1. Atmospheric Chemistry Products from the Ozone Mapping and Profiler Suite (OMPS): Validation and Applications NOAA Satellite Conference 2015 L. Flynn with contributions from the NOAA and NASA OMPS S-NPP Teams NSC2015_Session_2.3d_Flynn

2. Outline Instrument Overview and Performance Nadir Mapper Daily Global Products – Climate-quality Total Column Ozone (climate, monitoring, assimilation, UV Index) – UV-absorbing Aerosol Index (smoke, dust, volcanic ash) – Atmospheric SO2 for Hazards and Air Quality (volcanic warnings, forecasts, inventories, campaigns) Nadir Ozone Profile Products – Ozone vertical profiles for middle and upper stratosphere (monitoring, assimilation) – Solar Mg II index and spectral variations Limb Ozone Profile Products – High vertical resolution stratospheric Ozone profiles (Antarctic Ozone Hole monitoring, assimilation, climate) – High vertical resolution stratospheric Aerosol profiles (monitoring, transport) Research Products

3. Abstract The Ozone Mapping and Profiler Suite (OMPS) is designed to make measurements of scattered solar UV and visible radiance with three separate CCD detectors. The measurements contain information on atmospheric ozone and other trace gases. This talk reports on the performance of the instruments and the validation of the ozone products with particular attention to their ability to provide atmospheric chemistry products for use in ozone monitoring, air quality and aviation hazard applications. The OMPS measurements are well on the way to being fully validated with on-board calibration system performing as desired and thorough characterizations of stray light, wavelength scale shift, throughput degradation, non-linearity and dark current to provide models to correct for these effects. The Version 8 Ozone Profile and Total Ozone Algorithms used to create the earlier parts of the corresponding records have been adapted to produce similar products from OMPS measurements. The Version 8 Total Ozone Algorithms produces residuals used by a Linear Fit SO2 retrieval algorithm. Comparisons with NOAA-19 SBUV/2 and EOS Aura OMI products show excellent consistency. The OMPS Limb Profiler products are advancing with improved modeling and better estimates of stray light and pointing errors.

4. Nadir Mapper & Profiler Limb Profiler Main Electronics Each instrument can view the Earth or either of two solar diffusers; a working and a reference. The instruments measure radiance scattered from the Earth’s atmosphere and surface. The detectors are 2-D CCD arrays with one spectral and one spatial dimension. They also make solar measurements using pairs of diffusers. Judicious operation of working and reference diffusers allows analysts to track the diffuser degradation. The solar measurements also provide checks on the wavelength scale and bandpass. The instruments regularly conduct their internal dark and nonlinearity calibration sequences and have completed a full three years of solar measurements. Earth Mode Solar Mode Diffuser Entrance Aperture Diagram from Ball Aerospace and Technology Corporation 4

5. OMPS Ozone Mapping & Profiler Suite Global daily monitoring of three dimensional distribution of ozone and other atmospheric constituents. Continues the NOAA SBUV/2, EOS-AURA OMI and SOLSE/LORE records. Nadir Mapper (NM) Grating spectrometer, 2-D CCD 110 deg. cross track, 300 nm to 380 nm spectral, 1.1nm FWHM bandpass Nadir Profiler (NP) Grating spectrometer, 2-D CCD Nadir view, 250 km cross track, 250 nm to 310 nm spectral, 1.1 nm FWHM bandpass Limb Profiler (LP) Prism spectrometer, 2-D CCD Three vertical slits, -20 to 80 km, 290 nm to 1000 nm The calibration systems use pairs of working and reference solar diffusers. L. Flynn 5

6. 6 Comparison of NASA-derived and NOAA-derived solar diffuser measurement trends for the for the OMPS Nadir Profile sensor. NASA estimates include (a) correction for wavelength shifts determined independently and (b) correction for solar activity determined using the Mg II proxy and associated scale factors for each wavelength. NOAA estimates are calculated by simultaneously solving for diffuser trends, wavelength shift and solar activity represented by three separate terms in a multiple regression fit. Both sets of results agree well and the instrument shows excellent stability with throughput degradation less than 0.5%/Year even for the shortest channels for the reference diffuser measurements. OMPS Nadir Profiler Solar measurements show small degradation rates for diffusers and optics.

