search

RDCEP / CFS

Download data from the CFS (Climate Forecast System) archive and CFS2 rolling updates.
1 stars 2 forks source link

create symlinks to name according to pseudo-months (30-day intervals) #3

Open nbest937 opened 11 years ago

nbest937 commented 11 years ago

Using the 30-day interval produces a date sequence that is difficult to memorize/recognize.

Julian days doesn't solve the problem because leap years are different.

Pseudo-months might look like 1982.01, 1982.02, . . . for 1982-01-01, 1982-01-31, . . .

joshuaelliott commented 11 years ago

pseudo months are perfectly fine. not sure why julian dates are a problem though... the mapping from julian to gregorian is different in leap years, but the mapping from 30 day interals to julian is not. julian goes from 1..366 instead of 1..365 and a julian of 2008061 is a different calendar date than 2007061, but the julian date that is 60 days after jan 1 is still 061 whether its 2007 or 2008.

On Tue, Feb 19, 2013 at 3:41 PM, Neil Best notifications@github.com wrote:

Using the 30-day interval produces a date sequence that is difficult to memorize/recognize.

Julian days doesn't solve the problem because leap years are different.

Pseudo-months might look like 1982.01, 1982.02, . . . for 1982-01-01, 1982-01-31, . . .

— Reply to this email directly or view it on GitHubhttps://github.com/RDCEP/CFS/issues/3.

Joshua W. Elliott Research Scientist and Fellow University of Chicago and ANL Computation Institute 5735 S. Ellis Ave. Chicago, IL 60637 Links: Personal websitehttps://sites.google.com/site/joshuawrightelliott/ ; SSRN papers http://ssrn.com/author=1655092; E-mail: jelliott@ci.uchicago.edu

Adjunct Research Scientist Columbia University Center for Climate Systems Research NASA GISS; 2880 Broadway. NY, NY 10025; Rm 341 Tel: 212-678-5630

nbest937 commented 11 years ago

Maybe I'm using the wrong term, but what I mean is that a normal year uses days 1, 31, 61, 91, . . . and a leap year uses days 1, 31, 62, 92, . . . because they use the same calendar dates in all years so I thought of the "pseudo-months" to make the naming regular across years. Does that sound right?

joshuaelliott commented 11 years ago

Ah. That's fine then.

-- Sent from mobile -- Joshua W. Elliott Research Scientist and Fellow University of Chicago Computation Institute 5735 S. Ellis Ave. Chicago, IL 60637 On Feb 20, 2013 12:31 PM, "Neil Best" notifications@github.com wrote:

Maybe I'm using the wrong term, but what I mean is that a normal year uses days 1, 31, 61, 91, . . . and a leap year uses days 1, 31, 62, 92, . . . because they use the same calendar dates in all years so I thought of the "pseudo-months" to make the naming regular across years. Does that sound right?

— Reply to this email directly or view it on GitHubhttps://github.com/RDCEP/CFS/issues/3#issuecomment-13844909.