The ALTIUS atmospheric limb sounder Journal of Quantitative Spectroscopy & Radiative Transfer

(1)

ContentslistsavailableatScienceDirect

Journal of Quantitative Spectroscopy & Radiative Transfer

journalhomepage:www.elsevier.com/locate/jqsrt

The ALTIUS atmospheric limb sounder

Didier Fussen

^∗

, Noel Baker , Jonas Debosscher , Emmanuel Dekemper , Philippe Demoulin , Quentin Errera, Ghislain Franssens, Nina Mateshvili, Nuno Pereira, Didier Pieroux, Filip Vanhellemont

Royal Belgian Institute for Space Aeronomy, 3, Avenue Circulaire, Brussels B1180, Belgium

a rt i c l e i n f o

Article history:

Received 10 February 2019 Revised 17 June 2019 Accepted 17 June 2019 Available online 18 June 2019 Keywords:

Atmospheric remote sounding Stratospheric chemistry Ozone proﬁles Limb spectroscopy

a b s t r a c t

Thisarticledescribestheobjectives,conceptandexpectedperformanceoftheALTIUS(AtmosphericLimb TrackerfortheInvestigationoftheUpcomingStratosphere)missioninviewofthecontinuationofearth limbmeasurements foratmosphericscience.The instrument,consisting ofthreeindependent spectral imagersinthe UV–vis–NIRrange, willbeoperated fromseveralobservationgeometries withaglobal coverageina3-daysrevisit cycle.ALTIUS willretrieve verticalconcentrationprofiles ofozonewitha required3–10% totalerrorinthe stratosphere. Otherimportant scientific objectiveswilladdress NO₂, H2O,CH4,variousaerosoland cloudextinction profilesaswellas otherminor absorbers.Tomography, temperatureretrievalsandself-calibratingobservationswillbealsoimplemented.

ThistypeofmeasurementbecamerarewiththefailureoftheEuropeanENVISATmissionin2012and thenumberwillfurtherdecreasewhenseveralCanadian,SwedishandUSlimbmissionswillterminate withinthenextfewyears.Itisasmallmissioninitiativebasedonamicro-satelliteplatformofthePROBA (ProjectforOn-BoardAutonomy)class,withahighagilityallowingforatmosphericlimbobservationsin differentremotesensinggeometriesfromalowearthorbit.Theinstrumentconsistsofthreeindependent spectralimagerscoveringtheUV–vis–NIRranges.

TheALTIUSmissionwillbeimplementedasanelementoftheESAEarthWatchprogramme.

Thepaperidentifiesthegeneralscientificcontextoftheprojectandderivesthemission,instrument and scientific productsrequirements. The generaldesign ofthe payloadand platformsystems isdis- cussed.Thepreliminarydataprocessingchainispresented,fromtelemetrydatatoretrievedgeophysical profiles,withacomplementarydataassimilationlevel.Apreliminaryassessmentofthemissionperfor- manceisdiscussedwithfocusonozoneprofileretrievals,whicharethemainobjectiveofthemission.

ThisisanopenaccessarticleundertheCCBY-NC-NDlicense.

(http://creativecommons.org/licenses/by-nc-nd/4.0/)

1. Introduction

1.1. EvolutionoftheEarth’supperatmosphereduringthe21st century

It isnow acceptedthat theglobalandpolardepletionsofthe ozone layer can be attributed to the presence of halogen com- poundsreleasedby anthropogenicemissions.Inthe upperstratosphere, these trace gases are photolysed by solar UV radiation which releases active halogens destroying ozone by catalytic re- actions [22].The Montreal protocolhas causeda decrease inthe stratospheric halogenload anda slowing ofozone decline is expected to be thenaturalprecursor of acomplete ozone recovery

∗ Corresponding author.

E-mail address: didier@oma.be (D. Fussen).

toward2050. Clearly,themonitoringofozonestratosphericabun- dances is of crucial importance in assessing the milestonesof a clearrecovery processonaglobalscale[38,40].Basedonsatellite andground-basedobservations,thereispresentlyexperimentalev- idencethat upperstratospheric ozoneis increasing.On theother hand,trendsinlowerstratosphericozonearemostlynegative(al- thoughstatisticallyinsigniﬁcant)showingthenecessitytocontinue the record of long term ozone proﬁles [39,44]. In the Antarctic, there are indications that the size and depth of ozone hole has decreasedsince2000[44].

Among important trace gases, methane is very important for its impacton climatethrough alarge radiative forcing effectand theproductionof stratosphericwatervapour.The globalincrease of about0.7ppm in 1800 CE to 1.8ppm nowadays is diﬃcult to quantifybecauseofthediversityofthesources:wetlands,enteric fermentation,ﬁres andriceagriculture.The oddhydrogen family, https://doi.org/10.1016/j.jqsrt.2019.06.021

(2)

2 D. Fussen, N. Baker and J. Debosscher et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 238 (2019) 106542

Fig. 1. Past and future population of atmospheric limb sounders, according the WMO’s OSCAR database [27] . The colors represent the measurement techniques: limb scattering (blue), emission (red), solar occultation (magenta), and stellar occultation (cyan). The dashed lines illustrate the potential extended lifetime of missions currently operating beyond their nominal lifetime.

HOx,contains all active species,i.e. radicals that are involved in catalyticcyclesthatdestroyOx.TheHOx radicalsarederived pri- marilyfromtheoxidationofwatervapourinthestratosphereand it is thereforeessential to understand andto monitor the intru- sionofwatervapourintothestratosphere,especiallyintheregion ofthe tropicaltropopause[15].Forwatervapor intheuppertro- posphereandlower stratosphere,it wouldbe extremely valuable toaugment the balloonnetwork with satellite sensors,providing highverticalresolution(betterthan2km)measurementsofwater vapor in the UTLS on a time scale of decades [24]. Such a high verticalresolutionwillverylikelyonlybeachievablethroughlimb soundingsatelliteinstruments.

Similarly,theNOxfamilyisknowntoplayanessentialcatalytic role inozone destructionwith astrong diurnal cycleandits ob- servationrequiresday-andnight-timemeasurements(atdifferent localsolartimes)foracompleteunderstanding.

