Beta Findings for the OMPS

Nadir Mapper 1st Guess Total Column Ozone (INCTO)

andNadir Profiler Ozone Profile (IMOPO)

Products

JPSS and NPP OMPS Teams

Outline (1)• Beta Definition• OMPS Background• INCTO Performance

– Internal Evaluation• Flags – Sun Glint y, XXX Ozone, etc. Distribution• Reflectivity Map, XXX Distribution• Total Column Ozone Map, XXX Latitude Zonal means• Aerosol Index XXX Map, Distribution, cross-track• Cross-Track Dependence• Geolocation of Small FOV

– External Evaluation• Comparisons to OMPS V8TOZ• Comparisons to OMI and GOME-2 V8TOZ

– Known Deficiencies• EDR LUT (CCR #343)• Snow/Ice Data not updated (DR #)• Cloud Top Pressure (CCR #385)• Partial Cloud Logic (CCR #419)• Day One Solar and Trending (CCR #411)• Dark and Smear (DR #4750)• Stray Light / Radiance Coefficients• Wavelength Scale

Outline (2)• IMOPO Performance

– Internal Evaluation• Flags SAA, XXX tozcod, procod, • Initial and XXX Final Residuals• Reflectivity (also to INCTO)• Total Column Ozone (also to INCTO)• Ozone Profiles (Layer Amounts and Mixing Ratios)

– External Evaluation• Comparisons to OMPS V8PRO (Mixing ratios)• Comparisons to SBUV/2 V8PRO

– Known Deficiencies• Day One Solar and Trending (DR #)• Stray Light / Radiance Coefficients • Dark and Smear• Wavelength Scale• Update algorithm to V8PRO

• Product Status• Summary of Findings and Recommendations

– Promotion to Beta– Upgrades to V8TOZ and V8PRO– OMPS NM EDR: OOTCO

Outline (BU)• Irradiance Validation

– Solar diffuser measurements• Radiance Validation

– Signal-to-Noise– South Atlantic Anomaly – charged particle effects– Stray Light– Wavelength Scale

• Radiance/Irradiance Validation• SDR Status• Summary of Findings and Recommendation

5

Data Product Beta Maturity Definition

OMPS FundamentalsNOAA, through the Joint Polar Satellite System (JPSS) program, in partnership with National Aeronautical Space Administration (NASA), launched the Suomi National Polar-orbiting Partnership (Suomi NPP) satellite on October 28, 2011. The Ozone Mapping and Profiler Suite (OMPS) consists of two telescopes feeding three detectors measuring solar radiance scattered by the Earth's atmosphere and solar irradiance by using diffusers. The measurements are used to generate estimates of total column ozone and vertical ozone profiles. The total column sensor uses a single grating monochromator and a CCD array detector to make measurements every 0.42 nm from 300 nm to 380 nm with 1.0-nm resolution. It has a 110° cross-track FOV and 0.27° along-track slit width FOV. The measurements are currently combined into 35 cross-track bins: 3.35° (50 km) at nadir, and 2.84° at ±55°. The resolution is 50 km along-track at nadir, with a 7.6-second reporting period. The instrument is capable of making measurements with much better horizontal resolution.The nadir profile sensor uses a double monochromator and a CCD array detector to make measurements every 0.42 nm from 250 nm to 310 nm with 1.0-nm resolution. It has a 16.6° cross-track FOV, 0.26° along-track slit width. The current reporting period is 38 seconds giving it a 250 km x 250 km cell size collocated with the five central total column cells. The limb profile sensor is a prism spectrometer with spectral coverage from 290 nm to 1000 nm. It has three slits separated by 4.25° with a 19-second reporting period that equates to 125 km along-track motion. The slits have 112 km (1.95°) vertical FOVs equating to 0 to 60 km coverage at the limb, plus offsets for pointing uncertainty, orbital variation, and Earth oblateness. The CCD array detector provides measurements every 1.1 km with 2.1-km vertical resolution. The products for the Limb Profiler are not discussed here.

Instrument Fields of View. Schematic from Ball Aerospace and Technology Corporation.

Nadir Mapper & Profiler

LimbProfiler

MainElectronics

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. 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 have completed multiple passes through their internal dark and nonlinearity calibration sequences and are beginning to make regular solar measurements. (See information on the OMPS SDRs.)

EarthMode

SolarModeDiffuser

Entrance Aperture Entrance Aperture

Diagram from Ball Aerospace and Technology Corporation

Solar Irradiance

Earth Radiance

Radiance/Irradiance Ratio

Ozone Absorption Features

Solar Line

Typical spectra from 310 to 380 nm for OMPS Nadir Mapper.

Wavelength, nm310 380

OMPS Nadir Mapper Spectra• The plot at the top of the preceding page shows a sample

OMPS Nadir Mapper solar spectrum measured in January. The initial calibration, goniometry and wavelengths scales have been applied. Notice the Fraunhofer lines, e.g., a deep one near 360 nm.

• The plot in the middle shows a sample spectrum for the Earth View data for the nadir field-of view.

• The plot on the bottom shows the ratio of the first two spectra. Notice that much of the structure in the solar spectrum cancels out in the ration. Also notice the variations between 320 and 330 nm produced by differential ozone absorption with wavelength.

Sun Glint Flags

This figure gives the location of pixels with the OMPS NM Sun Glint flag set to 1 – viewing geometry has the potential for sun glint and the location is over open water – for the OMPS Nadir Mapper for March 5, 2012. The code is consistent in assigning this condition. (The VIIRS image in the upper left corner shows similar locations for Sun Glint but shifted to the South as the image is for Sun angles in January.)

