Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Relationship between spectral and grid resolution

IFS is a model of two resolutions: the spectral resolution determining the number of retained waves and the corresponding grid onto which these waves can be transformed. A greater spectral resolution can improve the forecast quality whilst reduced Gaussian grid resolutions can improve the efficiency of the model. However, aliassing of the quadratic (and cubic) terms in the model equations can result from using linear grids, which needs to be compensated for in the model formulation. A cubic grid removes aliassing and the need for additional filters.

Model timestep

The timestep depends on the horizontal resolution and the choice of the grid. ECMWF typically use a long a timestep as possible for efficiency reasons. However, there are times when this may result in some of the semi-Lagrangian trajectories going underground (messages can be seen in the model logfile), though generally this is not a problem.

Recommended maximum timesteps are given in the table below. Shorter timesteps can be used, though note that the model will not produce the exact same result between two forecasts with different timesteps.

Supported spectral and horizontal grid resolutions

The table below summarizes the supported configurations of the IFS and the relationship between spectral truncation, gridpoint resolution and the recommended timestep for optimum performance.

Info

Click in the column header to display the table according to that column.


Spectral truncationGrid
(lats pole to eq)
Grid typeResolution at equator
(npts at eq. / spacing)
Recommended
timestep (min)

Application

Tco1279O1280_45136 / 8km / 0.070o7.5Operational forecast resolution (43r3)
Tl1279N640l_22560 / 16km / 0.141o10Operational forecast resolution (40r1)

Tl1023

N512l_22048 / 20km / 0.176o10-
Tco799O800_43216 / 12.5km / 0.112o12-
Tl799N400l_21600 / 25km / 0.225o12-
Tco639O640_42576 / 16km / 0.140o12Operational ensemble resolution (43r3)
Tl639N320l_21280 / 31km / 0.281o15ERA-5 high resolution
Tl511N256l_21024 / 39km / 0.352o15-
Tco399O400_41616 / 25km / 0.223o15-
Tl399N200l_2800 / 50km / 0.45o20Operational ensemble resolution (40r1)
Tco319O320_41296 / 31km / 0.278o20-
Tl319N160l_2640 / 63km / 0.563o20ERA-5 ensemble
Tco255O256_41040 / 39km / 0.346o
20-
Tl255N128l_2512 / 78km / 0.703o45ERA-Interim
Tq213N160_2640 / 63km / 0.563o20-
Tco199O200_4816 / 49km / 0.44o20-
Tco159O160_4656 / 61km / 0.55o20-
Tl159N80l_2320 / 125km / 1.125o60ERA-40
Tq106F80_2320 / 125km / 1.125o60-
Tco95O96_4400 / 100km / 0.9o2030-
Tl95N48l_2192 / 209km / 1.875o60-
Tq63F48_2192 / 209km / 1.875o60-
Tq42F32_full128 / 313km / 2.813o30Development/testing only
Tq21F16_full64 / 626km / 5.625o30Development/testing only



Info
  • Grid type refers to the file suffix used at the end of filenames for the OpenIFS/IFS climatology files. It is also used as an argument to the 'gaussgr' command, see 4.5 OpenIFS: Grid tools.
  • The 'cubic' grid rather than 'cubic octahedral' is not listed here as it's infrequently used, the suffix for these files is '_3'.
  • T21 and T42 use the Eulerian dynamical core and a regular Gaussian grid.
  • Tco denotes cubic octahedral grid, Tl denotes a linear grid, Tq denotes a quadratic grid.
  • Other resolutions not listed here may be provided with the OpenIFS climatology files.


Note that a linear grid allows a higher spectral truncation for the same number of gridpoints as a quadratic or cubic grid. Linear and quadratic in this sense means enough gridpoints are available to compute the linear and quadratic terms in the equations respectively. Likewise a cubic grid has more gridpoints per spectral wavenumber and provides more gridpoints to describe the highest wavenumbers.



HTML Comment
hiddentrue

this table is taken from the draft OpenIFS 40r1 paper & supplemented with 43r3 info for the cubic grids.


% this should match the resolutions of the climate data provided by OpenIFS.
% see  bin/gaussgr for further definitions of the grids.

% for resolution information, look at prepIFS defaults in:
% ~rdx/prepIFS.defaults/current/40r1/setup_fc/resolution
% gives (ignoring some resolutions, only including those for which we have climate files):

Look in /home/rd/rdx/data/ifs/rtable* to confirm no. of long pts at lats closest to eq (see Paul's message below).
% T21, _full, Eulerian dyn., 1800sec, 32x64 % T42, _full, Eul , 1800sec, 64x128 % T63, _2 , Semi Lagr. , 3600 , 96x192 % T95, l_2 , SL , 3600 , 96x192 % T106, _2 , SL? , 3600 , 160x320 (LSLAG not set by 106.setup.save?) % T159, l_2 , SL , 3600 , 160x320 % T213, _2 , SL , 1200 , 320x640 % T255, l_2 , SL , 2700 , 256x512 % T319, l_2 , SL , 1200 , 320x640 % T399, l_2 , SL , 1200 , 400x800 % T511, l_2 , SL , 900 , 512x1024 % T639, l_2 , SL , 900 , 640x1280 % T799, l_2 , SL , 720 , 800x1600 % T1023, l_2 , SL , 600 , 1024x2048 % T1279, l_2 , SL , 600 , 1280x2560 % % _full, etc refer to the suffix on the model climate files, where _full mean regular (quadratic) Gaussian grid % _2 means reduced regular grid, l_2 means reduced linear grid.

FOr 43r3, also look in the prepIFS default directory.
However, there is a discrepancy between the namfpd@NLON values in those files and the
formula on the octahedral grid page: Nlat = 4 x i + 16 where i=1,.. N.
For an O1280 grid (Tco1279) -> Nlat = 4 x 1280 + 16 = 5136. This doesn't match namfpd@NLON = 5120.
I've used the 4i+16 formula in working out the NLONeq values above for Tco grids.

Checking with Paul Dando (the oracle of all things grids...)

The formula for the number of longitude lines at latitude=i = 4*i + 16 (starting with i=1 nearest the pole) is definitely correct.  So the latitude nearest the equator for the O1280 grid has 4*1280 + 16 = 5136 longitude points.

In fact, you can check this in the NAMGRI namelist (read from, e.g. /home/rd/rdx/data/ifs/rtable_41279).  Here you will see that line nearest the pole has 20 points and those nearest the equator have 5136 points

My conclusion is, therefore, that namfpd has the wrong number for the octahedral grids.

Of course, I don't know what namfpd@NLON refers to in fullpos.  It could be that this is the number of longituide points on the full Gaussian grid (which is always 4*N) ?

Or perhaps it's just not been updated and still refers to the old-style reduced grids which did have 5120 points at the latitudes nearest the equator.

I hope that helps.






Excerpt Include
Credits
Credits
nopaneltrue



Panel
borderColorgrey

On this page ...

Table of Contents
indent12px


Panel
titleIn this section...

Children Display
depth1
pageOIFS:4. OpenIFS: Grid and Resolution


Panel
titleOpenIFS User Guide ...

Children Display
depth1
pageOpenIFS User Guide