search

FIDUCEO / FCDR_HIRS

Other
1 stars 2 forks source link

Channel 1 BTs exhibit blocks from calibration #166

Open gerritholl opened 6 years ago

gerritholl commented 6 years ago

Channel 1 BTs exhibit blocks from calibration. This is worse in some cases than in others.

Question: if differences between those blocks are within the structured uncertainty, does that mean it may still be correct from a metrological point of view? Calibration uncertainty tends to be very large for channel 1.

Example from FIDUCEO_FCDR_L1C_HIRS2_noaa06_19790826121452_19790826135552_easy_v0.7_fv0.4 but it's more or less visible in quite many cases.

fiduceo_fcdr_l1c_hirs2_noaa06_19790826121452_19790826135552_easy_v0 7_fv0 4_ch1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

gerritholl commented 6 years ago

Need to check if this is consistent with the large structured uncertainty

gerritholl commented 5 years ago

Friendlier plot (still there in v0.8rc1):

image

plotted with:

hirs_orbit_map /gws/nopw/j04/fiduceo/Data/FCDR/HIRS/v0.8rc1/easy/noaa06/1979/08/26/FIDUCEO_FCDR_L1C_HIRS2_NOAA06_19790826121439_19790826135552_EASY_v0.8rc1_fv2.0.0.nc --channels 1 --without-bitmasks --split --range 30 60
jmittaz commented 5 years ago

Need to see if updated self emission model introduced in pull request #386 helps with this