7. Description of analysis and results for NP Solar Measurements The OMPS Nadir Profiler Solar measurements were analyzed to investigate changes over the first year of measurements. A total of thirty-two spectra with three from the reference diffuser (on March 21, 2012, August 31, 2012 and April 1, 2013) and 29 from the working diffuser (taken approximately every two weeks from March 7, 2012 to April 1, 2013) were used in the study. The spectra are taken at 147 wavelengths from 250 to 310 nm. For each measurement for each wavelength, the values are averaged over the spatial dimension of the CCD array (100 pixels). The initial analysis found three types of changes in the spectra as follows: those related to solar activity as we pass through the peak of Solar Cycle 24; those related to wavelength scale shifts most likely produced by annual variations in the instrument optical bench or aperture temperatures at the orbital locations of the measurements; and decreases in the measured irradiance values probably produced by uncorrected changes in the instrument throughput or diffuser properties over time. A model of the measurements was devised using three assumptions corresponding to the observed variations. For the first component, it was assume that the solar irradiance variations could be modeled by using a Mg II core-to-wing ratio and scale factors as described in Deland and Cebula (1993). That is, that a ratio of the measurements at the center of the Mg II feature at 280 nm to the measurements at the wings on either side would be a good proxy for solar activity and that a single wavelength pattern of scale factors could accurately predict the variations of a full spectrum from a variation in the Mg II Index. For the second component, it was assumed that relative shape of the variations produced by a spectral scale shift could be found by examining the average solar spectrum variations. Specifically, three consecutive measurements of the average spectrum were fit with a quadratic and the slope of the quadratic at the central point was used as an estimate of the effects of shift at that wavelength. These were normalized by the average spectrum values and the wavelength spacing to produce a relative change pattern with units of 0.4 nm^-1 (1/pixel spectral spacing). For the trend components simple linear trends in time were used with separate terms for the working and reference measurements. (Exponential models are often used to represent degradation but for small changes these are well approximated by linear terms.) The working diffuser exposure time grew at fairly consistent rate over this time period as there was a regular schedule of measurements. This means that calendar time is a good proxy for exposure time. Two sets of constants terms were also included as the working and reference spectra show a small, wavelength-dependent pattern of offsets. The model components are estimate in three steps. First a set of Mg II Index values are computed by fitting quadratics to three subsets of five irradiance values each centered at 277, 280 and 283 nm. The irradiance values at the vertices of these are used as the lower wing, core and upper wing values in the index. This gives two sets of indices. Mg2W(t) and Mg2R(t) for the working and reference diffusers on day t. (The wavelength coordinate of the vertices for the central fit tracks well with the wavelength scale shift values as found with the approach in the next paragraph. There are some small differences in the results depending on whether one removes the wavelength scale shifts before finding the Mg II Index values or removes the solar activity before finding the wavelength shifts.) The Mg II Index variations from OMPS also track well with variations of Mg II Indices from other satellite sources, e.g., MetOp-A GOME-2, over this period. Next a wavelength scale shift relative to the average spectrum is found by taking the relative differences of each spectrum with the average and finding the least squares fit of those differences with the spectral shift pattern described above. The fit is then removed to create a set of spectra with consistent wavelength scales. (You could include figure showing annual cycle in shift quantities in units of 1/pixel separation) SolarWS(w,t) = SolarW(w,t) – Shift(t)*Shift(w) SolarRS(w,t) = SolarR(w,t) – Shift(t)*Shift(w) Where SolarW(w,t) is the normalized irradiance spectrum for the working diffuser on day t (measured since January 1, 2012) and SolarR(w,t) is the same for the reference diffuser. These adjusted relative spectra were now used in multiple linear regression models of the form SolarWS(w,t) = c1(w)*t + c2(w)*Mg2W(t) + c3(w) SolarRS(w,t) = c4(w)*t + c2(w)*Mg2W(t) + c5(w) to obtain estimates of the trends and scale factors. This model is applied for each wavelength, w, individually. Note that the solar activity terms are shared but the trends and offsets are found independently for each diffuser. Ref: Matthew T. DeLand, Richard P. Cebula, “Composite Mg II solar activity index for solar cycles 21 and 22,” Journal of Geophysical Research: Atmospheres (1984–2012), Volume 98, Issue D7, pages 12809–12823, 20 July 1993, DOI: 10.1029/93JD00421 7

