Active Tropical Atlantic just now ....

[Martin Rowley [West Moors/East Dorset]]

... after a relatively quiet spell ... lots of interest stretching from Saharan west Africa right across to Bermuda! I don't have too much faith in medium-range forecast models for mid-latitudes at this time of year until these beasties have hibernated for winter: cause no end of trouble with injection of warmth, humidity, momentum into mid-latitudes :-)

Martin.

[Smartie]

Global 120hr skill scores have been rising lately-

http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/

so it will be interesting to see what effect the renewed activity does have.

[Jack Frost]

Out of interest... In terms of hurricane forecasting (and, looking at the skill scores, everything else), the CMC model (that's the Canadian one right?) always seems a bit rubbish. As such, (1) does anyone ever really pay attention to it, and (2) what is the point of it?! Are they hoping it will eventually develop into something to rival UKMO, ECMWF or GFS? Or does it have some other purpose, e.g. just Canadian weather forecasting?

Liam

[Jack Frost]

Sorry... this was in reply to Smartie's post!

Liam

[Smartie]

On Sunday, September 9, 2018 at 7:06:10 PM UTC+1, Jack Frost wrote:

Out of interest... In terms of hurricane forecasting (and, looking at the skill scores, everything else), the CMC model (that's the Canadian one right?) always seems a bit rubbish. As such, (1) does anyone ever really pay attention to it, and (2) what is the point of it?! Are they hoping it will eventually develop into something to rival UKMO, ECMWF or GFS? Or does it have some other purpose, e.g. just Canadian weather forecasting?

Liam

It looks like an also ran inthe verification statistics I've seen. It's actually quite hard to find detailed documentation but likely hampererd by the coarse grid (~25km?), hydrostatic dynamics, convection scheme (Kuo is old hat) and data assimilation putting it behind UKMO and EC IFS. These deficiences will be especially telling on TC forecasting performance. They're working on a 2.5km LAM for Canada itself, probabably pretty tricky given the topography and Arctic /sub-arctic domain.