UVvis data and NORS validation server F Hendrick

UV-vis data and NORS validation server F. Hendrick, B. Langerock, M. Van Roozendael, and De Mazière M. Belgian Institute for Space Aeronomy (IASB-BIRA), Brussels, Belgium NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 1

Context • To develop and implement a web-based application for validation of MACCII products using the NORS data products (NORS WP 8) MACC output at 0, 6, 12, 18, 24 h UT NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 2

1. Temporal interpolation For UV-vis data, three different cases need to be considered: • Total ozone columns (90° SZA AM and PM) • Stratospheric NO 2 columns and profiles (90° SZA AM and PM) + diurnal variation • Products from MAX-DOAS observations (every ~30’) NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 3

Temporal interpolation: total ozone • Choose a characteristic time-window (e. g. , 6 h as for FTIR): 6 h 6 h MACC output Time 90°SZA AM observation 90°SZA PM observation NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 4

Temporal interpolation: stratospheric NO 2 90°SZA AM 90°SZA PM MACC-II output Scaling factor x_MACCII(LTDOAS) = x_MACCII (LTMACCII). [ x_model(LTDOAS)/x_model(LTMACCII) ] with x= NO 2 VCD or [NO 2] at altitude z NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 5

NORS NO 2 photochemical correction tool Ø ‘Off-line’ calculation of look-up tables of NO 2 profile diurnal variation using of the BIRA photochemical box model PSCBOX Ø Model calculations for the following scenarios: altitude (24 levels), latitude (18 values covering 85°N→ 85°S by 10° step), longitude set to 0, months 1 -12 Ø For each month, use of the SLIMCAT chemical and meteorological fields corresponding to the middle of the month averaged over a period of 10 years (2000 -2009) Ø Look-up tables of NO 2 profile diurnal variation in ASCII Ø Creation of an interpolation routine in FORTRAN to extract the needed scaling factors NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 6

Sensitivity tests (1) March 15 th chemical field and p, T profiles averaged over 10 years (2000 -2009) March 15 th 2000 NO 2_model(LTDOAS) / NO 2_model(LT) NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 7

Sensitivity tests (2) March 15 th chemical field and p, T profiles averaged over 10 years (2000 -2009) March 15 th 2005 NO 2_model(LTDOAS) / NO 2_model(LT) NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 8

Sensitivity tests (3) denoxification NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 9

NORS NO 2 photochemical correction tool Ø ‘Off-line’ calculation of look-up tables of NO 2 profile diurnal variation Ø Use of the photochemical box model PSCBOX Ø Model calculations for the following scenarios: altitude (24 levels), latitude (18 values covering 85°N→ 85°S by 10° step), longitude set to 0, months 1 -12 Ø For each month, use of the SLIMCAT chemical and meteorological fields corresponding to the middle of the month averaged over a period of 10 years (2000 -2009) Ø For polar regions in late fall, winter, early spring, additional look-up tables are needed for denoxification conditions; the choice of LUTs (denoxification conditions or not) could be based on the measured NO 2 columns NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 10

Temporal interpolation: MAX-DOAS products • Smaller time-window due to the strong short-term variation: 1 h 1 h MACC output Time MAX-DOAS observations (every ~ 30’) NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 11

2. Spatial interpolation • As for FTIR, calculate the location of the effective airmass of every measurement • Interpolate the four surrounding MACC-II model data to this effective airmass location: Is it relevant to proceed like this with UV-vis measurements ? Do we have tools for such calculations ? NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 12

3. Smoothing and vertical gridding Profile comparisons: • MACC-II profile x. MACC must be regridded to the vertical grid of the NORS profile • The MACC-II data is smoothed with the averaging kernel of the NORS measurement using: xs, MACC = xa, NORS + ANORS (x. MACC – xa, NORS) • IF DOFS < 5, one defines partial column boundaries in between which there is about one DOF. Partial column boundaries should be defined by the PI of the instrument NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 13

3. Smoothing and vertical gridding Column comparisons: • MACC-II profile x. MACC must be regridded to the vertical grid of the NORS profile • The MACC-II data is smoothed with the column averaging kernel of the NORS measurement using: Cs, MACC = Ca, NORS + (x. MACC – xa, NORS) • Suitable for OEM-based retrievals but what we do for total ozone and stratospheric NO 2 column retrievals (no a priori data used) ? NORS/NDACC UV-VIS meeting, Brussels, 3 -4 July 2012 14