8. Description of results in NP_SOL_fit.pdf The first analysis was designed to find the scale factors, the wavelength shifts, and the trends in the Working Diffuser Solar measurements. We can now use the scale factors and the wavelength shift pattern to see how the reference diffuser measurements compare. (We can also check to see if the wavelength shifts are in common between the working and reference.) I agree that even without this analysis the reference spectra evolution looks fairly flat which implies that the trend is mainly from changes in the Working Diffuser, not the instrument throughput. That may not be the entire case. Since the reference measurements are paired with working measurements close in time, we can also compute the relative trends with just two pairs as is done in the second set of results in the figure on bottom of Page 3. We've continued the analysis with both working and reference. We first normalize the spectra by using the average of the working measurements Nsolar(w,t) = Solar(w,t)/AWSolar(w) - 1. We then compute an estimate of the wavelength shift pattern by using quadratic fits of three consecutive AWSolar(w) values and taking the slope at the midpoint as an estimate of the change in radiance per pixel shift. We then normalize these value to make them relative shifts. This gives us a shift(w) pattern. We then fit each Nsolar(w) with shift(w) to get a single shift estimate for each full spectrum relative to the average one. The first page of the attached files shows the relative shift pattern as %/one-pixel-shift on the top plot, and the shifts for each spectrum on the bottom plot. The three *'s are for the Reference Diffuser. There is good agreement in the timing of the shifts between the Working, +'s, and Reference even though the Working spectra are used to find the normalizing spectrum and shift pattern. Notice the suggestion of an annual cycle in the wavelength shifts. Multiple regression is now used for the normalized, shift-removed spectra including both Working and Reference. The model is NSRSolar(w,t) = HW*c1(w)*t + c2(w)*mg2(t) + HW*c3(w) + HR*c4(w)*t + HR*c5(w) where HW is 1 for working diffuse measurements and 0 for reference and HR is the opposite, and the Mg2 Indices are proportional changes relative to the average (the Mg II indices for the Reference were adjusted to remove a bias with the working index computations.) The Mg II scale factors, c2(w) term from the regression model, are in the top plot on Page 2. The Mg II indices, mg2(t) term from the model, are in the bottom plot. Notice that the Reference Diffuser measurements are taken at times with very different solar activity. The third page shows the trend terms from the model on the top figure. The upper line is the Reference Diffuser trend term, c4(w), and the lower line is the Working Diffuser trend term, c1(w). If we assume that the reference diffuser is stable, then the c4(w) pattern represents instrument throughput changes. The lower plot shows the difference in the Working and Reference trends, c1(w)-c4(w). The dotted line is the simple direct calculation one can do with double difference of the Day 81 paired diffuser measurements versus the Day 450 pair, namely, Work(w,450)/Work(w,81) - Ref(w,450)/Ref(w,81). The fourth page just shows the two constant terms from the multiple regression; working, c3(w), on the top and reference, c5(w), on the bottom. These are related to the choice of a Day 0, the use of the Mg II Index average to get deviations, and the use of the Working diffuser average for the normalization. The plot on the top of the fifth page shows the measurements and model results for the 11th wavelength. All the spectra were normalized by the average for the 29 working measurements. The data values in this figure are also after the wavelength shift adjustments have been applied. While one would expect that the three reference measurements (+ measured; * modeled) shown here would give a slightly positive trend, the two later measurements occur at peaks in the solar activity. The dashed lines show the measured data if we adjust for solar activity, that is, subtract the c2(w)*mg2(t) term. The change for the three reference data points is dramatic. The bottom figure on the fifth page shows sample spectra and models for the working diffuser. The spectra have been normalized relative to the working average and the wavelength shift adjustment has been applied. The sixth page shows sample spectra and models for the working diffuser (top figure) and reference diffuser (bottom figure). Both spectra are normalized but without the wavelength shift applied. They are compared to the model plus the fit wavelength shift. The model in the lower figures includes the separate offset (to account for the use of the working average as normalization) and trend calculations for the reference spectra. 8

