A lot of people here are aware of it - they car share, they take the train, etc - but it’s not enough. We should warn people earlier when there is a risk, this is why I'm planning to introduce a pollution index to Chamonix-meteo.com this year, so that people are better informed.

How do you spend your time when you are not working?

I still like to sail and to climb; I enjoy being in the mountains.

What are your favourite weather conditions?

It depends on where I am - in the mountains or at sea! I don’t like wind in the mountains, it is usually dangerous.

Chamonix-meteo.com is the most localised weather report relating to Chamonix with many people using it to plan their activities - how do you feel about that?

I only forecast the weather, I can’t tell people what to do with that information. I can tell them if there is a high avalanche risk but it is their decision whether to ski or not, I can tell them if it will be hot but they must be responsible for whether they take enough water into the mountains with them. Responsibility cannot be shared, each must make their own decisions.

In 1979 I took part in the Fastnet Race* as part of the French Accanito crew; we knew that there was a chance of strong winds and that the weather alerts were on red, and each team made their own decision whether to race that day or not. We decided to race but had to be towed to Cork after facing a hurricane and force 10 gales that broke our rudder. Perhaps the new weather models would have predicted the severity of storm and we might have decided differently.
(*an iconic yachting race which was struck by disastrous storms in 1979, causing 18 deaths and the largest ever rescue operation in peace-time)

So, finally the big question… when will it snow next?

Not within the next 10 days, the weather is due to get warmer in the short term. After that, we will have to wait and see!