VIIRS Blue MarbleJanuary 23, 2012

Distribution of Non-Fill Effective Reflectivity

Fewer than 1% of the 120000 values for this day are greater than 100 and only four values are less than 0.2.5 % of the effective reflectivity have fill values with most of these for SZA >= 88 degrees and all with N-values with fill – no error flags are set for fill N-value cases.

Cross-track View Position

|Lat|<20Lon<-100

Black Week 1Blue Week 2Green Week 3Red Week 4

OMPS NM Minimum Reflectivity Cross-Track Dependence (Need INCTO for March)

INCTO Cloud Top Pressure versus Latitude

Latitude, Degrees North

Clou

d Pr

essu

re, A

tmos

pher

es

The slide on the previous page shows cloud top pressures versus latitude for three retrievals for February 26, 2012.

The bottom one shows the cloud top pressures for INCTO with the at-launch climatology.

The middle one shows the cloud top pressures for INCOT with the new UV-based climatology.

The top one shows the cloud top pressures for the Version 8 processing at STAR also with the new data.

The new climatology gives new cloud pressures that are much lower (larger pressures) in the high latitudes of the Southern Hemisphere as the UV sees through thin cirrus. The switch to use this improved data set will be implemented in the near future.

INCTO Cloud Top Pressure versus Latitude

OMPS NM INCTO Minimum Reflectivity Cross-Track Dependence

The figure on the previous slide shows the weekly one-percentile effective reflectivity values for the month of February for all the data in a latitude/longitude box in the Equatorial Pacific versus cross-track view angle. (17 is the nadir position and 0 and 34 are the extreme viewing angles.) We expect this value to be approximately 4% from climatological measurements made by other instruments.

INCTO Total Ozone Map

Figures: J. Niu, NOAA/STAR (ERT) 1/26-27/2012

331-nm Channel Radiancesfor the first eight orbits of OMPS Nadir Mapper measurement (end of 1/26/2012 and start of 1/27/2012). This image shows the expected range of values and variations across the orbital track and with solar zenith angles at the times of the measurements. The white circle around the North Pole is the region of polar night during the Northern Hemisphere Winter. The OMPS needs scattered sunlight to make its measurements, so there are no data there.

Total Ozonefrom the multiple triplet retrieval algorithm in IDPS for the same eight orbits for the first pass ozone retrieval (IP product without CrIS or VIIRS information). The values show some cross track variations and are offset approximately 5% from another satellite ozone product. These uncertainty levels for preliminary products are consistent with the use of prelaunch calibration parameters and tables in the initial operational system.

Effective Reflectivity from the multiple triplet retrieval algorithm in IDPS for the same eight orbits. The quantity represents the UV reflectivity of the clouds and surface in each Field-of-View. Again, the range of values from bright clouds to dark open ocean scenes is as expected.

Cross-track View Position

Cross-Track Ozone Dependence

March weekly-mean total column ozone (OMPS Nadir Mapper First-Guess Multiple-Triplet algorithm retrievals) as a function of the cross-track view angle for the region defined by 20S to 20N Latitude and 100W to 180W Longitude. The persistent cross-track bias could be caused by deficiencies in the initial calibration or solar data values. Position 16 is the nadir FOV and 0 and 31 are the extremes.

|Lat|<20Lon<-100

2%

Tota

l Ozo

ne, D

U

Black Week 1Blue Week 2Green Week 3Red Week 4

OMPS Geolocation

C. Seftor, NASA/GSFC (SSAI) 1/26-27/2012

This image shows the effective reflectivity for the 380-nm Channel for part of an orbit of small Field-of-View (5 KM X 10 KM at Nadir) made by the OMPS Nadir Mapper in a special diagnostic mode. The Qatar peninsula sticking into the Persian Gulf in the middle of the picture lies along the nadir view of the orbital track and gives a preliminary assurance of the geolocation at better the 5 KM.

Daily Total Column Ozone map comparisons between (a) IDPS OMPS First Guess Multiple Triplet product, (b) NOAA OMPS V8 product, and (c ) NASA OMI V8.6 product for March 30, 2012. Cross-track features in OMPS products are related to the use of preliminary calibration values.

(a)

(b)

(c)

DU 100 300 500

February weekly-mean total column ozone (OMPS Nadir Mapper First-Guess Multiple-Triplet algorithm retrievals) as a function of the cross-track view angle for the region defined by 20S to 20N Latitude and 100W to 180W Longitude. The persistent cross-track bias could be caused by deficiencies in the initial calibration or solar data values. Position 16 is the nadir FOV and 0 and 31 are the extremes.

(b)

DU 100 300 500

(a)

(c)

Red OMPSBlue GOME-2Black N19 SBUV/2Green N18 SBUV/2

Tota

l ozo

ne, D

U

Comparison of daily mean total column ozone estimates from SBUV/2 NOAA-18 (BLACK) & NOAA-19 (GREEN), NPP OMPS NM (RED) and MetOP-A GOME-2 NOAA V8 Processing (BLUE) for February 2012 for a latitude/longitude box in the Equatorial Pacific.

SAA Effects on NP and NM

The noise/spikes below show the expected increases when an orbital path falls with within the SAA but return to normal levels after passing through it.Examination of South Atlantic Anomaly (SAA) effects and product flagging. The OMPS was designed to use the OMPS Limb Profiler retrievals in the SAA where the OMPS Nadir Profiler is affected by charged particles and cannot be used to produce ozone profiles.

Map of South Atlantic Anomaly effects on OMPS NM closed-door dark current measurements in December and November 2011, overlaid with OMPS NP SAA Flags (0 to 8) for 3/5-6/2012

Residuals

Daily zonal means for 20ºS to 20ºN (avoiding SAA effects by excluding longitudes from 0 to 90 West) for initial measurement residuals for the three shortest wavelengths (252, 274 and 283 nm) from the operational NOAA-18 (BLACK) & 19 (BLUE) SBUV/2 and the STAR NPP OMPS NP (RED) Version 8 ozone profile retrieval algorithm for February and March of 2012. The OMPS values are offset due to the use of a preliminary solar calibration and the lack of stray light corrections.