9. Wavelength Shift Effects < 1%Solar Activity Effects < 1% All Variations Instrument and Working Diffuser Degradation 9 Retrieval Channel Locations Multiple Linear Analysis of NP Solar Measurements for the Working Diffuser First Last

10. OMPS NP Working Solar residuals after fits 10

11. Spectral Patterns and Their Temporal Coefficients 11 Solar activity pattern

12. 12 Comparisons among Total Column Ozone Products from MetOp-B GOME-2 (NOAA Version 8 algorithm), NASA EOS Aura OMI (NASA Version 8.6 algorithm) and S-NPP OMPS-NM (NOAA Version 8 algorithm) for November 2, 2014.

13. OMPS Aerosol Index overlays on VIIRS RGB ozoneaq.gsfc.nasa.gov/omps/blog/ False color Aerosol Index data (with varying scales) plotted over VIIRS RGB. Top Right: Dust from deserts over Africa and the Middle East. Bottom Right: Haze/smoke from agricultural burning over India. Bottom Left: Smoke from wild fires in Australia. Notice the capability of OMPS to retrieve aerosols over clouds.

14. 14 High-Spatial-Resolution Capabilities of OMPS The image on the left shows a false color map of the OMPS effective reflectivity (from a single Ultraviolet channel at 380 nm) over the Arabian Peninsula region for January 30, 2012 when the instrument was making a set of high-spatial-resolution measurements with 5×10 km 2 FOVs at nadir. The color scale intervals range from 0 to 2 % in dark blue to 18 to 20 % in yellow. The image on the right is an Aqua Moderate Resolution Imaging Spectroradiometer (MODIS) Red-Green-Blue image for the same day. The OMPS Nadir Mapper instrument is very stable, extremely flexible, and has excellent SNRs.

16. SO2 estimates for the Mount Kelud eruption Top Right: NASA OMPS product Bottom Right: NOAA OMPS product Bottom Left: NASA OMI product Comparisons of NASA and NOAA, OMPS and OMI SO2 Indonesia volcano eruption closes three international airports

17. 17 OMPS NM measurements can be used to make state-of-the-art SO 2, NO 2 and Aerosol retrievals for air quality and hazard applications. Examples below are for Asia for 10/20/2013 (top) & 10/23/2013 (bottom) SO 2 (DU) NO 2 (10 15 cm- 2 ) UV AI b

18. OMPS SO 2 Measurements SO 2 (DU)

19. OMPS NO 2 Measurements NO 2 (10 15 cm- 2 )

20. OMPS UV Aerosol Index UV AI b

21. Chasing Orbit Comparisons to SBUV/2 Approximately every 12 days, the orbital tracks for the NOAA-19 and S-NPP spacecrafts align and allow comparisons of products for similar locations with small viewing time differences. The top figure shows convergence of the orbital paths. Products and residuals from the same retrieval algorithms for SBUV/2 and OMPS NP can be compared directly. The bottom figures shows ozone amounts for nine layers for the two Version 8 retrievals with the top left for the lowest layer and the bottom right for the highest layer. Additional monitoring plots provided at http://www.star.nesdis.noaa.gov/icvs/prodD emos/proOMPSbeta.O3PRO_V8.php show that the ozone profile differences are consistent with the initial measurement residuals computed relative to the first guess profiles. 21 4.0 hPa 1.0 hPa 16 hPa 2.5 hPa 0.64 hPa 10 hPa 1.6 hPa 0.4 hPa 6.4 hPa

22. SBUV/2 NH Assimilation Results Ozone Mixing Ratio - 2 hPa : Aug 27, 2014 SBUV/2 SH OMPS NH OMPS SH Comparisons between application results for assimilated SBUV/2 and OMPS NP products at NCEP. OMPS products are provide in NRT in BUFR files from NDE.

