#######################################
# This README file describes an off-orbit profile issue regarding EE files.
#
#       Last Updated: 2021-02-25        K.Lukens        NOAA/NESDIS/STAR
#######################################

---------------------------------------
In 2019, the following issue was reported to ESA:

	Off-orbit data points have been discovered in the EE files for both FM-A and FM-B data sets (both datasets are original and not reprocessed). It appears to be a longitude assignment issue. The off-orbit points tend to occur around hours 5-6Z, 12Z, and/or 18-19Z each day.

---------------------------------------
Solution:

	"profile_lon_average" is calculated as the average over the latitudes of the wind results which constitute a profile. For these particular "off-points" the longitude values of the single wind results are around the wrap-around value of 360 deg which is why the arithmetic mean does not yield meaningful values. To overcome this issue I would recommend to use the geolocation information of the single wind results, i.e the parameters of the mie/rayleigh_geolocation datasets

	The new L2BP v3.30 was activated for NRT processing at ECMWF on 2-Apr-2020, which solves this issue. Also the upcoming reprocessing will be based on this processor version, so also for this dataset the issue is solved.

	BUFR product is not affected as this problem only occurs when longitude values are averaged to profile values.

Update 2021-02-25:

	Off-orbit points have not been found in datasets with the M1 mirror temperature bias correction applied (Baseline B10 and onward).

#######################################