Slides for Ivan NPP Science TeamMaturity Matrix Time Line

ICVS Contenthttp://www.star.nesdis.noaa.gov/smcd/spb/icvs/proComparison.phphttp://www.star.nesdis.noaa.gov/smcd/jcsda/nsun/NPP/ipm_telemetry_npp_omps.php

Product upgrades near termLimited Validation

Products Longer TermNPP ST Synetgy

Plots for Ivan• Maps of INCTO, OOTCO, V8TOZ, OMI, GOME-2• Comparison of ozone profiles IMOPO, V8PRO,

N18, N19

OMPS Slides for Ivan’s NPP Science Team OverviewNote: Slide 3 has three layers set to fade in sequentially.

I will have an OMPS poster there as well.

OMPS Total Ozone Status• IDPS is processing OMPS Nadir Mapper data with two

versions of the Multiple Triplet Algorithm (an extension of the TOMS V7 algorithm) – a first guess product (INCTO) and a second pass product (OOCTO). The second pass product uses NRT information from the VIIRS and CrIS Instruments– A Beta maturity level findings meeting is planned for later

this month for the INCTO product• NOAA STAR is processing OMPS Nadir Mapper data

with the Version 8 Algorithm• A new Day One solar file will be implemented in IDPS

by Mx7 at the latest

OMPS Nadir Profiler Status• IDPS is processing OMPS Nadir Profiler and

Nadir Mapper Measurements with the Version 6 SBUV(/2) algorithm (IMOPO)– A Beta maturity level findings meeting is planned for

later this month for the IMOPO product• NOAA STAR is processing the data with the

Version 8 SBUV/2 algorithm• A new Day One solar file will be implemented in

IDPS by Mx7 at the latest• We are working on a Stray Light correction

Additional Information on OMPS

• Near term work on the total ozone products includes: • Evaluation of reflectivity and aerosol index products• Evaluation of cross track consistency • Modeling of along orbit wavelength scale variations

• Daily Maps of the OMPS INCTO and V8TOZ are available to Science Team Members at

http://www.star.nesdis.noaa.gov/smcd/spb/icvs/proComparison.php

• Near term work on the ozone profile products includes:• Comparison of profiles and initial residuals to SBUV/2• Modeling of stray light for the Nadir Profiler channels

• Information on IMOPO and V8PRO will be made available soon athttp://www.star.nesdis.noaa.gov/smcd/spb/icvs/proSBUV2operational.php

Background Slides

Sensor Algorithm Algorithm Specification Compliance Beta Provisional Val 1 Val 2 Val 3

OMPS RDR   Feb-12        

OMPS SDR - Ozone TC EVCCR approved to sever links between Cal and EV -

Changes to be implemented in Mx6 Code bug on DoY identified - To be corrected in

Mx5.3Mar-12 Mar-13 Jun-13    

OMPS SDR - Ozone NP EV No issues Mar-12 Mar-13 Jun-13    

OMPS SDR Ozone TC CalLimited distribution via OMPS SOC/PEATE (this is

the Cal SDR used by the OMPS PEATE) -* This may be the version that is used for "beta"

Mar-13 Jun-13 Sep-13

OMPS SDR Ozone NP CalLimited distribution via OMPS SOC/PEATE (this is

the Cal SDR used by the OMPS PEATE) -* This may be the version that is used for "beta"

Mar-13 Jun-13 Sep-13

OMPS SDR - Ozone TC Geo No issues Mar-12 May-12 Aug-12    

OMPS SDR - Ozone NP Geo No issues Mar-12 May-12 Aug-12    

OMPS EDR - Ozone TCPost-launch verification required; anticipate the

need for post-launch adaptation of the TOMS-EP Version 8 algorithm to OMPS TC to satisfy JPSS

L1RD accuracy and precision requirements

May-12 Jul-12 Jul-13 Sep-13 Feb-14

OMPS EDR(IP) - Ozone NP

No heritage performance requirements; post-launch verification of operability required and

anticipate the need for post-launch adaptation of the SBUV/2 Version 8 algorithm to OMPS NP to

satisfy JPSS L1RD accuracy and precision requirements

May-12 Jul-12 Jul-13 Sep-13 Feb-14

OMPS Ozone Products Maturity Matrix

Outline• Day 1 Solar and Radiometric Coefficients• Wavelength Scale• Geolocation Registration• NM 1st Guess IP (Repeat for MT EDR and V8 EDR)

– Reflectivity• Range and Error Flags• Pacific Box – Cross-track• Comparisons

– Ozone• Range and Error Flags• Pacific Box – Cross-track• Comparisons

– Aerosol Index• Range• Pacific Box – Cross-track• Comparisons

• NP Ozone Profile V6– Total ozone– Ozone Profile– V8 Initial Residuals– Stray Light– South Atlantic Anomaly – charged particle effects

Outline (2)

• NP Ozone Profile V6– Total ozone– Ozone Profile– V8 Initial Residuals– Stray Light– South Atlantic Anomaly – charged particle effects

These maps show false color plots of the total column ozone estimated from OMPS Nadir Mapper measurements for March 5, 2012. The top map shows the First Guess product from IDPS processing. The lower map shows a processing with the Version 8 algorithm at STAR. Both maps show cross-track bias patterns but generally detect the appropriate global ozone variations. (Figures created by W. Yu, ERT.)

