A common analysis workflow of the EAM developers/users includes calculating time averages and then remapping the simulation output from a cubed-sphere to a (structured) lat-lon grid. The resulting data files are often referred to as "climo" files. An example can be found at this URL.
In such files, the ncol dimension in an unstructured data file is replaced by a pair lat, lon. For example, the variable float AODDUST(time, lat, lon) in a climo file corresponds to float AODDUST(time, ncol) in the data file on the cubed sphere. The climo files typically contain the following 1D arrays that describe the structured mesh:
double lat(lat) ;
lat:long_name = "Latitude of Grid Cell Centers" ;
lat:standard_name = "latitude" ;
lat:units = "degrees_north" ;
lat:axis = "Y" ;
lat:valid_min = -90. ;
lat:valid_max = 90. ;
lat:bounds = "lat_bnds" ;
double lon(lon) ;
lon:long_name = "Longitude of Grid Cell Centers" ;
lon:standard_name = "longitude" ;
lon:units = "degrees_east" ;
lon:axis = "X" ;
lon:valid_min = 0. ;
lon:valid_max = 360. ;
lon:bounds = "lon_bnds" ;
@berkgeveci , can we add support for such files? We would treat the variables as point data. Thanks!
A common analysis workflow of the EAM developers/users includes calculating time averages and then remapping the simulation output from a cubed-sphere to a (structured) lat-lon grid. The resulting data files are often referred to as "climo" files. An example can be found at this URL.
In such files, the
ncoldimension in an unstructured data file is replaced by a pairlat, lon. For example, the variablefloat AODDUST(time, lat, lon)in a climo file corresponds tofloat AODDUST(time, ncol)in the data file on the cubed sphere. The climo files typically contain the following 1D arrays that describe the structured mesh:@berkgeveci , can we add support for such files? We would treat the variables as point data. Thanks!