Much remains to be learned about polar stratospheric clouds (PSC) that support the setup of heterogeneous chemistry in the standard model of the polar ozone hole and for which we do not know enough about particle sizes, crystal morphology and even composition. The same uncertainty holds for polar meso- sphericclouds(PMC),similar inappearance tothincirrusclouds, butlocatedatamuch higheraltitude,from80to87kmnearthe mesopause. PMCs onlyoccur at highlatitudes duringsummer (a few weeks before and after the solstice), when the mesosphere becomesextremely cold(withtemperaturesaslow as100 K)[5]. Their increasing occurrence may be a sign of climate change in the upper atmosphere. Some sensors can access to the micro- physicalinformation(asCALIOPanditsdual-polarization capabil- ity[29]) that allows classiﬁcation of PSCsaccording to composition.However, going back to the physicalnature of the observation,mostatmosphericsensors are actually measuring integrated extinctionproﬁles (oraerosolslantopticalthicknesses)with,pos- sibly,someinformationaboutthephasefunction.Thederivationof particlesizedistributions(andafortioriofPSC/PMC/aerosoltypes) requiresexternalinformation(composition,temperature,shape,...) and should be considered as a complementary optical inverse problem.

Ontheother side,thescientiﬁccommunityisrequiringanur- gentresponsetothepresentlackofatmosphericremotesounders

(see Fig. 1) witha high verticalresolution. Since the endof EN- VISATin2012,someaginglimbmissions/instruments arestillop- erating (ODIN/OSIRIS [19], AURA/MLS [9], ACE/FTS-MAESTRO [6], OMPS-NPP/LS[23])buttheyareallbeyondtheirnominallifetime.

ApartfromISS/SAGE-III (2016) recentlyoperated,JPSS-2/OMPS-LP (2022)istheonlyapprovedmissionsofarandthereisnoguaran- teeconcerningapossibleoverlapbetweenthem.

This article presents the concepts of a mission called ALTIUS (Atmospheric Limb Tracking for the Investigation of the Upcom- ing Stratosphere). The mission concept has evolved from an in- orbit-demonstrationto a semi-operational missionwith ozoneas mainoperationalobjectiveandothertracegasesassecondarysci- entific objectives. The expected ALTIUS capacities were found to haveafaircompliancelevelwiththerecommendationsissuedby the ESA study ’OPEROZ’ [43] to identify the minimum require- mentsforanozonemission.After technicalandscientificreviews byESA,ALTIUSwassuccessfullyproposedbytheRoyalBelgianIn- stituteforSpace Aeronomy(BISA) tothe BelgianScientific Policy Office(BELSPO)andhasbeen recentlyapprovedbythe European Space Agency (ESA) asan element of the ESA Earth Watch pro- gramme.Thegeneraldescriptionreadsas“ALTIUSaimsatthede- velopment,launch,in-orbitoperation,dataprocessing,dataarchiv- ingandproductsdistributionofalimbsoundermissionbasedon a small satellite. Its main objectiveis the monitoring of the 3-D distribution andevolution of stratospheric ozone athigh vertical resolution.The versatileinstrument, whichwillimage theEarth’s limbinthenearUV, visibleandnearinfraredspectral regions,in combinationwiththe agility provided by thespacecraft platform canallowmeasurementsofconcentrationprofiles ofotherspecies aswellasofaerosolextinctionverticalprofiles.”

Thepresentpapergivesanoverviewoftheobjectives,concept andexpectedperformancesoftheALTIUSmissionasproposed by BISA,onitswaytoapreliminarydesignreview(PDR),expectedat theendof2020.

1.2. ScientiﬁcobjectivesoftheALTIUSmission

Inviewoftheneedforatmosphericcompositionmeasurements onaglobalscaleandwithahighverticalresolution,itisusefulto

(3)

group the mission objectives intwo subsets ofprimary andsec- ondaryscientiﬁcrequirements(SR).

1.2.1. Primaryscientiﬁcrequirements

SR1Global andlong-termverticallyresolved ozonedata setsof observationsinthestratosphereareurgentlyrequiredtoas- sessmodelbehaviorandtestmodelpredictions,particularly intheuppertropospherelowerstratosphere(UTLS)domain, inpolarregionsandintheSouthernHemisphere.Today,the temporal recovery of ozone at different altitudes andlati- tudeshastobeconsolidatedandquantiﬁed.

SR2The measurement of ozone proﬁles in the middle strato- sphereat5%accuracylevelhastobecontinuedtoreinforce thesigniﬁcanceoftheexistingclimatologies.Aspecialeffort isneededintheUTLSregiontoimprovethedetermination oftrendsandtodetectchangesinupwelling[30].

SR3Thespreadofexistingozoneproﬁledatainozoneholecon- ditions is not satisfactory. Limb observations should allow determiningspatialconcentrationgradientsacrossandalong thevortexwhiledetectingPSC’sduringpolarnightfordata screeningandcorrelativeanalyses.

SR4ALTIUSshould alsoprovidehighlyresolved vertical proﬁles ofmesosphericO₃ [17],akeyspeciestounderstandtheat- mospheric coupling, not yet well understood,between the lower atmosphere andthe upperstratosphere, mesosphere andlowerthermosphere.

1.2.2. Secondaryscientiﬁcrequirements

SR5ALTIUSwillprovidehighlyresolved verticalproﬁles ofNO₂ from the UTLS to the upperstratosphere, mesosphere and lowerthermosphereatdifferentlocaltimes.Itwillquantify anti-correlations ofNO₂ andO₃ concentrations.It willalso focusonthedynamicsandthechemistryofstrongNO₂ en- hancementsintheupperstratosphere–mesosphere[13]. SR6Themissionwillprovideglobalverticalproﬁles ofH₂Oand

CH₄, with a special focus,for water vapor, on the tropical tape recorder and vortex dehydration. There exists a sub- stantialuncertaintyinsimulatedH₂O,evenwhenusingthe mostrecent reanalysisproducts [31]. Amajor objective for theseinterrelatedspecieswillbethemeasurement oftheir trendsinthelowerstratosphere.

SR7ALTIUS will provide extinction proﬁles, in the UV–vis–NIR wavelength ranges, from which particle size distributions of stratospheric aerosol can be derived. Depending on the volcanic state of the atmosphere, it will assess the relax- ation timesofthe volcaniccontributionortheevolutionof the background non-volcanic contribution. ALTIUS aerosols areexpectedtocontributetotheglobalspace-basedstrato- spheric aerosol climatology (GloSSAC) recently published [42].

SR8ItisimportanttomeasurethefrequencyofPSCoccurrences inside the polar vortex and their temperature dependence (fortypeidentiﬁcation).

SR9ALTIUSwillmeasure thetrendandthephaseinthe occurrence of PMC’s as well astheir median altitude and their horizontalextent,inbothhemispheres,aroundthesummer solstice.

SR10 ALTIUS will measure OClO (in polar [2] and equatorial [10]regions),BrOandNO₃, thatare importantminortrace gasesinvolvedinthestratosphericchemistry.

SR11 In solaroccultationmode, by making useof imagingtech- niques atlarge S/N ratio, ALTIUS will retrieve density and temperatureproﬁles up tothe mesosphere,fromrefraction anglemeasurements.

SR12 Horizontal concentration gradients of relevant trace gases willbeobservedinalongtrackandacrosstrackgeometries,

allowing fortomographic retrievals duringsuccessive revo- lutions.

