Little_R format, SRFC = SLP, PW (DATA,QC,ERROR).

Problems with data in 3DVAR.

Little_R format, SRFC = SLP, PW (DATA,QC,ERROR).

Postby allenea » Wed Dec 20, 2017 3:06 am

I am looking at the obs.ascii and obs_gts_2008-02-05_12/00/00.3DVAR files from the WRFDA test case.

In the top section says;
SRFC_FMT = (F12.3,I4,F7.2,F12.3,I4,F7.3)

FM-13 SHIP 2008-02-05_16:00:00 SHIP or TEMP MOBIL Id >>> WDD2074 1 33.700 -118.200 0.000 WDD20
102450.000 0 200.00 -888888.000 -88 0.200
102450.000 0 160.00 4.116 0 1.10 10.000 0 1.10 0.000 0 6.00 285.950 0 2.00 -888888.000 -11 2.00 -888888.000 -11 10.00

(I am aware that this is created after obsgrid.exe).
The Sea Level Pressure (SLP) is the same as the observed pressure from the ship. I am going to assume PW means precipitable water. Are we supposed to calculate that and if so what should we make the error value? I do not think that I do this when writing the data to Little_R format, at least I have not come across this after reading README.obs_format and everything else.

I have a bunch of data from a boat. For each time (sometimes multiple obs per minute) there is an observations I should be giving it the header, single row of data, ending record.... and at the end of the file the tail integers.

The header should contain:
- Latitude, Longitude, ID, Name, Platform, Source, Elevation, and Date String

then I should give it the one surface observation in the Little_R format:
Pressure, QC, Height, QC, Temperature,QC, Dew Point, QC, Wind Speed, QC, Wind Direction, QC, U-Wind, QC, V-Wind,QC, Relative Humidity,QC, Thickness, QC
Format: 10(F13.5,I7) -- Data (missing data = -888888.00000), errorVal =0

Next is the Ending Record (as shown):
-777777.00000 0 -777777.00000 0 -888888.00000 0 -888888.00000 0 -888888.00000 0 -888888.00000 0
-888888.00000 0 -888888.00000 0 -888888.00000 0 -888888.00000 0

Then do the cycle over again for the next observation..... continue in the same ascii file until the end, then use the 3 tail integers to end the file.

Tail Integers: Format (3(I7)) : 0 0 0 <-- Some use it to store the number of observations, errors, and warnings, but these could be anything.

Can someone tell me if my thought process is correct or where I am going wrong.

Could you also explain how it eventually gets the PW in the ob.ascii after running obsgrid.exe... Is there a fortran program that does the calculation?

Lastly, it is my understanding that if I am using "Quality Control on Individual Reports" then all my QC values are 0. Correct?

Thank You,

Posts: 3
Joined: Tue Nov 15, 2016 8:32 pm

Re: Little_R format, SRFC = SLP, PW (DATA,QC,ERROR).

Postby allenea » Sun Dec 31, 2017 3:05 pm

Disregard.... I first tried using WRFDA, which was helpful because I figured out the Little_R format even though I thought it might be wrong since it was not working, but then when I tried OBSGRID it worked once I had the file name corrected. PW is not required (as I originally thought) but if your observations are from the surface you can just set SLP = P
Posts: 3
Joined: Tue Nov 15, 2016 8:32 pm

Return to Data Issues

Who is online

Users browsing this forum: No registered users and 1 guest