About storm numbering (PDF document):
https://www.nrlmry.navy.mil/atcf_web/docs/pdf/ATCF_Storm_Numbering.pdf
From ("About ATCF"): https://www.nrlmry.navy.mil/atcf_web/
Basically, they start out at 90 each year in each basin and go from 90 to 99. When they reach 99 in a basin, they go back and start at 90. "L" designates Atlantic.
When a storm is upgraded to either a depression or a potential tropical cyclone, it gets a number from 01 to 49. (again, each basin has its own numbering) Start at 01 for the year and those numbers don't repeat.
Here's where the best track and model system is located for the NHC:
http://ftp.nhc.noaa.gov/atcf/
"btk" folder, that's where best track data comes from. That's how we know when there's an invest.
"aid_public" is where the publicly available models are that the NHC releases. (They have access to models we don't, like Florida State Superensemble, this is all the public has access to.)
"fix" folder is where most of the center positions as determined by satellite are.
Those three things are downloaded and processed at HurricaneCity:
http://hurricanecity.com/models/
I also download Euro data here:
https://www.ecmwf.int/en/forecasts/datasets/public-wmo-and-acmad-datasets#Access
Once a storm gets a storm number. (01 to 49)
Another thing worth checking out in the NHC's ATCF system, the "stext" folder where SHIPS intensity output is.
As for invest numbering, they also have an additional numbering system that isn't really used too much. It is in the archive. That is:
A0 to A9
B0 to B9
C0 to C9
And so on.
First invest of the season in a basin is A0. That is 90.
Second 90 of the season is B0. In this numbering system, every invest has a unique identifier. You don't see this in the public facing best track and model system at HurricaneCity. When a storm is renumbered from 90 something to a storm number like 01, the best track line that indicates the storm has been renumbered. That line has that invest numbering scheme. In this file example from Irma:
http://hurricanecity.com/models/archive/al/2017/11/atcf/nhc/Best_Track.txt
Available as "Original Raw File" on the page here:
http://hurricanecity.com/models/models.cgi?basin=al&year=2017&storm=11&display=best_track
But here is where things get confusing, and I haven't even figured it out yet. For Irma you had these lines:
AL, 11, 2017083006, , BEST, 0, 162N, 280W, 35, 1007, LO, 34, NEQ, 30, 0, 0, 0, 1012, 180, 60, 0, 0, L, 0, , 0, 0, INVEST, M, 0, , 0, 0, 0, 0, genesis-num, 023, TRANSITIONED, alB32017 to al112017,
AL, 11, 2017083012, , BEST, 0, 163N, 298W, 45, 1006, TS, 34, NEQ, 20, 0, 0, 30, 1012, 180, 20, 50, 0, L, 0, , 0, 0, IRMA, M, 0, , 0, 0, 0, 0, genesis-num, 023, DISSIPATED, al112017 to al932017, TRANSITIONED, alC32017 to al112017,
This was the second 93 of the season in the Atlantic, and B3 was correct. But then for some reason they did something and then eventually called it C3 too. In the archive that is created after the season, the time as an invest for 93L should be listed as "alB32017" for the year.
2016 archive:
http://ftp.nhc.noaa.gov/atcf/archive/2016/invests/
But I still haven't figured out that renumbering system. The model system at HurrcaneCity use multiple techniques to try to determine when a storm has been renumbered from an invest to a storm number. (01 to 49)
I'm not sure if they will ever use the A0 type numbering system for invests. That's a lot newer. |