1.3.ALTIUSScientiﬁcproductandmissionrequirements

InTable1,wereportthetargetscientificproductrequirements related to the abovementioned scientific objectives. The baseline assumptionforALTIUSsupposesalimbsounderonboardamicro- satellite platform (a PROBA class carrierwas pre-selected for its agility and pointing performance). This choice ensures a global coverage (including polar regions) if launched in a standard he- liosynchronouslowearthorbit(LEO)witha10:00 a.m.localtime atdescendingnode(LTDN),andthecapabilitytoachievethehigh- estvertical resolution. As a thresholdrequirementfor theuse of ALTIUSdata in present assimilation systems, the globalcoverage shouldbe sampledon agridfinerthan 5–10^◦ inlatitudeand10^◦ inlongitude.

Globalcoverageandhighverticalresolutioncanbeachievedby combiningseveralobservationmodesbasedon theinteraction of thelightemittedby acelestialbodyandtheatmosphere,suchas thelimb-scatteredsolarlight,solar(andlunar)occultations,stellar (andplanetary)occultations.Indeed,brightlimbmeasurementsof- feralargesamplingonthedaysideoftheorbit,whileoccultations addanumberofmeasurementpointsonthenightside.

This multimode capacityrequires some agility,autonomy and stability from the satellite, which is precisely what has been demonstratedbythe PROBAplatform.Afurtherconstraintisthat ALTIUS requires to be the only payload with pointing requirements onboard. Limb measurements (not based on atmospheric emissions)areverysensitivetothetangentaltituderegistrationof theline-of-sight(LOS),i.e.the closestpoint tothe localgeoid,es- pecially wheretheconcentration profiles show large verticalgra- dients. Previous limb-scatter instruments have experienced seri- oustangentheightmisregistrationissues(SCIAMACHY[25],OSIRIS [19],OMPS[23]).Occultationsalsoneedtokeepthelightsourcein thefield-of-view(FOV).TheoriginalapproachproposedforALTIUS istouseanimagingsystemwithaFOVmatchingtheapparentsize ofthebrightlimb(0–100km).Withthisfeature,in-flightcalibra- tionmethods are more easily implemented to solvethe pointing issue.Theentireatmosphereisprobedatonce,whichisanadvan- tagecomparedtoscanningsystemswhichtaketensofsecondsto completethescan(andloseinalong-trackresolution).Inaddition, inertial pointingto the occulted celestial bodiesis done without theneedforcomplexlightsourcetrackingsystems(asitwasthe caseforGOMOSonENVISAT).

2. Instrumentrequirementsandoperationconcepts 2.1. Instrumentconceptandrequirements

In order to solve the tangent height misregistration problem, ALTIUSis designed asa limb imager forwhich the ﬁeld-of-view canbe calibratedby differenttechniques.Assuminga rectangular FOV,theatmosphericlimbshallbeimagedbetween0and100km (about34 × 34mrad)althoughstellarandplanetaryoccultations canbeperformedwithasmallerFOVof3.4 × 3.4mrad.Thepixel instantaneousﬁeld-of-view(IFOV)willbelessthanorequalto0.2 mradallowingforaverticalsamplingbetterthan0.6km.

Clearly,limb imagingimposes stringent pointingrequirements on the system. If we consider a Cartesian Boresight Reference Frame(BRF) centeredat thedetectorwith theZ axis pointingto thelimbhorizon,theX andYaxisrespectivelyperpendicularand paralleltothehorizon,wecandeﬁneangularuncertaintiesrelated totherotations RZ,RX andRY aroundtheseaxes.The correspond- ingrequirementsarereportedinTable2fortheMeanPerformance Error (MPE), Mean Knowledge Error (MKE), Relative Performance

(4)

4 D. Fussen, N. Baker and J. Debosscher et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 238 (2019) 106542 Table 1

Table of ALTIUS scientiﬁc product requirements.

Scientiﬁc objectives Altitude range [km] Vertical resolution [km] Horizontal resolution [km] Total error [%] (target/threshold) Coverage (revisit time)

SR1 [O 3] 15–45 0.5 50 5/20 Global (3 days)

SR2[O 3] 20–45 0.5 50 3/10 Global (3 days)

SR3 [O 3] 15–45 1 20 10/30 Polar (3 days)

SR4 [O 3] 60–100 1 50 10/20 Global (3 days)

SR5 [NO 2] 20–45 1 25 15/40 Global (3 days)

SR6.1 [H 2O] 15–30 1 25 5/20 Global (3 days)

SR6.2 [CH 4] 15–45 1 50 2/5 Global (3 days)

SR7 [Aer] 15–35 1 10 10/100 Global (1–3 days)

SR8 [PSC] 10–25 0.5 10 30/100 Polar (1–3 days)

SR9 [PMC] 75–90 0.5 10 30/100 Polar (1–3 days)

SR10.1 [OClO] 15–50 2 50 20/50 Global (3 days)

SR10.2 [BrO] 10–30 2 50 5/10 Global (3 days)

SR10.3 [NO 3] 20–45 1 25 15/40 Global (3 days)

SR11 [T] 15–100 2 20 0.5/2 Global (3 days)

SR12 [ ∇^n] ^15–45 ¹ ² ^5/50 Polar (1–3 days)

Table 2

Summary of instrument pointing and operational requirements.

FOV 100 km × 100 km IFOV 200 μrad (0.6 km)

Pointing Axes Requirement [mrad] Timescale [s]

MPE R Y, R X, R Z < 2, < 2, < 2.5 1

MKE R Y, R X, R Z < 0.1, < 1, < 2.5 1

RPE R Y, R X, R Z < 0.1, < 1, < 2.5 1

RKE R Y, R X, R Z < 0.1, < 1, < 2.5 1

PDE R Y, R X, R Z < 0.1, < 1, < 0.25 10

KDE R Y, R X, R Z < 0.1, < 1, < 0.25 10

Limb scattering Average number Typical duration

observations/orbit 50 10 s/observation

images/channel/observation 4 1 s/image total images/orbit 600

Stellar occultations Average number Typical duration

occultations/orbit 10 125 s

observations/occultation 25 5 s/observation images/channel/observation 4 0.5 s/image total images/orbit 30 0 0

Solar occultations Average number Typical duration

occultations/orbit 2 125 s

observations/occultation 125 1 s/observation images/channel/observation 4 0.1 s/image total images/orbit 30 0 0

Error(RPE),RelativeKnowledgeError(RKE),PerformanceDriftEr- ror(PDE) andKnowledgeDriftError (KDE)(see [28] foran exact deﬁnitionoftheseerrorterms).

