SOFT CALIBRATION FOR BUV OZONE DATA Steven Taylor, SSAI

1. SOFT CALIBRATION FOR BUV OZONE DATA Steven Taylor, SSAI PHILOSOPHIC OVERVIEW GENERAL PROCEDURE AN EXAMPLE FOR SBUV SOME RECOMMENDATIONS OTHER EXAMPLES

2. OVERVIEW In short, for an Identified Calibration Problem apply multiple Validation Tests to consistently demonstrate the problem and then derive Measured Albedo Corrections from Select Data. Internal Validation = use data from only subject instrument External Validation= directly compare with another instrument Soft Calibration is not an empirical correction to the Science Data. All known errors have NOT been fixed by Soft Calibration. (Example: Errors in NOAA-9 Profile Ozone NOT fixed, the Instrument Calibration Problem was NOT identified.)

3. SOFT CALIBRATION GENERAL PROCEDURE Step 1: A Validation Test A validation looks at differences in the Science Data where there are 1) sensitivities to an Identified Calibration Problem and 2) there are Expected Constraints in the differences. Expected Constraints for Internal Validation from 1) proven theory 2) experience with other calibrated instruments 3) a guess (a working hypotheses) Step 2 – Repeatability Tests For the same Identified Calibration Problem repeat Step 1 using other Validation Tests Question: Are all Validation Tests demonstrating the Identified Calibration Problem. If Yes, then Step 3 Step 3 - Make Soft Calibration Correction from Select Data Corrections derived from the Validation most sensitive to calibration error relative to its Expected Constraints.

4. EXAMPLE CORRECTIONS TO "DAY 1 ALBEDO CALIBRATION" FROM OVERLAPPING SBUV A combination of External and Internal Validations Step 1) Albedo Correction for 331 nm (surface reflectivity channel) based on Antarctica Ice. (Internal Validation using theory and experience with EP TOMS) Step 2) Albedo Corrections for 273 to 302nm (Profile Channels) from other instrument profile ozone near the equator accounting for ozone gradients. (External Validation - other Satellite Measurements starting with Space Shuttle SSBUVs) Step 3) Albedo Corrections = 306 to 318nm residues from Algorithm Profile using 273 to 302 nm, at pacific equator, no sea glint, with Corrections 1) and 2). (Internal Validation using theory and ozone algorithm climatology)

5. GENERAL CATEGORIES OF INSTRUMENT PERFORMANCE ERRORS 1) Multiplicative Errors - radiative calibration constants 2) Additive Errors - stray light - PMT non-linearity - thin film contamination Multiplicative Errors can often be fixed with Soft Calibration. Corrections for Additive Errors are often inaccurate/incomplete with remaining Additive Error degrading the measurements. RECOMMENDATIONS Message to the Instrument - No significant Additive Errors Message to the Programs - Spend the money on correcting Additive Errors

6. GENERAL LIST OF SOFT CALIBRATION CORRECTIONS APPLIED N7 SBUV Profile Trend - Ascent/Decent Method Decent translated into Ascent Conditions Total Trend - D-pair Method 306-313nm pair less sensitive to calibration error at equator, low solar zenith NOAA-9 Significant Error in Profile NOT fixed since Instrument Problem could NOT be identified. NOAA-11 PMT Non-linearity - Ice Observations Large signal level change over ice versus solar zenith Day 1 Albedo - per SSBUV Flights

7. N7 TOMS Ozone and Aerosol Trend - Spectral Discriminationand Extrapolation Relatively small radiance error for 313 and 318 extrapolated from 331, 340, 360 and 380nm radiance. EP TOMS - (New Version to be released in the near future) Contamination - mix of Recently Developed Internal Validations plus some External Validations OMI Day 1 Albedo - Recently Developed Internal Validations