Page 1 of 1

Terrain data

PostPosted: Mon Oct 16, 2017 8:51 am
by markwang
I am wondering about who had displayed and compared
terrain data (e.g. topo_10m) and WRF terrain height (hgt).
If someone did, please spend a little time to read this email, really appreciate your reply.

The fortran code seems working correctly to read those geography data,
e.g. 00001-00120.00001-00120, ... and ... 02041-02160.00961-01080,
total 162 files for 10m resolution, and 648 files for 30s resolution globally,
with "halo" (tile_bdr=3 in the index file) in each file (or tile).

And, the fortran code wrote out in binary format as:
+ ,convert='little_endian')

Its ctl file is:
dset ^all_topo_10m.bin
undef -999999.99
options sequential
title all_30s_topo
xdef 2160 linear 180.0 0.16666667
ydef 1080 linear -90.0 0.16666667
zdef 1 levels 1000
tdef 1 linear 00Z09JUN2017 6hr
vars 1
h 1 0 terrain height [m]

However, the WRF-terrain seemingly shift eastward of the geog-terrain.
(I can't upload jpg files here......)
Therefore, my question is, did I make something wrong?

Thanks for reading this email.

Re: Terrain data

PostPosted: Tue Oct 17, 2017 11:56 am
by markwang
I have found some errors.
However, after I corrected the errors, the 10m (globally) showed better pattern,
but the 30s (regionally) still remain shifted compared with WRF terrain height.

After test, the correct ctl file (for 10m globally) should be:
xdef 2160 linear 180.083333 0.16666667
ydef 1080 linear -89.916667 0.16666667

I also changed the ctl file (for 30s regionally) to:
xdef 4800 linear 100.00416 0.00833333
ydef 3600 linear 10.00416 0.00833333