ALTIUSmaybedescribedasanimagingspectrometer ofmod- eratespectralresolution.The instrumentwill bea tuneablespec- tralimager capable ofobservingthe atmosphericlimb inthe UV (250–400nm),vis(420–800nm)andNIR(800–1800nm)domains, ensuringaspectralresolutionalwaysbetterthan10nmforvisand NIRandbetterthan2.5nmforUV.Adistinctspectraltuneableel- ement will ﬁlter observations in each of these three wavelength ranges.

In general, the instrument will operate at different measure- mentratescorrespondingtothe differentobservationgeometries.

Forlimbscatteringobservations, thealong tracksamplingrate is consistentwiththeeffectiveopticallengthalong theline-of-sight (about500km).It isimportantto noticethat alimb pointingin- strumentisunabletoachieveaverylargeswathlikeanadirlook- ing sensordue to the Earthcurvature. In Table 2, we report the typical number of images (spectral snapshots) and observations (minimumensembleofimagesatdifferentwavelengthstoretrieve thespectralabsorberproﬁles).

InTable3,weidentifytheusefulspectralwindowstomeasure theconcentration proﬁles ofthe main target gasestogether with therequired maximumspectral width, minimumSNR andmaxi- mum acceptable pointing error.In limb mode, a typical set of 4

wavelengthsperchannelwillberecordednominallyin10stoob- tain suﬃcient spectral information content with a maximal geo- graphical resolution. However, it will be possible to increase the measurementtimeupto50s(oreven100sintheworstcase)to improvetheS/N ratio.Pixelbinning (uptothefull detectorrow) could alsobe madepossibleforthesamepurposeatthe priceof areducedhorizontalresolution.

2.2. Instrumentdesign

Ifthegoalwastotakespectralpicturesatﬁxedwavelengths,a ﬁlterwheelcouldhavebeenused.Butthenthewavelengthswould befrozen,withasupplementaryriskofmechanicalfailureinspace environment.Toovercometheseissues,ALTIUSwilluseatunable approachbasedon:

• acousto-optical tunableﬁlters (AOTF)forthe vis andIR chan- nels.

• Fabry–Pérotinterferometers(FPI)fortheUVchannel.

AOTFsaresmallbirefringentcrystals(typicallyafewcubiccen- timetres) serving as interaction medium between the incoming light andan acousticwave propagatingin thecrystal[3,4,45].By carefullyselecting theacoustic wave frequency, theacousto-optic interaction induces a Bragg diffraction regime that deﬂects light withaspeciﬁc wavelengthaway fromitsincident direction(bya

(5)

Table 3

Table of ALTIUS spectral windows and target species. Atmospheric domains: UT = upper troposphere, LS = lower stratosphere, US = upper stratosphere, MS = mesosphere.

Species Region Spectral range (nm) Spectral width (nm) Limb Stellar Occ. Solar Occ. Minimal SNR

O 3 UT-LS 550–650 10 x x x > 200

O 3 US 300–330 3 x x x > 70

O 3 US 550–650 10 x x x > 50

O 3 MS 250–300 3 x x > 20

O 3 MS 1260–1280 3 x > 10

NO 2 LS-US 440–470 2 x x x > 1500

H 2O UT-LS 90 0–180 0 5 x x x > 500

CH 4 UT-LS 160 0–180 0 5 x x > 500

Aerosol/PSC UT-LS 250–1800 10 x x x > 200

PMC MS 250–1800 10 x > 20

OClO UT-US 320–400 2 x x > 100

BrO UT-LS 320–360 2 x x x > 500

NO 3 LS-US 662 5 x > 50

T UT-MS Visible 50 x x > 10 0 0

[ ∇n] UT-LS Visible 10 x > 50

few degrees).Thedivertedbeamcanthenbe collectedby anoff- axisopticaldetector.Usedinsideanimagingsystem,suchadevice offers manyadvantages: it is smalland lightweight, contains no moving parts, it consumes only 1–3W and can be tuned to the required wavelength in a few milliseconds. An AOTF works over a broad spectral range(hundreds ofnanometres) witha variable bandwidth(quadraticinwavelength)thatcanbeoptimizedbyde- sign(fromsub-nmto10nm).

AsAOTFcrystalsintheUVregionarenotyetmatureenoughto be considered asa suitable technology forspace application, the ﬁlter element forthat region will consist of a cascadeof Fabry–

Pérotinterferometers.

ALTIUS benefits from a relatively simple optical design. The three spectral channelssharethe sameidea: an aperture collects the(ir)radiance ofthescenewhichisthen reflectedbyanumber ofmirrorstoformanimageontoadetectorarray.Locatedapprox- imately half-way betweenthe aperture andthe detector,a tune- ablespectral filtercapturesafractionofthe lightspectrumwhile therestoftheincidentbeamisblockedbythecombinedactionof a beamstop andcross-oriented polarizers(AOTF), orbythe filter itself(Fabry–Pérot).

Eachchannelismadeofreﬂectiveopticsandcontainsthesame setoffunctionalmodules:

• amechanicalstructure

• afront-endopticsgroup(FEO),toguidetheincominglightto- wardthespectralelement

• aspectralﬁlter(AOTFmoduleorFPIassembly)

• a back-end optics group (BEO) to focus the ﬁltered image on thedetector

• adetectorwithoptionalcoolingcapacity

• adrivingandread-outelectronicboard.

At the detectorlevel, a512×512 CMOS image sensor is pro- posed, suitable for atmospheric absorption measurements as a hyper-spectralapplicationintheUVandvisiblewavelengthranges.

For the NIR detector, the required SNR imposes a 190 K work- ingtemperaturewhichwillbeobtainedby usingaStirlingcooler.

Such adeviceiscapableofsigniﬁcantlyreducingthechipoperat- ing temperatureattheexpenseofapowerconsumption ofabout 5Wandalimitedlifetimethatwillhavetobetakenintoaccount inthemissionscenario.

An extended set ofon-ground calibrations is foreseen tofully characterize the sensor. Instrument performance will be checked atanaccuracylevelcompatiblewiththesensorrequirements(typ- ically1/10pixelforLOSandpointing,1–3%forradiometricquan- tities,1/10ofspectralbandwidthforwavelength).Inparticular,the followingcharacteristicswillbecalibrated:

• LOSandpixelmappingforFOV,IFOVanddistortion.

• Temperaturedependenceofpixelmapping.

• MTFandpointspreadfunction.

• Radiometric response: PRNU, SNR, pixel non-linearityanddy- namicrange.

• Straylight.

• Spectral response:wavelengthcalibrationandtemperaturede- pendence,bandwidthandout-of-bandrejection.

