Hi Carl,
Thank you for getting back to me so quickly.
The duplicates I'm seeing are in the raw netCDF data from the v4 release. They seem to fall in patterns/groups of three and before moving to a new time. The following is output from IDL after the data is read in using a custom function:
ibtracs = read_ibtracs_v4()
% Compiled module: READ_IBTRACS_V4.
% Loaded DLM: NCDF.
% Compiled module: JULDAY.
IDL> ii = WHERE(ibtracs.season EQ 2009 AND ibtracs.storm_name EQ 'MELOR', cnt)
IDL> cnt
1
IDL> print, ibtracs.storm_time[*,ii]
2455103.5 2455103.5 2455103.8 2455104.0 2455104.0
2455104.0 2455104.2 2455104.5 2455104.5 2455104.5
2455104.8 2455105.0 2455105.0 2455105.0 2455105.2
2455105.5 2455105.5 2455105.5 2455105.8 2455106.0
2455106.0 2455106.0 2455106.2 2455106.5 2455106.5
2455106.5 2455106.8 2455107.0 2455107.0 2455107.0
2455107.2 2455107.5 2455107.5 2455107.5 2455107.8
2455108.0 2455108.0 2455108.0 2455108.2 2455108.5
2455108.5 2455108.5 2455108.8 2455109.0 2455109.0
2455109.0 2455109.2 2455109.5 2455109.5 2455109.5
2455109.8 2455110.0 2455110.0 2455110.0 2455110.2
2455110.5 2455110.5 2455110.5 2455110.8 2455111.0
2455111.0 2455111.0 2455111.2 2455111.5 2455111.5
2455111.5 2455111.8 2455112.0 2455112.0 2455112.0
2455112.2 2455112.5 2455112.5 2455112.5 2455112.8
2455112.8 2455112.8 2455113.0 2455113.0 2455113.0
2455113.2 2455113.5 2455113.5 2455113.5 2455113.8
2455114.0 2455114.0 2455114.0 2455114.2 2455114.5
2455114.5 2455114.5 2455114.8 2455115.0 2455115.0
2455115.0 2455115.2 2455115.5 2455115.5 2455115.5
The times for the storm all have the Julian time of 1858-11-17 added to them as part of the reader function. I have attached the reader function for your reference if needed.
Ultimately, with the duplicates removed, there is little change to the track overall, but the duplicates cause problems when trying to select a single time or interpolating the location/other data to the time of another data point. My solution for now is to just use only the unique time values via the IDL UNIQ function. Let me know what you find out/suggest. Thanks again!
- Kevin Nelson, PhD Candidate, Texas A&M University - Corpus Christi