I’ve been toying with the idea of a Real-time summer index [SISI Summer Index] for many years. Of course the idea of a summer index is very subjective – everyone’s perfect summer day is different. I use a very simple algorithm of total cloud amount, present weather, relative humidity, surface visibility, mean wind speed and of course temperature to calculate it with. Results look promising. There’s no reason why the idea couldn’t be extended to use NWPNWP Numerical weather prediction uses mathematical models of the atmosphere and oceans to predict the weather based on current weather conditions. data for forecasting summer index for many days ahead. The values are using my preferences for an ideal temperature of 25°C. This is not ideal because what I would like to have used was anomalies. A fixed temperature also means the SI will oscillate through the day as things warm up, so it’s far from perfect. A perfect day to me really requires an okta of cumulus or cirrus cloud rather than complete blue skies. The perfect day also requires some wind, no more than five knots, perfect visibility and low humidities. I factor all of them into the algorithm, but thats still very much work in progress. The 25°C ideal temperature may be a little too high for me, but at that rules out SI levels much above 75 across the Mediterranean in the summer.