I promise them only weather!
Weather forecasting is extremely difficult. It is easier if you have a network of Cray supercomputers and sophisticated algorithms. It is very difficult, even in the short term, if the only data you have is from a single, isolated weather station in the foothills of the Troodos mountains, feeding in to a PC. The software I use revises its forecast every hour, but what it says can apply to 10 minutes to 48 hours ahead. If you wait long enough it is often right. For example, it currently says "partly cloudy and cooler". Well, there is a fairish chance it will be 100% correct in the next few hours. If I look at the curves at http://www.cypenv.org/weather/wx2.htm I can offer a complement of guesses. For example, I can see that it will be chilly tonight. I can see we are climbing out of a zone of instability, which means I'd make an educated guess tomorrow will be relatively fine and sunny but probably some clouds, with little or no rain. What I cannot see is where the depression which has just passed us is going. It is not impossible that it could backtrack on us, in which case today's or yesterday's weather could repeat itself. Alternatively, another one could gallop in to take its place: there is no way I could know without an analysis of weather system data from the mid-Atlantic to India and Finland to the equator. I don't have access to this data nor the computing power to process it. That is why I cannot give reliable forecasts; I can tell you only what I measure and what I can interpret from these data,