23. Wavelength Shift Effects < 1%Solar Activity Effects < 1% All Variations Instrument and Working Diffuser Degradation 23 Retrieval Channel Locations Multiple Linear Analysis of NP Solar Measurements for the Working Diffuser First Last

24. OMPS NP: Error Impacts on Precision/Accuracy The sensitivity of the ozone retrievals to radiance/irradiance ratio errors is approximately 1.6%::1%. Wavelength scale errors produce radiance variations of ±1% – 1.6%/1% x 1% = 1.6% O3 effects and ozone cross-section, alpha, of ±0.4 %, – 0.02 nm x 100%/5 nm x 1%/1% = 0.4% O3 effects Solar activity produces irradiance variations of ±1% – 1.6%/1% x 1% = 1.6% O3 effects Instrument degradation is -0.5%/year at 253 nm – 1.6%/1% x 0.5%/year = 0.8%/year O3 effects (assuming annual updates to Calibration Factor Earth tables) Stray light errors are now approximately 1/3 of the original errors with radiance variations of ±1% – 1.6%/1% x 1% = 1.6% O3 effects 24

25. Comparison of OMPS Total Column Ozone and Limb Profile Ozone, by C. Seftor, NASA GSFC Outside and Inside the Antarctic Ozone Hole The OMPS Limb Profiler has 1-KM reporting and 3-km vertical resolution.

26. Research Products Combined UV/IR Ozone Profile retrievals (OMPS+CrIS) – Simple top+bottom approach in operations -- TOAST product Global UV reflectivity (OMPS NM) UV Cloud Optical Centroids (OMPS NM) Column NO2 for Air Quality (OMPS NM) Combined UV/Visible Aerosol retrievals (OMPS+VIIRS) Tropospheric Ozone Estimates (Total Column - Limb, DCC) Polar Mesospheric Clouds (OMPS NM) Polar Stratospheric Clouds (OMPS LP) Stratospheric Rocket Plumes and Meteor Plumes (OMPS LP Stratospheric NO2 Profiles (OMPS LP) Upper atmosphere Temperature/Pressure profiles (OMPS LP)

27. Summary & Conclusions The on-board monitoring systems are providing good characterizations of the time-dependent changes. The heritage Version 8 algorithm products are more than capable of continuing the Climate Data records and provide continuation of SBUV/2 records. The Linear Fit SO2 algorithm provides good estimates by using the Version 8 residuals. The Limb Profiler products are progressing and are achieving high-vertical resolution performance for ozone and aerosol profiles. We are just beginning to explore the real-time applications for the OMPS aerosol products. Additional products are in the research queue.

28. Support Acknowledgement The implementations at NOAA of the Version 8 algorithms for use with OMPS were supported by the NCDC Climate Data Record Program and the JPSS Program. The implementations at NASA were supported by the MEASURES Program and the OMPS S-NPP Science Team. The OMPS Limb Ozone Profile retrieval development and validation is primarily the work of the NASA OMPS S-NPP Science Team.

29. Useful OMPS Resources Instrument Description and Recent Publications – Dittman et al., 2002, “Nadir Ultraviolet Imaging Spectrometer for the NPOESS OMPS,” SPIE 4814. DOI:10.1117/12.453748 – Flynn et al., 2014, ”Performance of the Ozone Mapping and Profiler Suite (OMPS) products,” J. Geophys. Res. Atmos., 119, 6181–6195, doi:10.1002/2013JD020467.10.1002/2013JD020467 – Seftor, et al. 2014, Postlaunch performance of the Suomi National Polar- orbiting Partnership Ozone Mapping and Profiler Suite (OMPS) nadir sensors, J. Geophys. Res. Atmos., 119, 4413–4428, doi:10.1002/2013JD020472.10.1002/2013JD020472 Operational Algorithm and Algorithm Theoretical Basis Documents and Status Briefings – http://star.nesdis.noaa.gov/jpss/ATBD.php http://star.nesdis.noaa.gov/jpss/ATBD.php – http://npp.gsfc.nasa.gov/documents.html http://npp.gsfc.nasa.gov/documents.html – http://www.star.nesdis.noaa.gov/star/meetings.php http://www.star.nesdis.noaa.gov/star/meetings.php Monitoring – http://www.star.nesdis.noaa.gov/icvs/prodDemos/proComparison.php http://www.star.nesdis.noaa.gov/icvs/prodDemos/proComparison.php – http://www.star.nesdis.noaa.gov/icvs/status_NPP_OMPS_NM.php http://www.star.nesdis.noaa.gov/icvs/status_NPP_OMPS_NM.php Data Sets – http://www.nsof.class.noaa.gov/saa/products/welcome http://www.nsof.class.noaa.gov/saa/products/welcome Direct Broadcast Programs – http://directreadout.sci.gsfc.nasa.gov/?id=software http://directreadout.sci.gsfc.nasa.gov/?id=software

