ECMWF Forcing metre wind component

[avatarraz]

Hello,i am new with ROMS. Try to build model with wind component. I use tool d_ecmwf2roms.m

This script say to download v10u and v10v. but ECMWF does not have this type.

so i use 10v and 10u.

1. can i use wind components with both of this? do i have to change variable names?

2. i'm litle confuse about ROMS base date.
mybasedate = datenum(2000,1,1,0,0,0); at this line, do i have to change to 1900-01-01?

[kate]

The variable names for the winds don't matter - just match what is in your varinfo.dat to the the names in the files.

The time origin doesn't matter either, as long as you are consistent among all your inputs and you don't ask ROMS to have a negative time. Starting in 2000 is fine for all recent years, but not fine if you start in say 1980.

[avatarraz]

Ok. Thank you very much.

[avatarraz]

hello, kate. now i have problem with reading initialization forcing time.

the goal is to change "ana_smflux" with sustr and svstr.

so, i have already download ecmwf forcing, with time : 3,6,9,12. check the forcing was divided per 3-hours.

i got this error :

CHECK_MULTIFILE - Error while processing Forcing multi-files:
data does not include initialization time = 2017-01-02 00:00:00.00

Found Error: 04 Line: 366 Source: ROMS/Utility/check_multifile.F
Found Error: 04 Line: 757 Source: ROMS/Nonlinear/initial.F
Found Error: 04 Line: 188 Source: ROMS/Drivers/nl_ocean.h


time value start from 0.0625, so i decided to add time (-0.0625). so i got value from zero.
but the error is still there.

i have checked my forcing sms.nc set to days since:

sms_time
. Size: 2944x1
. Dimensions: sms_time
. Datatype: double
. Attributes:
. long_name = 'surface momentum stress time'
. units = 'days since 2017-01-02 00:00:00'
. calendar = 'gregorian'

sustr
. Size: 481x191x2944
. Dimensions: lon,lat,sms_time
. Datatype: single
. Attributes:
. long_name = 'surface u-momentum stress'
. units = 'Newton meter-2'
. time = 'sms_time'
. coordinates = 'lon lat sms_time '

svstr
. Size: 481x191x2944
. Dimensions: lon,lat,sms_time
. Datatype: single
. Attributes:
. long_name = 'surface v-momentum stress'
. units = 'Newton meter-2'
. time = 'sms_time'
. coordinates = 'lon lat sms_time '

i think my problem is my initial.nc does not has sustr and svstr, or time is not in right set up?

[kate]

I don't know about a negative time. You can just set the time value to zero for that record and it should work.

[avatarraz]

hello kate, i solve this. thank you..