20120315 Updateto

OMPSInitial Performance Evaluation

List of Golden Day Datasets

• OMPS NM: SOMTC, GOMTC, OOMTC, INCTO• OMPS NP: SONPS, IMOPO, GONPO• Calibration SDRs?• GOME-2 data, Level 1, 2, 3• SBUV/2 data, Level 1, 2 • Assimilations?• Ground-based• CrIS O3 IP?• OMI Level 3 Ozone, aerosol and reflectivity?

20120308 Updateto

OMPSInitial Performance Evaluation

Summary New Bullets• Collection of Golden Day (20120225) data sets is

proceeding.• STAR data repository (scdr-files system) has begun

receiving IDPS products from CLASS; ozone profile products are present in the March files received by scdr.

• ICVS monitoring plots for OMPS are starting to be populated.

• Surface reflectivity values for the Nadir Mapper are sometimes greater than 1. This has been traced to the climatology.

• Comparisons of the observed solar to synthetic spectra show consistent biases and features.

Comparisons between 201201027 Solar & synthetic data. Notice the features are shared between the three. The figure on the right shows that all three have the same pattern of average bias as a function of cross-track position.

Top Figure -1*0.01 nm shift for SAO synthetic. Wavelength scale shifts are responsible for most of the smaller scale structure in the differences between the synthetic and observed. New figures will be provided after further analysis.

20120301 Updateto

OMPSInitial Performance Evaluation

Summary new Bullets

• Work is proceeding to produce a provisional Day 1 Solar spectra

• Offline V8 processing (TOZ and PRO) is active• NP stray light investigation will compare estimates to

those from modulation transfer functions• G-ADA processing of February data has been delayed

due to other priorities. The Day-of-Year error plaguing February processing does not appear to be present for March for total ozone.

• There is a dark current discretization error that will be present until the next build caused by the non-linearity correction

• We need to develop a screen to filter transients for the smear correction

Histogram of Dark Counts for TC Cal SDR

The OMPS TC Dark counts show severe discretization errors. In this histogram, there are approximately 13 empty bins between each populated one. This problem has been traced to a nonlinearity correction that uses the corrected value for the nearest table node instead of an interpolated value.

Comparison of solar measurements and synthetic spectra in the NP/NM overlap region. Further study of the radiance/ irradiance ratios will need to be conducted once a preliminary Day 1 solar is provided.

OMPS Earth View SDR Status (1)

• Dark Current levels and the increase in Hot Pixels are consistent with prelaunch characterization and expectations.

• On-orbit Non-Linearity measurements match laboratory data very well.• Solar Spectra is consistent with synthetic spectra from instrument bandpass

convolved reference spectra.• Preliminary analysis of Signal-to-Noise levels show performance above

requirements.• Stray Light is present in the Nadir Profiler spectra and will need a correction.

Measurements were taken in the laboratory to provide these corrections.• The Wavelength Scale shows small intra-orbit variations tracking the optical

bench temperatures• Charge particle effects in the South Atlantic Anomaly are apparent in the

Nadir Profiler spectra in the expected region.• Preliminary evaluation of the Geolocation Registration shows it well within

specifications• The CCD array Detector Temperatures are within allowable tolerance of the

design values.

• A good amount of Earth-View SDR data has been obtained over the last month for both Nadir instruments.

• A minor code bug and a minor table bug have led to missing fields for scales and solar flux in the EV SDRs. The radiances are complete. This has prevented the creation of IPs and EDRs at IDPS since February 1. The issues will be resolved on March 1.

• In the meantime, the GRAVITE-ADA has begun reprocessing February data to provide the missing information (both SDRs and IPs).

• OMPS EV SDRs have been used successfully to generate total ozone and ozone profiles in offline processing (at the PEATE and STAR) with various flavors of the Version 8 Algorithms.

• The process to provide new and updated SDR Calibration tables to IDPS is now working well. Fast Track Tables have been identified and can be replaced with less bureaucratic overhead.

• Several DRs have been identified (both SDR and EDR related) and solutions are proceeding through the system.

• The teams are awaiting the introduction of measured day one solar into the products.

• Delivery and archiving of JPSS-Funded, PEATE-processed Calibration SDRs is an open concern.

OMPS Earth View SDR Status (2)

20120223 Updateto

OMPSInitial Performance Evaluation

andFirst Images

Summary new Bullets• NP has stray light; a two term correction using

radiances at 305 nm and at 298 nm is envisioned. The mean magnitude of these corrections will affect any radiance/irradiance validation.

• Pre-launch solar reference results for NM: NASA PEATE vs IDPS show ±4% wavelength dependent offsets

• Increase in hot pixels as expected

NP Stray Light Study via Correlation MatrixA smooth polynomial fit as a function of latitude was computed for each wavelength using three orbits of NP spectra. The correlation matrix was computed for the differences between the measurements and these fits. The results for six wavelengths are presented to the left. The peaks at 1.0 identify the selected wavelengths. The high correlations between wavelengths below 290 nm with those above 300 nm is direct evidence of stray light.E.g., the thin dotted line for 275 nm drops off as expected until 292 nm where it rises back up due to stray light contamination. Over half of the variations at this wavelength could be explained by variations at a channel between 300 and 310 nm. Similar results hold for all channels below 287 nm.

The single wavelength proxy model gives stray light ranging from 3% at high SZAs to 7% at low SZAs for the four shortest profiling wavelengths (253-288 nm) with additional variations when scene reflectivity variations are present in 300 nm.This signal contamination will complicate any attempts to find a Radiance/Irradiance adjustment. That is, we need to come up with a first-order stray light correction, before we try to adjust the NP calibration.Further investigation of the NP stray light in Earth View with a single wavelength proxy suggests that a weighted proxy may be needed. The relative weights can be provided by the BATC data.