30. Additional Product and Validation Examples

31. 31 Comparison of TACO (OMPS and CrIS) with TOAST (SBUV/2 and TOVS/HIRS) Combined products use UV retrievals for the stratosphere and IR retrievals for the Troposphere

32. SATELLITE TANGENT HEIGHT LINE OF SIGHT SOLAR IRRADIANCE SURFACE CLOUDS MULTIPLE SCATTERING SINGLE SCATTERING Limb Scatter Schematic Sensitivity of the 305-nm (left) and 600-nm (right) channel limb radiances to layer ozone changes for 1-km layers for a 40º SZA and mid- latitude 325-DU profile. Each curve gives the ratio of changes in the limb radiance for a given tangent height to changes in the ozone amounts as a function of altitude. The changes in the natural log of both quantities are used to give a % radiance change / % ozone change interpretation to the results. The orange curve with the highest peak for 305-nm is for the 45- km tangent height case.

33. Limb Aerosol Extinction Retrievals From a talk by Nick Gorkavyi Dust from the Russian Chelyabinsk Meteor

34. Time evolution of the Russian Meteor Cloud The Meteor Skybelt has a vertical depth of about 5 km, a width of about 300 to 400 km, a density of about 1 particle per cc. Total particulate mass within the skybelt is estimated to be 40-50 metric tons. From a talk by Nick Gorkavyi

35. Extending the Records: OMPS Ozone Products Lawrence Flynn with contributions from the NOAA and NASA OMPS S-NPP Teams Will products from OMPS (NOAA’s new generation of ozone monitoring instruments) measurements continue the past Ozone CDRs with adequate fidelity? This talk will present the following results: The on-board calibration systems are functioning as designed and are identifying changes in the optical throughput and dark current with high accuracy. Validation of the SDRs and EDRs showed good stability but identified areas where improvements were needed. Adjustments and models to apply corrections have been created to reduce errors from stray light and wavelength scale variations to acceptable levels and are progressing into the operational product processing system. Offline processing with the heritage Version 8 algorithms creates products with good consistency with existing elements of the Ozone CDRs. These algorithms are moving down the pipeline for operational implementation. The figure on the right compares daily averages (in Dobson units) for total ozone products over a latitude x longitude box in the equatorial Pacific Ocean for December 2014. The black line and symbols are the results from the Version 8 algorithm applied to the NOAA-19 SBUV/2 measurements and the red line and symbols are for the same algorithm applied to S-NPP OMPS measurements. They show good agreement. The other colors are for total ozone values from a variety of instruments and algorithms.