Sofar, the ALTIUSsensor design has beenperformed through manytechnicalstudiestodemonstrateits feasibilityanditscom- plianceto themission requirements.All thesestudyphaseshave been thoroughly reviewed by several technical committees. No showstoppers were found whereas missing studies or optimiza- tions have been identiﬁed: a detailed description of the payload/platform interface, a global optimizationof the UV channel, the detector layout, the payload electrical design, a consolidated thermalanalysisfocused ontheassessmentof thethermo-elastic effects(andtheirimpactonthepointingrequirements)andacom- prehensive straylight analysis.All these missing technicalstudies will haveto be conductedbefore the PreliminaryDesign Review (PDR)presentlyforeseenattheendof2020.

2.3.Platform,orbitandoperations

2.3.1. Platform

ThepayloadhasbeendesignedtofitinanevolvedPROBAplat- formdesign,asshowninFig. 2(internalview)andFig. 3(external view)below. The platform generic design hasbeen modified inordertoimplementa propulsionsystem, toprovidetheneces- sary accommodationvolume forthe instrument aswell assuffi- cientpowerto sustainthemission,andtoremain withinvolume constraintsforasetofrealistic launcherscompatiblewiththeAL- TIUSmission.

2.3.2. Orbit

Heliosynchronicity is a standard requirement for a limb- scatteringinstrumentasairmassesare soundedataconstantlo- cal solar time, while it makes the radiative transfer problem of the limb-scatteredlight more easily addressed by narrowing the rangeofsolaranglestoasmallsubset(driftoftheorbitallocalas- cendingnode will be limitedto 0.5h over themission lifetime).

Arevisit time ofabout3 days wasselected (similarto ENVISAT) withsomelooserequirementsontheexactoverlappingoftheS/C groundtracks. Fora low earth circularorbit (LEO)at an altitude of680km,thisallowsforaninteresting trade-off betweenashort revisittimeandadensegeographicalsamplinggrid(about800km

(6)

Fig. 2. PROBA-ALTIUS internal view: the payload is red and equipped with three star trackers.

Fig. 3. PROBA-ALTIUS external view with solar panels deployed.

betweensuccessivetracksatequator)thatmatchesthehorizontal resolutionofdataassimilationmodels(f.i.BASCOE[7]).

2.3.3. Baselineoperations

During 80% of the mission time, ALTIUS will be operated ac- cordingtoaso-called”baseline” scenariodepictedinFig.4:

1.Onthedayside(brightlimb),itobservesscatteredsolarradi- ationinthebackwarddirectionwithrespecttothevelocity vector

2.When approaching the terminator,a spacecraft manoeuvre brings theSunintotheFOVtoobserveasunsetoccultation ininertialpointingmode

Fig. 4. ALTIUS observation geometry and baseline operations: (1) backward limb, (2) sunset occultation, (3) star occultation, (4) Moon occultation, (5) planet occultation, (6) sunrise occultation.

3. Onthenightside(darklimb),5–10bodies(starsandplanets) areselectedandralliedtoobservetheiroccultation 4. Attheterminatoragain, ALTIUSobservesasunriseocculta-

tion

5. Backtostep1,foranewcycleofobservations.

Ashortanimation explainingmostofthepossibleobservation geometrieshasbeenproducedandisavailableonline[12].Allsys- temresources(energy,thermalbudget,telemetry,...)havebeenop- timized fromthisscenario. However,20% ofthe missionscenario isspareforlessregularactivities,includinginstrumentcalibration campaigns. The sequence ofnight objects (stars, planetsand the Moon)selectedforcandidateobservationswasoptimizedwithre- specttothenumberofnecessarysatellitemanoeuvres.

InFig.5,wereporttheensembleofsuccessiveobservationsfor one satellite revolution in the baseline scenario, with successive sunrise occultation, limb scatteringobservations indayside, sun- setand, inthenightside,stellar,planetaryandlunaroccultations.

AfulldayofALTIUSobservationsisplottedinFig.6,wherethege- olocationsofobservationsare progressingwestwardbyabout24^◦ betweentwosuccessiverevolutions.Thesimulationwasperformed foraquasi-circular(eccentricity =1.05×10⁻³ degree)orbitwith asemi-majoraxisof7050kmandaninclinationof98.07^◦.Figs.7 and8respectivelyrepresenttheseasonalevolution ofbrightlimb andsolaroccultationgeolocations, thelatter onesbeingconﬁned to ratherhighlatitudes asaconsequence ofthealmost polaror- bitinclination.Ontheotherhand,lunaroccultationsindarklimb are also possible andallow for tropical observations (see Fig. 9).

Finally,wealsoreportthegeolocationsofbrighteststarandplanet occultations inFigs. 10and 11, witha ratherlimitedcoverage of ozoneholeconditions.

2.3.4. Specialobservationmodes

Asaserendipitousmode ofmeasurement,it shallalsobe pos- sibleto perform stereoscopic observationsof a species3D distribution ﬁeld in the low latitude regions. Due to the remarkable (andfortuitous)factthattheazimuthalangulardistancetotheside horizonisclosetotheanglebywhichtheEarthwillhaverotated atthe next LEO revolution, itis possible tocombine a backward limbobservation followedbya dedicatedsideward limbobserva-

(7)

Fig. 5. Observation in the baseline scenario for one revolution. ALTIUS successively performs limb (green) observations, solar occultations (yellow), planet occultation (magenta) and stellar occultations (blue) measurements. For the occultations, the geolocations of line-of-site tangent points are shown. The beginning and the end of the occultation measurements are associated with the corresponding orbit sectors. Only stars/planets with a magnitude brighter than 3 are considered. The terminator position, that changes during a revolution, was calculated at each longitude when the orbit crosses the longitude.

Fig. 6. Daily coverage for the orbit (3-day revisit time). See text for the orbital parameters. Green: limb measurements, blue: stellar occultations, yellow: solar occultations.

(8)

Fig. 7. Annual coverage for bright limb observations, limited by the extension of the polar night. We consider here “pure” bright limb situations with a maximal solar zenith angle at the tangent point of about 70. The color scale refers to the number of observations per day.

Fig. 8. Annual coverage of solar occultations (sunrises and sunsets).

(9)

Fig. 9. Annual coverage of lunar occultations. The minimal SZA at the tangent point is 102, the minimal phase 0.5.

Fig. 10. Annual coverage of bright stars with different colors for different star tracks of the remotely sounded geolocations. Stellar magnitudes are less than 1, the minimal SZA at the tangent point is 102.

(10)

Fig. 11. Annual coverage of planets. Magnitudes are less than 3, the minimal SZA in the tangent point is 96.

tionattheconsecutiveorbit. Thismode iscalledstereoscopicbe- causethe samelocation is observed witha smalldelay ofabout 1.5h, from 2 almost orthogonal directions in equatorial regions, whichallowsforsome 3-D informationretrievalof thegeophysi- calfields.Inparticular,inhomogeneitiesalongthefirstline-of-sight canbedetectedbyacomplementarysidewardobservationandcan possiblybeused toconsolidatetheretrieval ofavertical concen- trationprofile.

