ERROR: Gaussian_latitude_computation

[meteoipb]

Hi!!
I use the project data ERA-Interim for April 29, 2017 for box with coordinates: latitude = 33 - 44, longitude = 128 - 147 (It's territory of Japan). When working metgrid.exe for the indicated box, the program aborted with the following error:
Processing 2017-04-29_00
FILE
ERROR: Gaussian_latitude_computation
Oops, something is not right with the Gaussian latitude computation.
The input data gave the starting latitude as 43.860.
This routine computed the starting latitude as +- 89.463.
The difference is larger than 0.01 degrees, which is not expected.
2021-09-22 10:58:03.107 --- Getting input fields from FILE
2021-09-22 10:58:03.114 --- Processing TT at level 200100.000000.
2021-09-22 10:58:03.133 --- ERROR: Gaussian_latitude_computation

 

[mgduda]

If I recall correctly, the ungrib program only supports global Gaussian grids -- limited-area subsets of Gaussian grids aren't supported, unfortunately. Does ungrib run successfully if you use the full global dataset?

 

[meteoipb]

Hi!
I work with limited grids and there were no problems. The problem may be that ground level data were missing for this area. But I added it to Metgrid.tbl. (fill_lev= 200100100000). But in the REAL program I did not sent : use_surface=.false. I do not know how to do it . I read about this in forum publications https://forum.mmm.ucar.edu/phpBB3/viewforum.php?f=75. The question is not resolved.

 

[mgduda]

I don't think lack of surface fields would lead to the error

ERROR: Gaussian_latitude_computation
Oops, something is not right with the Gaussian latitude computation.
The input data gave the starting latitude as 43.860.
This routine computed the starting latitude as +- 89.463.
The difference is larger than 0.01 degrees, which is not expected.

If you can do so, I would strongly recommend trying to use the full, global dataset as a test to see whether that resolves the "ERROR: Gaussian_latitude_computation" problem.

When you say that you "work with limited grids and there were no problems", were these other limited-area grids subsets of a Gaussian grid, or were they limited area subsets of other types of grids (e.g., regular latitutde-longitude)?

 

[Ming Chen]

Hi,
I believe Michael is right, --- the ungrib program only supports global Gaussian grids.
if the global data is in regular lat-lon grid, then ungrib can process a subset of the global data.

 

[meteoipb]

Hi!
Oh sure. I worked for the territory of Ukraine, and now for the territory of Japan, surrounded by the ocean. I think that problems are connected with this. I will try to choose the maximum possible data for the surface.