Hello, I wrote a small utility for parsing NetCDF files, and am not
understanding how the variable parameters [vsize] and [begin] are used when
a record dimension is present. A typical file containing monthly
precipitation has dimensions lat, lon and time, with time being unlimited.
The [begin] values for lat and lon are as expected. The [begin] for precip
is what I would expect -- since time values do not seem to be stored, but
rather are derived from the metadata, I would expect the precip values to
begin immediately after lat and lon. The [vsize] is not at all what I'd
expect. I would think it should be n-lat * n-lon * n-time-records *
sizeof(precip), but instead it is n-lat * n-lon * sizeof(precip). That leads
to the [begin] value for time, which I would not think should even exist.
The [vsize] for time is what the documentation explains, which is
sizeof(time). In short, the header suggests that the precip consists of a
single time interval, and that there is a block of time values. Obviously
I'm misinterpreting the values. I have not been able to find any
documentation that makes this clear. Any direction would be appreciated.
Duane
