Post on 01-Apr-2015
UPDATE ON SMOS LONG-TERM BIASES OVER THE OCEAN AND ROUGH SURFACE SCATTERING
OF CELESTIAL SKY NOISE
Joe TenerelliSMOS L2OS Progress Meeting
Arles, France, September 26, 2011
SMOS PASSES USED IN THIS ANALYSIS
DRIFT IN 2010 ONLY
NO LOSS MODEL
ONE-SLOPE LOSS MODEL
TWO-SLOPE LOSS MODEL
EXTENSION THROUGH JUNE 2011
NO LOSS MODEL
ONE-SLOPE LOSS MODELNotice the significant desc-asc bias now appearing with both one and two slope loss models between Nov 2010 and April 2011. It is of opposite sign to that between Sep and Dec 2010 observed with no loss model. It appears that the loss correction is too strong in this period.
TWO-SLOPE LOSS MODELThe two-slope model seems to degrade desc-asc bias relative to the one-slope model.
ONE-SLOPE LOSS MODEL: IMPACT OF GALACTIC REFLECTION MODELIt turns out that, for the metrics used here, the impact of the formulation of galactic noise scattering at the surface on desc-asc bias is quite small even for descending passes in September.
Kirchhoff scattering model for galactic noise
ONE-SLOPE LOSS MODEL: IMPACT OF GALACTIC REFLECTION MODELIt turns out that, for the metrics used here, the impact of the formulation of galactic noise scattering at the surface on desc-asc bias is quite small even for descending passes in September.
Flat surface reflection model for galactic noise
DRIFT METRICS
LONG TERM DRIFT JAN-DEC 2010
LONG TERM DRIFT JUN-DEC 2010
LONG TERM DRIFT JUNE 2010 – JUNE 2011
Considering the June 2010 to June 2011 period, there is very little net drift in the AF and EAF FoVsIn the ascending passes.
SHORT TERM DRIFT IN 2010 AND 2011
The two-slope model seems to degrade desc-asc bias relative to the one-slope model.
Two-slope degrades desc-asc consistency
LATEST OPERATIONAL BIAS CURVES
The following are bias curves for operational DPGS Level 1B data:
The following are bias curves for operational DPGS Level 1A data and using JRECON with no direct sun correction:
Comparing Original DPGS with L1PP
NO LOSS MODEL
LATEST OPERATIONAL BIAS CURVES
UPDATE ON GALACTIC NOISE MODEL
UPDATE ON GALACTIC NOISE MODEL
UPDATE ON GALACTIC NOISE MODEL
UPDATE ON GALACTIC NOISE MODEL
UPDATE ON GALACTIC NOISE MODEL
CELESTIAL SKY NOISE MAPS
HOVMOLLER PLOTS OF REFLECTED SKY NOISE
SMOS RESIDUALS BASED ON COMMISSIONING REPROCESSING DATA
SMOS RESIDUALS BASED ON COMMISSIONING REPROCESSING DATA
FITTING A GEOMETRIC OPTICS MODEL TO THE RESIDUALS
FITTING A GEOMETRIC OPTICS MODEL TO THE RESIDUALS
FITTING A GEOMETRIC OPTICS MODEL TO THE RESIDUALS
FITTING A GEOMETRIC OPTICS MODEL TO THE RESIDUALS
FITTING A GEOMETRIC OPTICS MODEL TO THE RESIDUALS
CROSS SECTIONS THROUGH THE CELESTIAL SKY MAPS
CROSS SECTIONS THROUGH THE CELESTIAL SKY MAPS
CROSS SECTIONS THROUGH THE CELESTIAL SKY MAPS
CROSS SECTIONS THROUGH THE CELESTIAL SKY MAPS
CROSS SECTIONS THROUGH THE CELESTIAL SKY MAPS
NEW SEMI-EMPIRICAL MODEL
CROSS SECTIONS THROUGH THE CELESTIAL SKY MAPS
ORIGINAL KIRCHHOFF MODEL
CROSS SECTIONS THROUGH THE CELESTIAL SKY MAPS
NEW SEMI-EMPIRICAL MODEL: Ascending Passes
CROSS SECTIONS THROUGH THE CELESTIAL SKY MAPS
NEW SEMI-EMPIRICAL MODEL: Descending passes
IMPACT ON CELESTIAL SKY BIASES
BIAS: ORIGINAL KIRCHHOFF MODEL
BIAS: KIRCHHOFF MODEL EVALUATED AT 3 M/S WIND SPEED
BIAS: EMPIRICAL GEOMETRIC OPTIC
NEW SEMI-EMPIRICAL GO MODEL SWATH RESULTS
IMPACT ON SWATH BIASES: ALIAS-FREE FIELD OF VIEW
IMPACT ON SWATH BIASES: ALIAS-FREE FIELD OF VIEW
IMPACT ON SWATH BIASES: ALIAS-FREE FIELD OF VIEW
IMPACT OF GALACTIC MODEL ON BIAS TREND ANALYSIS
IMPACT OF GALACTIC NOISE MODEL ON BIAS TREND ANALYSIS
The curves show the contribution of scattered celestial sky noise to FoV biases, based upon the Kirchhoff model evaluated at a 3 m/s wind speed.
IMPACT OF GALACTIC NOISE MODEL ON BIAS TREND ANALYSIS
Similar to previous slide but for the flat surface reflected celestial sky noise.
IMPACT OF GALACTIC NOISE MODEL ON BIAS TREND ANALYSIS
The bias curves below were computed using the original Kirchhoff model evaluated at a 3 m/s wind speed.
IMPACT OF GALACTIC NOISE MODEL ON BIAS TREND ANALYSIS
With the new empirical geometric optics model the curves change very little.
IMPACT OF GALACTIC NOISE MODEL ON BIAS TREND ANALYSIS
Even switching to flat surface reflected galactic noise makes little difference:
IMPACT OF GALACTIC NOISE MODEL ON BIAS TREND ANALYSIS
For these curves we have used the original theoretical roughness emission model. This also makes little difference.
IMPACT OF GALACTIC NOISE MODEL ON BIAS TREND ANALYSIS
IMPACT OF GALACTIC NOISE MODEL ON BIAS TREND ANALYSIS