Septorcher 2026 Discussion - Endless Summer

[WxWatcher007]

1 minute ago, weatherwiz said:

In terms of the statement, "looking at normal temp 15 days out is more accurate" around 15 days out or so climatology is going to have more "skill" versus a forecast model. Something I learned recently but the reason why say the GFS goes out to 15 days is because that is around the period when climatology holds better skill than forecast models. 

It's really interesting to see the strides made by AI guidance in recent years. We'll see if that continues into this winter. As I remember it, AI ensembles were great least year when legacy guidance was too quick to kick our coastal events out to sea. 

[weatherwiz]

Just now, WxWatcher007 said:

It's really interesting to see the strides made by AI guidance in recent years. We'll see if that continues into this winter. As I remember it, AI ensembles were great least year when legacy guidance was too quick to kick our coastal events out to sea. 

For sure. While I've been a huge negative on the AI side of things, I do think they are going to provide value but so long as they are used properly. Unfortunately though, we're also going backwards with a lot of the changes done within the field. The lack of weather balloon launches I think too has been extremely detrimental to forecasting, particularly short-term. Once we, if we, get to the point of quantum computing we will see substantial strides forward with modelling and forecasting.