The Grid Aggregation entry type lets you aggregate any number of gridded data files.
The common use case is to aggregate the children grids on the time dimension, so for example,
if you have a number of the same grids but one time step per file you can aggregate these
files and act as if they are one grid.
For example, NOAA provides Max Stream flow grids (in GRIB2 format) available at
[1]. Each of
these GRIB files contains one time step. Go and download the files of interest. Then in RAMADDA,
Select Pick a Type... and look for Grid Aggregation:
You can leave the default settings:
| The grid aggregation works by dyamically constructing a
NCML file.
The Aggregation type, Time Coordinate and Fields are are used to control how the
NCML is configured.
More documentation on this in the future.
|
| Note: The Files, File Pattern, Recurse, etc fields allow an admin
to define that this grid aggregation dynamically loads gridded files from
the server file system. More documentation on this in the future.
|
Then from the new Aggregation entry choose Pick a Type... and select Gridded Data File:
| Note: you can either upload each GRIB file individually or you can zip them up (as below) and upload
the zip file
|
Once done you will have a listing like below. You can access each of the grid files individually
but you can also access subset and OpenDAP services on the aggregation.
Note: There can be some performance issues as RAMADDA has to crack open each individual file when
accessing the aggregation.
You can use the Subset Grid form on the aggregation to subset the aggregated grid and to
publish it back to RAMADDA as a single NetCDF file. This can address some of the performance
issues with with aggregation. Optionally specify spatial and temporal bounds. At the bottom of
the form you can specify a RAMADDA folder to add the new NetCDF grid file to.