EvenifthemissionscenarioofALTIUSisdrivenbythebaseline observationmodesdescribedabove,theplatformhighmanoeuvra- bilityenablesapointinginmanydirectionsfordedicatedandlim- itedcampaigns.Takingintoaccountthetotalavailablepower,ther- malbudgets,datatransmission,exposureofstartrackerstodirect solarlight,etc,thefeasibilityandprogrammingofthesecampaigns willrequirecase-by-casestudies. Thereforethey will onlybe im- plementedaftersuccessfulcommissioning.Asanon-exhaustivelist ofdedicatedobservationmodes,wecanmention:

• Speciﬁcairglowemissions

• Lunaroccultations

• Inter-validationwithnadirlookinginstruments

• Volcanicevents

• Colourofcloudsmeasurements

• Effectofsolarprotonevents

• “Tangent” occultations (celestial body does not set for a very obliqueoccultationandisgrazingthehorizon)

• Polarization-sensitivemeasurements(byrotatingthesatellite)

• Photochemicaleffectofsolareclipses 2.4.Multimodeobservations

2.4.1. Brightlimbimages

InFig.12,brightlimbradiancesimulationshavebeenobtained fromthe radiative transfer code MODTRAN 5, for a US standard atmosphere 1976 (mid-latitude summer model) and climatologi-

calvaluesforthemainabsobers,withabackgroundstratospheric aerosolloadandnoclouds.The tangentpoint islocatedat45^◦N, 0^◦E. ALTIUS looks at this location from a sun-synchronous orbit at10:30localtimeatdescending node(LTDN),backwardwithre- specttotheplatform velocityvector. Theradiancewascalculated for19tangentaltitudesandexhibitsalargedynamicalrangewith respecttothetangentaltitude.

Subsequently,rawALTIUSimagesweresimulatedforanywave- lengthineachchannel.Whenpossible,thepixelgainandtheex- posure timehavebeenoptimizedtomaximizethesignal without saturation.Ingeneral,verylittlevariationoftheatmosphericcom- positioncanbeexpectedalongthehorizontaldirection.InFig.13, weshowatypicalALTIUSbrightlimbimageintheUVchannel,not takingintoaccountthe cloudinessofthescene.Thenoise isrep- resentativeofthetotalradiometricuncertaintyandaverticalcross sectionallowstoeasilyidentifytheradiometricknee.

2.4.2. Solaroccultationimages

TheSunisanextendedlightsourceofwhichtheapparentsolid angle is fully covered by the 2-D detectors. A model of the so- larabsoluteradianceﬁeldwasdevelopedin-house.Itiscapableof simulatingthesundiskatanywavelengthandtangentaltitude,in- cludingthesolarlimbdarkeningeffect.Araytracingmodelisalso available tocompute the pathandtheextinctionthrough theat- mosphere.OneimageofthehighSunhasbeensimulatedinFig.14 at

λ

=500nm.Atinstrumentlevel,theonlyradiometricdifference betweenabrightlimbacquisitionandaSunimageistheinsertion ofaneutraldensity(ND)ﬁlterintheopticalpathbyamechanism.

2.4.3. Staroccultationimages

AsobservedbyALTIUS,starsarepoint-likesourcesslightlyde- focused by the point spreadfunction (PSF) ofeach channel. Raw ALTIUSspectralimagesofstarsarethereforeessentiallynoisewith a bright spot spread over a few pixels. A simple Beer–Lambert

(11)

Fig. 12. Limb-scattered spectral radiance as a function of tangent altitude (see curve label in km) and wavelength. Simulated with MODTRAN 5 with the US standard 1976 mid-latitude summer atmospheric model at 10:30 a.m. local time.

Fig. 13. Left: simulated raw image acquired by the UV channel at the central wavelength of 300 nm, an exposure time of 2 s, the sensor in high gain and at a temperature of 20 ^◦C. X - and Y -axis are labeled according to the tangent point coordinates. The color scale shows the registered signal in digital numbers (DN). The noticeable noise is representative of the total uncertainty on the electronics signals (containing the signal and dark current shot noises, the readout and the quantization noises). Right: vertical cross section taken in the image at column 100 (black curve), or from the averaging of columns 101–117 (red curve). The signal proﬁle exhibits the well-known radiometric knee used by other limb-scatter instruments to constrain their tangent height uncertainty.

modelhasbeendevelopedtosimulatestellaroccultations.Itcom- putestherefractedopticalpathandtheextinctionbyallabsorbing (orscattering)speciesthroughtheatmosphericlayers.Forthesim- ulationoftheexpectedatmosphericslanttransmittancepresented in Fig. 15, we used averaged data fromthe occultationof Sirius asmeasuredbyGOMOS,andsolved theraytracingproblemfora standardatmosphere(summermid-latitude).

2.5. In-ﬂightcalibrations

Due to volume limitations, the platform will not be carry- ingcalibratedlightsources orspectroscopicreferencestomonitor drifts ofinstrumental characteristics. Onlyvicariousmethods will provide the needed information. When unavailable during nominal observations, calibration campaigns will require acquisitions andmanoeuvres thatwill interrupt thebaseline data production.

The heliosynchronicity of the orbit will introduce repetitive pat-

terns overdifferent time scales (orbit, seasons) allowing for pre- dictabilityof the change in performance. Also, the rather simple instrumentdesignlimitsthesourcesofpossibleperformancedrift anditisexpectedthatthetimeallocatedtocalibrationcampaigns willremain well below the limitsofthe 10% spare missiontime reservedfornon-baselineoperations.TheimagingcapabilityofAL- TIUSis a strongadvantage especiallyforthe pointingcalibration.

Furthermore,theagilityofthePROBAplatformallowsforasetof differentapproaches,potentiallybringinganysourceoflightinthe ﬁeldofview.Hereafter,threemaincalibrationtopicsareidentiﬁed:

radiometriccalibration(andimageacquisitioningeneral),pointing registrationandspectralregistration.

2.5.1. Radiometriccalibration

Atmosphericlimbsoundingby theoccultationtechniqueisra- diometricallyself-calibrating becausethe transmittance along the optical path is computed by dividing the observed spectrum by

(12)

Fig. 14. Simulated spectral image of the Sun as seen from ALTIUS at 500 nm. The solar limb darkening, and the atmospheric refraction are the key parameters of the solar occultation forward model.

