Move historic CDFs to single netCDF file

[lizaclark]

Historic CDFs are currently written (once) to one ascii file per variable per grid cell per day of year (365). A single file is read in per variable for grid cell every day in the real-time system. The Weibull plotting positions are also calculated each day. It should be faster to write these values (once) to a separate netCDF file for each day of year (365) so that only one file needs to be read in in real-time everyday when calculating current percentiles.

This would look something like:

netcdf cdf.mm-dd {
dimensions:
        lon = 245;
        lat = 195;
        rank = 150;
variables:
	double lon(lon) ;
		lon:long_name = "longitude" ;
		lon:units = "degrees_east" ;
		lon:standard_name = "longitude" ;
	double lat(lat) ;
		lat:long_name = "latitude" ;
		lat:units = "degrees_north" ;
		lat:standard_name = "latitude" ;
	int rank(rank) ;
		rank:long_name = "rank" ;
                rank:description = "rank of variable relative to its own time series" ;
		rank:units = " " ;
	double weibull(rank) ;
		weibull:long_name = "weibull_plotting_position" ;
                weibull:description = "calculated as rank/(n+1)"
		weibull:units = "percent" ;
	double swe(lat, lon, rank) ;
		swe:long_name = "snow water equivalent" ;
		swe:units = "mm" ;
	double sm(lat, lon, rank) ;
		swe:long_name = "soil moisture" ;
		swe:units = "mm" ;
	double tm(lat, lon, rank) ;
		swe:long_name = "total moisture" ;
		swe:units = "mm" ;