Dear all,
I am working on verification of seasonal forecast from AOGCM (primarily CFSV2) over Greater Horn of Africa. For the time being, my focus is on boreal summer season (JJA). I am seeing a pattern where skill scores for 2 and 3 month lead hindcast ( JJA hindcast issued in April and March respectively) are better than scores for 1-month lead (JJA hindcast issued in May). I was hoping to know if "spring predictability barrier" can partly explain this skill differences.
Thank you
Hi Andualem,
I hope all is well with and you have sorted out this issue. As far as I understood, if this is something to do with spring predictability barrier the reverse statement should be true. May initialized seasonal system supposed to give a better skill for JJA hindcast, partly because in May spring predictability barrier must be over and the seasonal forecast system can mimic initial condition and should provide a better hindcast. Please let me know the progress of your work.
I am also evaluating the skill of multi models ensemble forecast over GHA.
Cheers,
Titike.