NP Stray Light Study with 300 nm Proxy

Comparisons of pre-launch solar flux.The plot on the top shows the solar flux from the IDPS IP versus the solar flux from the PEATE product for the nadir scene (IDPS position 18; for the PEATE a simple average of scenes 18 and 19). There is a definite, wavelength-dependent difference between the IDPS and PEATE solar flux values (that is surprisingly large).The PEATE product is made from a reference solar flux that has been convolved with the appropriate OMPS BPS and CBC data.

C. Seftor, NASA/GSFC (SSAI) 1/26-27/2012

IDPS and Star Solstice proxy comparison are shown on the left. Differences are less than 3% and show high frequency features. The spectral scale have less than 0.01 nm differences in all 35 cross track position. The PEATE proxy comparisons to the STAR Solstice one are shown on the right. The differences range from ±5%.

Figure: J. Niu, NOAA/STAR (ERT)

Support Material forOMPS SDR to Beta Maturity DecisionPerformance compared to Pre-launch

Compilation of results fromJPSS and NPP OMPS Teams

67

Data Product Beta Maturity Definition

Outline• Dark Current• Hot Pixels• Non-linearity • Solar Spectra• Signal-to-Noise• Stray Light• Wavelength Scale• South Atlantic Anomaly – charged particle effects• Geolocation Registration• Detector Temperature

69

• Raw dark image for prelaunch• Exposure time of 1247.1 second• The number of coadds is 100 • Only one hot pixel is identified

Dark Transition from Ground to Orbit

Histogram function

Prelaunch

On-orbit

On-orbitPrelaunch Gaussian distribution

• Raw dark image for orbit• Exposure time of 1247.1 second• The number of coadds is 100 • A few potential hot pixels are observed.

Bin index

Hot pixels increasing as expectedHot pixel defined as 8 above pre-launch distribution

Pre-launch Orbit 1080

71

Linearity PerformanceTransition from Ground to Orbit

Observed OMPS TC LED drifts over time meets the system requirement of < 1% per minute

Sensor nonlinearity is compliant with the specification of < 2%

LED signal drifts System nonlinearity

On-orbitPrelaunch

On-orbitPrelaunch

Spectral Calibration Transition from Ground to Orbit

On-orbitPrelaunch

EOF/SNR analysis for six orbits (~1800 scans) OMPS NM on 1/28/2012 using three wavelength ranges (305-325, 320-345, 340-380) for 35 CT

0.1 is equivalent to SNR of 1000.

Maximum Residuals.Spikes could be filtered.

Average Radiances

Stray Light and SNR for OMPS NPThe scatter plot on the left compares simple three-wavelength Mg II core-to-wing ratio variations (280 nm to the average of 277 nm and 282 nm) with longer wavelength variations at 305 nm. The core average is ~0.4 of wings, so this -0.5% to 1.5% variation represents (1.0/0.4 – 1 =) 1.5 times the wing variations.

The RMSR estimates on the right were done on clean data for 230 spectra. A spectral screen detected and removed one deviant value on average from each of the 120-wavelength spectra. The removal of the second EOF pattern was somewhat arbitrary as it and the third EOF may be stray light patterns. The EOF used a 6th order polynomial for normalization.

Wavelength Shift and Ring Effect/Stray Light

EOF analysis 365 to 380 nm for parts of six orbits on 1/28/2012. The first two patterns contain 90% of the variability after removing a 3rd order polynomial from Rad/AvgRad. They are combinations of a Wavelength Scale Shift and Ring Effect/Stray Light. Given the radiance levels, a 0.01 in the figures at the top equates to approximately 1% variation.

* EOF pattern–– 0.02-nm shift

* EOF pattern–– ∝ (1/Avg-poly)

SAA Effects on NP

Figure: J. Niu, NOAA/STAR (ERT)

The noise/spikes show the expected increases when an orbital path falls with within the SAA but return to normal levels after passing through it.

CCD Temperatures for the last ten daysDesign temperature is -30∘C.Deviations yet to be investigated.Small intra-orbit cycle observed.

Design temperature is -45∘C.Deviations yet to be investigated.Small intra-orbit cycle observed.

20120208 Updateto

OMPSInitial Performance Evaluation

andFirst Images

Summary Bullets (Compiled by L. Flynn 2/8/2012)

• OMPS NM SDRs– SNRs as expected (EOF and NN analysis)– Provide reasonable ozone and reflectivity (IDPS IP)– Geolocation is good at 5-KM level (Small FOV)– Intra-orbit Wavelength scale stability of 0.02-nm (EOF) – First Solar at 1% relative to laboratory– Solar is not in SDRs starting in February (Fill in ozone products)– Good Aerosol Index after NPP Science Team tweaking

• OMPS NP SDRs– SNRs as expected (EOF and NN analysis)– No wavelength scale in product; all zeros – Fill in Profiles– Will need corrections/filtering in SAA as expected– Stray light is present (correlation of 280-nm variations with 305-nm

variations gives 3%::50% response ratio)– First Solar at 5% relative to laboratory (possible 1-pixel offset in

wavelength scale from expected)

C. Seftor, NASA/GSFC (SSAI) 1/26-27/2012

We've been working on a (very) preliminary set of soft calibration factors using the extremely limited amount of data we have so far. Using this calibration, we've generated the first daily aerosol index map from OMPS, which I'm attaching here. Sun glint has been screened out (for the most part), and clouds are represented by the OMPS reflectivity measurements. The MODIS Blue Marble image is being used for the background. Note the missing orbit over North / South America.

