Forecasting with several models

#1
I would like to know how to deal with (presumed) independent models that provide different forecasted values.

For example, let's assume that I need to forecast the weather and I have one model based on the correlation between the weather and the levels of radiation emitted by the sun, and another model based on the correlation between the weather and the ozone levels. For next year, each model gives me a different probability of experience an increase of the global temperature. Which probability should I infer from both results? Or better, how should I deal with this kind of models?

Thanks