36. Provisional Maturity for S-NPP OMPS EDR products The Ozone Mapping & Profiler Suite (OMPS) was launched on the Suomi National Polar Partnership (S-NPP) satellite on October 28, 2011. The OMPS is the next generation of operational ozone sensors for NOAA. Its products will replace existing ones from the Solar Backscatter Ultraviolet (SBUV/2) instruments. NOAA/NESDIS has provided validation and analysis of the Total Column and Nadir Profile operational ozone products from the Interface Data Processing Segment (IDPS) to justify their advance to Provisional Maturity. This will be applied retroactively to the products as of March 1, 2013. Findings presentations & ReadMe files are provided with the products at CLASS. Significance Significance: The OMPS instruments provide the measurements to continue monitoring atmospheric ozone. Slide courtesy of L. Flynn; Maps by W. Yu. Supports the CLIMATE Mission and the WEATHER & WATERGoals Supports the CLIMATE Mission and the WEATHER & WATER Goals Total Ozone (DU) from S-NPP OMPS (top left), NASA EOS Aura OMI (top right) and MetOp-A GOME-2 (lower left) total ozone maps for October 4, 2012. Ozone Hole conditions (< 220 DU) are present at the southern tip of South America. The two figures on the right compare the OMPS Limb Profiler and Nadir Profiler ozone profiles for September 6, 2012. They show the zonal mean mixing ratios at pressures from 100 to 1 mbars versus Latitude. The Limb data has been smoothed vertically to be similar to the Nadir resolution. Refinements of both products are continuing. (Limb Profiler data and figures are from the NASA OMPS NPP Science Team.)

37. New Products from S-NPP OMPS The Ozone Mapping & Profiler Suite (OMPS) was launched on the Suomi National Polar Partnership (S- NPP) satellite on October 28, 2011. The OMPS is the next generation of operational ozone sensors for NOAA. Its products will replace existing ones from the Solar Backscatter Ultraviolet (SBUV/2) instruments. NOAA/NESDIS are now providing BUFR versions of the OMPS Nadir Ozone Profile and Total Column maps to users in NRT from the NDE system. This format is compatible with applications for the NWS and other operational users. A newer version of the combined UV/IR total ozone maps (called TACO) using information from the CrIS Ozone product in place of HIRS is now in operational production at OSDPD. As soon as the OMPS Ozone Profile product advances to validated maturity, it will use that information in place of the SBUV/2 data. Significance Significance: The OMPS instruments provide the measurements to continue monitoring atmospheric ozone. Slide courtesy of L. Flynn; Maps by J. Niu. Supports the CLIMATE Mission and the WEATHER & WATERGoals Supports the CLIMATE Mission and the WEATHER & WATER Goals The two figures on the right compare the current TOAST Product using HIRS and SBUV/2 (top) with the new (TACO) one using CrIS and SBUV/2. Validation studies confirm that the CrIS ozone product information is superior to the HIRS single channel retrievals.

38. Taco Umkehr Layer Analysis 38

39. Optical throughput change < 1% NM NP Reference Working

40. Pacific Box Comparisons Product statistics and cross-track variations are monitored in an uneventful region of the globe extending from 20°S to 20°N and 100°W to 180°W. The following figures each show four weekly means for May 2014 versus cross-track view position. The highly repetitive patterns are produced by cross-track channel biases. These will be removed by soft calibration adjustments using the Calibration Factor Earth tables.

41. Current Total Ozone Product Deficiencies 4% Small uncertainties in the calibration produce a cross-track dependence in the total ozone with satellite view angle The wavelength quadruplet chosen for the SO 2 index is too sensitive to calibration uncertainties. This leads to a large number of false positives. The figures show weekly averages for June 2014 for a latitude/longitude box in the equatorial Pacific. The horizontal axes are cross-track position. These patterns are repeatable and will be removed with soft calibration adjustments. SO 2 exclusion

42. 42 Weekly 1-percentile Effective Reflectivity and Aerosol Index values, for June 2014 for a latitude / longitude box in the Equatorial Pacific versus cross-track pixel. Internal Consistency and Vicarious Calibration / Validation Generation of soft calibration coefficients (CFE) – Use Equatorial Pacific Box, Minimum Reflectivity = 4%, no aerosol, no SO2, TOz set to OMI TOz. Pacific Box Internal Consistency

43. Time series of daily zonal mean ozone for Pacific Box 43 Daily equatorial Pacific mean time series of total column ozone for a variety of products from OMPS, OMI, GOME-2, and SBUV/2 3%

44. Ground-Based Comparisons Comparisons to Ground-base station Total Column Ozone estimates for matchup overpass data sets. Top right is summary for 22 Dobson stations. Lower plots are for a single well- calibrated NOAA/ESRL station at Boulder CO. Differences for OMPS and OMI products are shown. 44