Fig. 15. Atmospheric slant path transmittance as a function of its apparent tangent altitude for several wavelengths (in nm). The extinction is caused by neutral air and aerosol scattering, O 3and NO 2absorption and refractive dilution.

anexo-atmosphericreferencespectrum.Brightlimbretrievalsalso rely on normalized measurements with respect to a reference spectral radiance recorded at a selected tangent altitude (gener- allyabove the vertical region of interest) andan absoluteradio- metriccalibrationistherefore notmandatory.Forphoto-response non-uniformity (PRNU), no ﬂat ﬁeld device will be available onboard,sotwopossibleapproachesareconsidered:

• to builda syntheticﬂatﬁeld froma numberofcloudyscenes imaged at nadir: the complex patterns found in independent imagesarecompletely smoothedoutby therandomcombina- tionoftensofthem.

• tousetheMoonortheSunasextendedlightsourcesthatcan bescannedacrosstheﬁeldofviewbyasatellitemanoeuvre.

Residualstraylightanalysiscanbeperformedbymakinguseof aslightdepointingoftheline-of-sightand,fortheAOTFchannels,

by exploitingthe deﬂection of the ﬁltered beam at the selected wavelength (i.e. straylight issued from the complementary spec- trumwillnotfollowthesameopticalway).

2.5.2. Pointingcalibrationandtangentaltituderegistration

From ﬂight experience withprior missions (f.i. PROBA-2), the PROBAplatformhasdemonstratedarelativeperformancepointing error (RPE)of 25

μ

^rad ^(at ²

σ

signiﬁcance) over 5 s, well below the100

μ

^rad^for^the^ALTIUS^required^RPE.^During^limb^scattering

observations,theplatformwillberequiredtopointtheinstrument LOStowardthelocalgeoidatthehorizon.ThePROBAattitudeand orbitcontrol system(AOCS)uses threestartrackers todetermine thespacecraftattitude.Thisinformationwillbetime-taggedonev- ery acquiredimage. In limb geometry, the AOCS istherefore the primarysourceofinformationforthedeterminationoftheinstru- mentlineofsight(LOS)pointing.

The on-ground calibration will determine the relationship between the star trackers LOS, and the individual boresight refer- enceframes(BRF)ofthethreechannels.Asthisrelationshipmay change afterlaunch (0-grelaxation) andunder differentthermo- elasticconditions(orbitalandseasonaleffects),thereisaneedfor anindependentassessmentoftheinstrumentpointing.Obviously, during occultations the absolute pointing is calibrated by astro- nomicalephemeridesanditcanbe usedtovalidatethe AOCSin- formation.This validatedquantity can be re-usedfor brightlimb measurements, although therelatively differentillumination con- ditions,andhencethermo-elasticenvironments,haveto betaken intoaccountandmodeled.Thebaselineapproachforthevalidation of thestar trackerinformation duringbrightlimb measurements consistsinidentifyingbrightstars(orplanets)shiningthroughthe highestatmosphericlayers.Thisensuresthatthestarsignalholds strong,andthat no refractiveeffectperturbs thetangent altitude determination.Itturnsoutthatthelikelihoodofﬁndingusefulstar candidatesforalmostanylatitudeishigh.Eventually,thepointing calibrationcanalsobecross-calibratedwithrespecttogeographi- cal details (ondayside)andenlighted cities(on nightside)with orwithoutdepointingtowardnadir.

2.5.3. Spectralcalibration

AnAOTFcanbeeasilysweptoversmallspectral windowscon- tainingknown starspectral features (Fraunhoferlines). The iden- tiﬁcation ofthesefeatures enables directwavelengthregistration.

Withitsspatialextension,theSunisanidealtargetanditalsoof- fersthepossibilitytodiscoverwavelengthshiftsproducedbyther- malgradientsinsidetheAOTF.Inaspectralcalibrationsimulation (using asweep ofspectral imagesoftheSun), thewavelengthof the minimum of the hydrogen

α

^Balmer ^absorption ^line ^around

656nmcouldbedeterminedbypolynomialﬁtwithanuncertainty of 0.13nm with only one1 pixelintensity, orvirtually nouncer- taintybysummingallthepixels(seeFig.16).

For the FPI, the approach is a bit more complex as the peak transmission of the filter is a less smooth function of the mir- ror gap length (and henceas a function of wavelength). The di- rectidentificationofspectralfeaturesrequiresthereforetheproper considerationof thefilterresponse.As on-ground calibrationwill haveprovidedaccurate estimationofthe FPIconstants, thephys- icalmodel ofthefilterwill beavailable. Preliminaryexperiments with the solar spectrum around 360nm havedemonstrated that the de-tuningerror can be detected with an uncertainty smaller than0.1nm.

3. Dataprocessing 3.1. Dataleveldeﬁnition

ThefollowingdeﬁnitionswillapplytotheALTIUSdatalevels.

(13)

Fig. 16. Solar spectral irradiance (SSI) as a function of wavelength. Data from the SOLSPEC instrument onboard ISS sampled at 0.1 nm (black curve). The red curve shows the result of a sweep of the vis channel around 655 nm.

• LR concerns raw telemetry data as received at the downlink station

• L0 data product: formatted data frames produced from LR, checked for integrity and ingested into the mission data base/archive.

• L1dataproduct:Detectorreadoutscalibratedandconvertedto geolocatedphysicalunits,afterprocessingofallcalibrationpro- tocols.L1dataprocessingishighlyautomatized butisveriﬁed withaqualityassurancecontrol.

• L2dataproduct: geophysicaldataproduced fromL1data,further processed to such a level that scientiﬁc analysis can be performed for individual geolocated observations with error budget.During andafterthe mission,the algorithms are usually improved and validated; L2 data are therefore released with a reference to the used version of the data processing model (DPM). DPM optimization and quality assurance of L2 dataneedhighlevelscientiﬁc expertiseinatmosphericchem- istryandphysics.L2dataarearchivedandreleasedtotheuser community.

• L3 data product: These are publishable science products, for which L2 data products are used as input. L3 data are pro- ducedbycombiningdifferentobservationsinordertoproduce griddedassimilateddata,climatologiesandtime series.Assim- ilated L3products willbe releasedto supportcross-validation andcoverage gapﬁllingthatcan,potentially,supporttime de- layedoperationalobjectives.

3.2. Processingchain

TheﬂowchartoftheprocessingchainisillustratedinFig.17.

3.2.1. FromL0toL1data

Although there are a number of common steps, fundamental differencesexistbetweenlimbandoccultationdatawhichdeserve speciﬁctreatment.Similarly,solaroccultationsandstellarorplane- taryoccultationshavetheirowncomplexityrelatedtothefactthat theSunisanextendedsourcewhereasstarsarepointsources.

The ﬁrstinput totheL1 processingchainisobviouslytheAL- TIUSL0data.Itconsistsofﬁlescontainingtherawspectralimages

acquiredduringone brightlimb observationor one entireoccul- tation.Therelevantauxiliarydata(withtime tags)such asspace- craft state vector andattitude,or payload housekeeping dataare extractedfromtheL0 ﬁle.Thesedata mustbe complementedby on-groundcharacterizations,in-ﬂightcalibrationsandthebestes- timateofthepolarizationellipseoftheincomingradiance.

