The seasonal climate forecast for the region of Komotini shows the monthly mean temperature and precipitation anomalies for the next 6 months in the top panel. The forecast is regional for areas of 100 km by 100 km or larger.
The seasonal forecast provides climate characteristics such as mean values or anomalies for an entire month. Anomalies are deviations from the climatological mean. Thus, a negative temperature and precipitation anomaly indicates cooler and drier than average conditions. Climatological information allow little inference on the expected weather. Assume a month with a positive anomaly of +1 degree. It is very unlikely that every hour of this month is 1 degree warmer. A more realistic scenario is that some days are significantly warmer than average, while others are on average. Most importantly, there might also be some days that are colder or even significantly colder than average, so the positive anomaly is not at all a guarantee to have e.g. no frost.
A seasonal weather forecast for particular day is not technically possible: it is statistically more unreliable than a climatic average. The reason is that daily weather is subject to larger swings influenced by mesoscale or microscale events, and originating factors cannot be measured precisely enough, so daily weather forecasts become statistically more unreliable than a climatic average about 10-14 days ahead. You probably noticed the unreliability of a 10-day weather forecast and predicting several months is clearly more difficult.
Unfortunately, also these climatological forecasts can be more or less reliable. Therefore, a 6 months forecast computed today may be different from the seasonal forecast computed yesterday.
To better understand how quickly the seasonal forecast is changing, new forecasts (actually an entire ensemble) are computed every day. We average all forecasts computed within a ten-day period, giving three periods of different forecast age for the last 30 days. The most current period (0-10 days) thus includes today’s forecast and the forecasts from the previous 10 days. This most recent forecast period is used to draw the maps for the next 3 months in the lower diagrams. If you see that forecasts of different age are very different, and even contradict each other, then there is very little hope of forecasting the season for that period of change.
There are some regions and situations where seasonal forecasts can be quite accurate. The most well-known examples are El Niño and La Niña situations.
The seasonal forecasts presented here are computed by NOAA and ECMWF.