45. 45 Red OMI V8 Blue OMPS INCTO Blue OMI V8 Red OMPS INCTO

46. Comparisons of INCTO to three very good Dobson ground stations 1/2012 to 6/2013. Notice the shift in biases in June 2012 with the introduction of new solar flux and wavelength scales. 46 INCTO and Dobson Mauna Loa 2012/2013 INCTO and Dobson Boulder 2012/2013 INCTO and Dobson Lauder, NZ 2012/2013

48. Statistics for Dobson Match-Up Data Sets in preceding slides SiteSat.Avg G Avg S mGmG mSmS mEmE σδε  Min E Max E # DaysNameDU BOUOMPS308.7293.90.900.980.940.026.76.30.97-10.6-24.1 N=335OMI308.7306.30.931.000.960.026.46.10.97 0.3 -8.1 MLOOMPS256.6 c 259.40.991.131.060.034.74.90.93 0.4 c 5.9 c N=217OMI256.6 c 266.91.031.171.100.034.44.80.94 6.0 c 15.6 c LAUOMPS304.5300.20.971.000.990.024.84.70.99 -3.3 -5.8 N=270OMI304.5304.40.971.010.990.025.35.20.98 0.6 -1.1 48 c The Dobson station is near the top of Mauna Loa. Satellite FOVs include ocean scenes. Adjustments from 6 to 12 DU have been used to account for these scene differences based on Hilo HI ozonesondes and standard ozone profiles. The OMPS Bias estimates at the maximum and minimum data values for each station show negative biases.

49. Summary The OMPS instruments are performing well and delivering ozone products to continue the over 30-years of satellite monitoring. Validated nadir total column ozone and ozone profiles will be available operationally by mid- 2014. The limb ozone profiles provide global coverage of the ozone layer with high vertical resolution. The OMPS measurements can be used to provide other atmospheric chemistry and composition products at good horizontal resolution. 49

50. Center Slit, OMPS Limb Ozone Profile Retrievals for one Orbit on October 22, 2013 Event # Along Orbit from South to North Altitude in kilometers Ozone Amounts in Volume Mixing Ratio, 10^12 mol/cm^3 High vertical resolution structure of the Antarctic Ozone Hole ozoneaq.gsfc.nasa.gov/omps/about/

51. 51 Current Status of NUCAPS CrIS Trace Gas Products gas Range (cm -1 ) PrecisionInterference UTH1400-150015%T(p) O3O3O3O31025-105010%H2O,emissivity CO*2080-220030%H2O,N2O CH 4 1250-1370 20 ppb H2O,HNO3 CO 2 680-7952375-2395 2 ppm H2O,O3 HNO 3 860-9201320-133040%25%emissivityH2O,CH4 N2ON2ON2ON2O1250-13152180-225010%10%H2OH2O,CO SO 2 1340-13801000%H2O,HNO3 CFCl 3 (F11) 830-86020%emissivity CF 2 Cl (F12) 900-94020%emissivity CCl 4 790-80550%emissivity * If sub-sampled CrIS SW is used. 15% if full resolution SW is acquired Product Available Now Work in Progress Held Fixed

52. 52 Summary of products from Science Code gasRange (cm -1 )Precisiond.o.f.Interfering GasesScience Code T650-800 2375-2395 1K/km6-10H2O,O3,N2O emissivity 100 levels H2OH2O1200-160015%4-6CH4, HNO3100 layers O3O3 1025-105010%1+H2O,emissivity100 layers CO2080-220015%  1 1 H2O,N2O100 layers CH 4 1250-13701.5%  1 1 H2O,HNO3,N2O100 layers CO 2 680-795 2375-2395 0.5%  1 1 H2O,O3 T(p) 100 layers Volcanic SO 2 1340-138050% ??< 1H2O,HNO3flag HNO 3 860-920 1320-1330 50% ??< 1emissivity H2O,CH4,N2O 100 layers N2ON2O1250-1315 2180-2250 5% ??< 1H2O H2O,CO 100 layers NH 3 860-87550%<1emissivityBT diff CFCs790-94020-50%<1emissivityConstant