Closed BenjaminRuston closed 2 months ago
@smaticka this does warrant a thorough test most concerned with this line that requires the arrays to be in the order of the obs_data.keys()
. Since these keys were renamed think we should be ok, but again should be certain.
@smaticka this does warrant a thorough test most concerned with this line that requires the arrays to be in the order of the
obs_data.keys()
. Since these keys were renamed think we should be ok, but again should be certain.
This shouldn't be a problem since we define the order of the columns here: https://github.com/JCSDA-internal/ioda-converters/pull/1501/files#diff-dbe4cc8b28ed4b9fdc60d3e49d64bd91b7547c9ff88c4de04e7ec35364139bdbL140-L146
So the order won't change if they change their data again. Other things will break though if they change :) (like the variables we search for and store as our own, and as you found, the compatibility when converted 2 versions at the same time)
Description
there have been changes to the wind speed files being ingested from Windborne, this includes changing from wind speed in the
x
andy
tou
andv
.The winds were switched in the API from mis-labeled
speed_x
andspeed_y
tospeed_u
andspeed_v
. The data was backprocssed.However on 20240322T161500Z the files began showing both
speed_u
andspeed_v
as well as thespeed_x
andspeed_y
changes have been made to accommodate these modifications
Issue(s) addressed
Resolves #1500
Impact
correctly decoded balloonsondes, must confirm if the
speed_x
andspeed_y
are now corrected to be the longitudinal and latitudinal windsChecklist