Long range (5-10 day) computer forecasts are notoriously inaccurate and must be used with care. They should be used only for long range planning, to pick a general area for travel, as they are likely to show many changes before August 11 arrives. The best sites will lie beneath an upper level ridge ahead of, or in, a surface high pressure region, if these structures are forecast. This does not guarantee good weather, but it does greatly enhance one's prospects. If the numerical models don't predict these conditions, then more sophisticated decisions must be made which will probably require the services of a meteorologist.
Computer forecasts become much more reliable within five days, and still more accurate within 48 hours. The amount of detail they reveal also increases, with fields of relative humidity, precipitation, cloud cover, winds and temperature becoming available as the critical date approaches. A logical approach would be to select two or three specific sites within a convenient travel distance at this time, and then make a decision between them at the last possible moment.
Long range forecasts for Europe from the National Weather Service (NWS) in the United States and the European Centre for Medium Range Weather Forecasting (ECMWF) are readily available on the World Wide Web from Purdue University and a number of other locations. At Purdue the U.S. medium range forecast (MRF) model is usually available each morning after 12:30 UT. The ECMWF forecast out to six days is also available at the same site.
MRF charts show the 500 millibar flow (about 5000 m above the surface) and the surface pressure pattern. Look for an upper level ridge on the 500 mb chart, and the position of highs and lows on the other. These two charts will allow an early evaluation of the weather over Europe (most models are global, though not all parts of the globe are shown in the Web sites). The really hard part is whether or not the models are trustworthy, especially eight or ten days into the future. What clues are available to indicate that the numerical weather patterns will match the real ones on eclipse day?
Consistency is one clue is the same general pattern forecast from one day to the next? Another is whether or not two different models forecast the same general patterns. Most likely there will be small differences between the two, but if highs and lows are hundreds of kilometers apart, or upper features don't line up, then use the charts with caution. Try to find a spot which looks good on both charts, and wait another day to see if the agreement improves. Don't finalize long range plans at this stage.
Once the eclipse is within five days, the greater reliability and detail in the models allows serious planning, for now we know not only the location of weather systems, but also how the cloud is draped around them. But be careful! Predictions are not gospel, and thunderstorms especially are difficult to predict with accuracy when more than a day or two away. Low level cloud is often unforecast, or indicated only in a subtle variation in the humidity pattern. Keep looking for the best spot within your travel range according to the position of the low and the upper ridge. By now of course, normal meteorological forecasts will be available and you may simply have to watch the television.
Long range forecasts over Turkey and the Middle East are not likely to be as informative as those over Europe, where weather systems are more active and changeable. Individuals and groups who are not mobile may wish to consider sites where climate statistics work in their favor, from Hungary eastward through Turkey, rather than take their chances with the westerlies in western Europe or England.
Table of Contents