Stray Light and SNR for OMPS NP

The scatter plot on the left compares simple three-wavelength Mg II core-to-wing ratio variations (280 nm to 277 and 282 nm) with longer wavelength variations at 305 nm. The core average is ~0.4 of wings, so this -0.5% to 1.5% variation represents (1.0/0.4 – 1 =) 1.5 times the wing variations.

The RMSR estimates on the right were done on clean data for 230 spectra. A spectral screen detected and removed one value on average from each of the 120-wavelength spectra. The removal of the second EOF pattern was somewhat arbitrary as it and the third EOF may be stray light patterns. The EOF used a 6th order polynomial for normalization.

EOF/SNR analysis for six orbits (~1800 scans) OMPS NM on 1/28/2012 using three wavelength ranges (305-325, 320-345, 340-380) for 35 CT

Wavelength Shift and Ring Effect/Stray Light

EOF analysis 365 to 380 nm for parts of six orbits on 1/28/2012. The first two patterns contain 90% of the variability after removing a 3rd order polynomial from Rad/AvgRad. They are combinations of a Wavelength Scale Shift and Ring Effect/Stray Light. Given the radiance levels, a 0.01 in the figures at the top equates to approximately 1% variation.

* EOF pattern–– 0.02-nm shift

* EOF pattern–– ∝ (1/Avg-poly)

20120202 OMPS

Initial Performance Evaluation and

First Images

Summary Bullets (Compiled by L. Flynn 2/2/2012)

• OMPS NM SDRs– SNRs as expected (EOF and NN analysis)– Provide reasonable ozone and reflectivity (IDPS IP)– Geolocation is good at 5-KM level (Small FOV)– Intra-orbit Wavelength scale stability of 0.02-nm (EOF) – First Solar at 1% relative to laboratory

• OMPS NP SDRs– SNRs as expected (EOF and NN analysis)– No wavelength scale in product; all zeros– Will need corrections/filtering in SAA as expected– Stray light is present (correlation of 280-nm variations with

305-nm variations gives 3%::50% response ratio)– First Solar at 5% relative to laboratory (possible 1-pixel

offset in wavelength scale from expected)

OMPS Geolocation

C. Seftor, NASA/GSFC (SSAI) 1/26-27/2012

This image shows the effective reflectivity for the 380-nm Channel for part of an orbit of small Field-of-View (5 KM X 10 KM at Nadir) made by the OMPS Nadir Mapper in a special diagnostic mode. The Qatar peninsula sticking into the Persian Gulf in the middle of the picture lies along the nadir view of the orbital track and gives a preliminary assurance of the geolocation at better the 5 KM.

Figures: D. Rault, NASA LaRC

OMPS Limb ProfilerThese curtains represent the ozone profile in vertical slices through the atmosphere along the three paths shown above.They demonstrate the ability of the research retrieval algorithm for the OMPS Limb Profiler in use at the NASA Ozone PEATE. The gaps at the top in the middle of the plots occur when the satellite encounters charged particles as it passes through the South Atlantic Anomaly; these are consistent with the modeled effects.The profiles regularly extend down below 15 KM in altitude.

OMPS Nadir Mapper and Nadir Profiler First Solar

C. Seftor NASA/GSFC SSAI

OMPS TC IP Retrieval – January 27, 2012

89

OMPS TC IP ozone values are plotted between 100 and 450 DU. Data as available on January 27. There are values outside of this range that need to be investigated. (M Novicki and B Sen, NGAS)

OMPS TC Surface Reflectivity – January 27, 2012

90

Surface reflectivity values between -5% and 120% are allowed by the OMPS TC EDR operational algorithm. Investigating ≈120% surface reflectivity values at high SZAs over Antarctic, southern Greenland and Hudson Bay. (M Novicki and B Sen, NGAS)

V8.6 Ozone Retrieval from NASA Ozone PEATE

C. Seftor NASA/GSFC SSAI

Backup

EOF/SNR analysis for six orbits OMPS NM on 1/28/2012

EOF/SNR analysis for six orbits OMPS NM on 1/28/2012

EOF/SNR analysis for six orbits OMPS NM on 1/28/2012

Sensor Algorithm Algorithm Specification Compliance Beta Provisional Val 1 Val 2 Val 3

OMPS RDR   Feb-12        

OMPS SDR - Ozone TC EVCCR approved to sever links between Cal and EV -

Changes to be implemented in Mx6 Code bug on DoY identified - To be corrected in

Mx5.3Mar-12 Mar-13 Jun-13    

OMPS SDR - Ozone NP EV No issues Mar-12 Mar-13 Jun-13    

OMPS SDR Ozone TC CalLimited distribution via OMPS SOC/PEATE (this is

the Cal SDR used by the OMPS PEATE) -* This may be the version that is used for "beta"

Mar-13 Jun-13 Sep-13

OMPS SDR Ozone NP CalLimited distribution via OMPS SOC/PEATE (this is

the Cal SDR used by the OMPS PEATE) -* This may be the version that is used for "beta"

Mar-13 Jun-13 Sep-13

OMPS SDR - Ozone TC Geo No issues Mar-12 May-12 Aug-12    

OMPS SDR - Ozone NP Geo No issues Mar-12 May-12 Aug-12    

OMPS EDR - Ozone TCPost-launch verification required; anticipate the

need for post-launch adaptation of the TOMS-EP Version 8 algorithm to OMPS TC to satisfy JPSS

L1RD accuracy and precision requirements

May-12 Jul-12 Jul-13 Sep-13 Feb-14

OMPS EDR(IP) - Ozone NP

No heritage performance requirements; post-launch verification of operability required and

anticipate the need for post-launch adaptation of the SBUV/2 Version 8 algorithm to OMPS NP to

satisfy JPSS L1RD accuracy and precision requirements

May-12 Jul-12 Jul-13 Sep-13 Feb-14

Northrop GrummanOMPS TC IP versus OMPS TC EDR Analysis

May 16, 2012

M. Novicki, B. Sen

OMPS Total Column – IP vs. EDR

• The OMPS Total Column EDR and IP Retrievals from April 14, 2012 were compared• Retrievals greater than 650 DU:

