Reliability seasonal models
Computer calculations for a whole season ahead are still in their infancy, but they do add something to the statistical approach. Monthly and seasonal forecasts come true about 60 percent of the time. By comparison, the forecast for 5 days ahead comes true 90 percent of the time. That is why a seasonal forecast really needs to be viewed differently.
The seasonal forecast for the first month is logically more reliable than for the third month. Moreover, monthly and seasonal forecasts for the summer and winter come true more often than those for spring and autumn. This is because the predictability of the atmosphere is less in the transitional seasons, where cold and heat often alternate.