|
Hi all, Golla Nageswararao wrote about a dataset which is a big collection of profile data at a set of global locations and sampled over depths. We had taken this discussion off-line and found a way to sample the dataset so there is a collection only of the Indian Ocean data. However the question remains of what to do with a set of scattered XY locations, each with a profile that does not contain the same set of depths. There seems to be more and more of this kind of data that people are trying to work with. It is not gridded data and it doesn't necessarily make sense to try to interpolate it onto a grid. The CF standard for netCDF datasets has added some sections on organizing such data collections under the heading Discrete Sampling Geometries, including collections of time series, collections of profiles, and so forth. Please see Chapter 9 of the documentation, http://cf-pcmdi.llnl.gov/documents/cf-conventions/1.6/cf-conventions.html#discrete-sampling-geometries and the examples, shown in Appendix H of the same document http://cf-pcmdi.llnl.gov/documents/cf-conventions/1.6/cf-conventions.html#appendix-examples-discrete-geometries We have upcoming projects in Ferret and LAS development to work with this sort of data. We'll want to write these kinds of Discrete Geometries files Ferret, and to come up with some tools to work with them. This will include ways to subsample sets of profiles or time series from a big collection, and new ways to visualize this kind of data. As always we encourage contributions of Ferret scripts, Python tools that might be incorporated using PyFerret, and also general ideas for handling this sort of data. So if you're thinking about non-gridded data, please lend your ideas. Ansley On 4/17/2013 4:36 AM, golla
nageswararao wrote:
|