– IP had 3 retrievals greater than 650 DU– EDR had 25 retrievals greater than 650 DU– No macro-pixels had out of range retrievals from either products

• Cloud fraction values– Cloud fraction values are expected to have significant differences– IP cloud fractions are an internally calculated effective cloud fraction– EDR cloud fractions are derived from the VIIRS Cloud Top Pressure product– Areas in the EDR with zero cloud fraction produced large differences in the total column retrieval

• Internally calculated best surface reflectivity had significant differences.– Differences over non-polar regions appear to be correlated with cloud fraction differences, as

expected.

99

OMPS Total Column – Total Ozone from IP and EDR• The OMPS total column retrievals in the following two slides are shown for April

14, 2012. • Slide 4 shows the retrievals from the OMPS TC IP (INCTO) product• Slide 5 shows the retrievals from the OMPS TC EDR (OOCTO) product. • The trends in the polar regions are very similar. • The EDR product shows many low total ozone retrievals in the non-polar regions.

– Typical total ozone values in these regions are around 200-300 DU, but they are frequently lower than that in the EDR product.

– The patterns of the low total ozone values suggest cloud fraction may be a significant factor. • The difference between the two retrievals is shown on slide 6

– 1 plot of all differences between -300 and 300 DU– 1 plot of all differences between -100 to 100 DU

100

OMPS Total Column – Total Ozone from IP Product

101

Total column retrieval from the OMPS TC IP (INCTO) product for April 14, 2012.

OMPS Total Column – Total Ozone from IP Product

102

Total column retrieval from the OMPS TC IP (INCTO) product for April 14, 2012.

OMPS Total Column – Total Ozone from EDR Product

103

Total column retrieval from the OMPS TC EDR (OOTCO) product for April 14, 2012.

OMPS Total Column – Total Ozone Difference (IP – EDR)

104

Shown here is the difference between the OMPS TC IP retrieval and the OMPS TC EDR retrieval. The patterns in the differences suggest that the cloud fraction is a significant contributor. There are whole swaths with similar differences (such as to the west of the tip of South America) and areas with significant differences (south of Greenland) which need further investigation. These areas have been outlined in red above and throughout this presentation. (M. Novicki and B. Sen, NGAS)

OMPS Total Column – Cloud Fraction from IP and EDR

105

Shown here are the cloud fractions used by the OMPS TC algorithm for the IP (left) and the EDR (right). The IP internally calculates an effective cloud fraction, while the EDR gets the cloud fraction information using the VIIRS Cloud Top Pressure product. The IP calculates a significantly higher cloud fraction over the polar regions, likely due to the presence of snow/ice. The VIIRS product is only available to 80 degree solar zenith angle. Over the non-polar regions, the EDR has significantly more cloud cover and has some areas of obvious deficiency (e.g. to the west of the southern tip of South America). (M. Novicki and B. Sen, NGAS)

OMPS Total Column – Impact of Cloud Fraction Differences on Total Column

106

On the left is shown the difference in the total column retrieval from the IP and the EDR, with the range truncated to ±100 DU to highlight the differences. On the right is shown the difference between the cloud fraction values used by each product. The significant differences in cloud fraction over the polar region does not have a large impact for the most part on the total ozone retrieval. Over the remainder of the globe, the differences in total column appear to correlate with the differences in cloud fraction. Key examples are highlighted by the red boxes. (M. Novicki and B. Sen, NGAS)

OMPS Total Column – Column differences compared with cloud fraction differences

107

The four panels shown here illustrate the differences in cloud fraction versus the differences in total column amount as a function of latitude. There is a clear negative trend around the equator and into the low-mid latitudes (top two panels) . This trend falls off at higher latitudes and the spread increases significantly. (M. Novicki and B. Sen, NGAS)

OMPS Total Column – Column differences where VIIRS cloud fraction is zero.

108

Shown above are histograms by latitude of the difference in total column between the IP and EDR retrievals for locations where the EDR has cloud fraction equal to zero. The effective cloud fraction and surface reflectivity calculations from the IP product indicate that there are likely clouds present in some of the areas that all have zero cloud fraction in the EDR. These regions are at mid- to high-latitude, which is reflected in the histograms. (M. Novicki and B. Sen, NGAS) EDR CLFRAC = 0.

OMPS Total Column – Best Reflectivity Value from IP and EDR

109

The left hand panel shows the best reflectivity values for the OMPS TC IP product, and the right hand panel shows the best reflectivity values for the OMPS TC EDR product. The OMPS TC EDR product indicates an overall ‘brighter’ scene than the IP product over non-polar regions. The polar regions are more consistent between the two products, but there are some noticeable differences, for example Greenland is less reflective in the EDR product. Reflectivity is a key component of the total column retrieval algorithm, and therefore likely contributes to the differences between the total column retrievals. The reflectivity values seem to have a cloud-like structure. (M. Novicki and B. Sen, NGAS)

OMPS Total Column – Cloud Fraction Differences and Reflectivity Differences

110

The difference between the best reflectivity values calculated by the OMPS TC IP and the OMPS TC EDR products are shown in the left-hand panel. The differences for the cloud fraction values for the same products are shown in the right-hand panel. There are clear correlations in the differences outside the polar regions. (M. Novicki and B. Sen, NGAS)

OMPS Total Column – Cloud Fraction Differences vs. Reflectivity Differences

111

The differences in the best reflectivity retrieved between the OMPS TC IP and the OMPS TC EDR are clearly correlated with the differences in the cloud fraction used in the same products. This is expected over the snow-free regions. (M. Novicki and B. Sen, NGAS)

OMPS Total Column – Total Column Differences vs. Reflectivity Differences

112

The differences in the best reflectivity retrieved between the OMPS TC IP and the OMPS TC EDR are negatively correlated with the difference in the total column amount. There are some obvious deviations from this correlation that should be investigated further. (M. Novicki and B. Sen, NGAS)

OMPS Total Column: Cloud fraction vs. Latitude

113

Shown here are the cloud fractions for the OMPS TC IP (top) and OMPS TC EDR (bottom) for cases where the difference in total ozone retrieval was greater than 40 DU. The EDR cloud fraction does not go below zero, as expected, while the IP effective cloud fraction goes below zero in some significant locations. These latitudes correspond to the areas highlighted throughout the presentation. (M. Novicki and B. Sen, NGAS)

OMPS Total Column Path Forward• As no other algorithms on NPP are using the OMPS TC product, it is

reasonable to only produce either the IP or EDR• To produce just an EDR product, we recommend the following actions:

– No longer bring in the VIIRS CTP data• Use CTP climatology recently submitted to DPE instead of VIIRS CTP (DR 4269)• Implement effective cloud fraction calculation currently under review (DR 4266)

– Use the VIIRS rolling tile Snow/Ice GIP• The rolling tile is in the process of being updated on a monthly basis

– This would function like a climatology • Eliminate attempted use of Snow and Ice Concentration EDR/IP• No climatology is currently built into the operational system, so there would be no

gracefully degraded product.• Current implementation is to use the rolling tile GIP for both the IP and EDR

• Continue analysis:– EDR product as it evolves – Comparisons between the IP and EDR products

114

Northrop GrummanOMPS Status

April 18, 2012

Bhaswar Sen, Megan Novicki, Wen-Hao Li

NGAS Activities and Highlights• Investigation of Sb as a function of bandpass change

– ADR 4418• Analysis of outlier retrievals in the TcIp data product• G-ADA processing to test new code for replacement of VIIRS cloud fraction

– ADR 4266– Baseline tests have been run for 3 TC granules– New code compiled and 3 TC granules produced with new code– Examining results of new code before proceeding to generation of test results for a full day

• Testing of new day one solar table in progress– ADR 4616– Baseline full day SDR processing performed with Ops TC–SOL-LUT– Ingested PEATE TC-OSOL-LUT in G-ADA for SDR processing

• Download of Mx6 binary tables is in progress– Plan to verify that tables were properly updated

• Verify content and format

117

Variation of Sb with bandpass

118

Investigation of Sb variance with bandpass• Used ozone profile for M325, pressure = 1• The ideal wavelengths were used• Sb was convolved with the bandpass for each of the sensor zenith angles for all wavelength

channels• Change in Sb over the bandpass space was small • Addresses ADR 4418

119

Sb difference from the mean value

120

The mean value of Sb was calculated for each wavelength channel, and then subtracted from the value for each of the sensor zenith angles. Differences are small (~10-4). Sb ranges from about 0.3 to 0.4.

(M.Novicki, NGAS)

Sb percentage difference from the mean value

121

The percentage difference of the Sb value for each channel versus the mean value for that channel. Differences are within ±0.5%. The greatest differences are seen at the edges of the CCD (sensor zenith angle indices 6 and 12).

(M.Novicki, NGAS)

Sb differences from the maximum value

122

Shown here is 0.5*(Sb/max(Sb)-1), and demonstrates the band center effects across the CCD. The values are small (~10-4).

(M.Novicki, NGAS)

Analysis of TcIp Outliers

123

Examination of TcIp outliers• Collected all retrievals with total ozone greater than 650 DU for March 1 through March 25• Total of 117 outliers• Algorithm error flag values range from 0 to 14

– Flag triggered by• Reflectivity out of range• SO2 out of range• Greater than 3σ differences between triplets• Total Ozone < 0 DU or Total Ozone > 750 DU

–Flag = 0 is the most common value (45 TcIp retrievals)–Five outlier retrievals are not identified by the

algorithm error code• V7 ozone is within range for all cases• Snow/ice fraction is equal to 1 in all cases• Addresses ADR 4266

124

125

Algorithm Error Flag Values for Outlier Retrievals

The algorithm error flag values for outlier retrievals (total ozone greater than 650 DU) are shown. Out of 117 retrievals, more than 1/3 have an error flag value of 0, indicating no problem in the execution of the algorithm (e.g. reflectivity out of range, high so2index, etc.) The range error that triggers this flag in the science code looks for values greater than 750DU.

(M.Novicki, NGAS)

126

Outlier Retrievals with Algorithm Error Flag = 0

Shown here are the total ozone values greater than 650 DU, but with an error flag equal to zero. Most of these cases are in the lowest ozone bin, possibly indicating that the retrievals are only slightly greater than 650 DU. The error flag indicates a problem within the retrieval algorithm, but not a value out of range.

(M.Novicki, NGAS)

127

V7 Total Ozone for All Outlier Retrievals

The v7 ozone retrievals for each of the outlier total ozone retrievals are all within the expected range. Most of the values are large, but there are a handful of cases with very low v7 ozone. The outliers with low v7 ozone retrievals should be examined for unusual Nvalues and location on the globe.

(M.Novicki, NGAS)

Testing of PEATE Day One Solar LUT

128

129

Baseline March 5 Earth SDR from G-ADA: Ops OMPS-TC-OSOL-LUT used in calculation

130

Difference between Earth SDR from G-ADA and Ops: Correct suite of SDR tables installed in G-ADA