TheL1dataprocessingstagethenproceedsthroughthefollow- ingfork:

• Forbrightlimbimages,in-ﬂightcalibrationdatacombinedwith S/Cattitudeknowledgeareusedtodeterminethespatialcoor- dinates of the apparent tangent point (TP) ofeach pixel line- of-sightabove thereferencegeoid.Digital radiances(andtheir uncertainties) are computed for each pixel by taking into ac- countthepolarizationstate,andareﬂaggedforthepresenceof clouds.

• Forstellar andsolaroccultations,theatmospheric slanttrans- mittanceisobtainedbydividingthestarspectralirradianceby themeasuredexo-atmosphericreference.

Itisworthpointingout asmalldifference,relatedtoLOSsam- pling,betweengrating basedspectrometersandspectral imagers.

The former record all useful wavelengths at once but at a ﬁxed tangent altitude. The latter capture all tangent altitudes at once butataﬁxed wavelengthforeach channelduringtheintegration time(althoughthe3ALTIUSchannelsoperateindependently).Both remotesensingtechniquesassumea locallyhomogeneousspheri- calatmosphereinordertoapply simpleinversiontechniques(not tomographicones).The sameremarkholds forlimbscatteringas foroccultationswheretheinstantaneoustangentpointmaymove by several degreesin latitude and/or longitude.This sampling is howeverconsistentwiththegeometricalproperty ofatmospheric remotesoundingfromaLEOsatellite.Mostofthegeophysicalin- formationarisesfromtheregionaroundthetangentpoint,whose effectiveopticallength isabout√

2

π

^RH≈500kmwhereRandH respectively stand for Earth radius and atmospheric scale height (≈7.5km). Thetypical durationofan ALTIUSlimbscatteringob- servationis10s whichcorresponds toabouta70kmalong-track displacementforaLEO orbit.Also, ifthenumberoflimbscatter- ingobservationsperorbitcouldbeincreasedupto100(insteadof 50) dependingon the ﬁnal platform performance, the possibility ofalong-tracktomographicretrievalsfromsuccessiveobservations, compatiblewiththevertical sensorresolution,willbeconsidered intheoperationaldataprocessing.

3.2.2. Radiativetransfer

The developmentof theALTIUS radiative transfer models isa longtermactivitydrivenbyatrade-off betweenprocessingspeed andaccuracy,forall consideredobservationmodes. Hereafter,we simplyoutlinethegeneralframe inwhichtherelevantalgorithms will be constructed, tested, optimized and ﬁnally integrated ina so-calledforwardmodel.

The densitydistribution ofthe atmospherictrace constituents is not directly observable. The remote sensing of an atmosphere is only made possible by measuring the electromagnetic energy outgoingfromtheatmosphere. The observedenergyis theresult ofabsorptionandscatteringprocesses,causedbytheatmospheric constituents,that modifythelimbradianceand/or bythermalra- diationgeneratedintheatmosphereitself.Todeterminetracecon- stituentdensitiesthusrequiressolvingthefollowinginverseprob- lem: givenan observed radiation ﬁeld, calculatethe atmospheric compositionthatproducedthisobservation.

It is the task of the retrieval algorithm to solve this inverse problem. This istypically done by minimizing the difference between the observed radiation and the radiation calculated by a model of the atmosphere, in an iterative loop by adjusting the trace constituent densities in theatmospheric model. The model

(14)

Fig. 17. ALTIUS data processing model for the data acquired in the nominal observation geometries. The L1 processor takes care of the signal correction, the geolocation of the measurements, and the data reduction. The L1c ﬁles are the end product of the L1 processor, ready for dissemination for further processing. The L2 processor is responsible for retrieving the atmospheric species concentration proﬁles, with the help of forward models or LUT.

(15)

Table 4

Stokes vector look-up tables.

Wavelengths (nm) 250, 275, 300, 350, 400, 500, 600, 750, 1000, 1400, 1800 Tangent heights (km) 0, 2.5, 5, 7.5, 10, 15, 20, 25, 30, 40, 50, 60, 75, 100 Solar zenith angles (deg.) 0, 15, 30, 45, 60, 75, 90

View zenith angles (deg.) 0, 15, 30, 45, 60, 75, 90, 105, 120, 135, 150, 165, 180 Relative azimuth angles (deg.) 0, 30, 60, 90, 120, 150, 180

Albedo 0.0, 0.3, 0.6, 0.9

that calculates the radiation,produced by an atmospherewith a givencomposition,iscalledtheforwardmodel.

Anacceptableforwardmodelmustbecapableofaccuratelyand eﬃcientlysimulatingthefollowingkeypropertiesandeffects:

1. Oblatespheroidgeometryoftheatmosphere 2. Polarizationofthelight

3. Atmosphericrefraction

4. Single(Rayleigh)scatteringbytheairmolecules(intheIR) 5. Groundalbedocontribution

6. Linearizedsinglescatteringmodelcalculation

7. OrbitalmovementsoftheEarth(w.r.t.the Sun)andALTIUS (w.r.t.Earth)

8. Presenceofmoleculeshavingspectracomposedofmanyin- dividualabsorptionlines(intheIR)

9. Multiplescatteringbytheairmolecules(intheUV–vis) 10. Linearizedmultiplescatteringmodelcalculation

The ALTIUS team is presently developing a radiative transfer model (RTM)that will addressall above-mentioned features. The performance of theRTM will beassessed withrespect to several existingRTcodes:SIRO-2.2[26],SCIATRAN-3.1[34]andSASKTRAN [46].Recently,outstandingperformanceshavebeendemonstrated by SMART-G[32] formassiveparallelizationthrough Monte-Carlo runson graphicalcards andare presently consideredfor adirect implementation.

In the future,we will consider two strategiesto compute the polarizationandmultiplescatteringeffectsintheradianceﬁeld:

1. Theclassicalapproach,whichistobuildasetofStokesvec- torlookuptables,producedforanensembleofatmospheric conﬁgurations and a grid of observational geometries, by running anaccurate (possiblyslow) tool, such asaMonte- Carlo program.Intheretrievalalgorithmloop,we canthen interpolatewithinthesetablestodeterminetheStokesvec- torfortheatmosphericconﬁgurationathand.Wewilltyp- ically need to tabulate over at least 6dimensions: albedo, wavelength, tangent height, solarzenith angle,view direction zenith angle and relative azimuth angle between the solarandviewdirection(seeTable4).

2. A more refined approach, which also starts by building a setoflookuptablesasinthe firststrategy,andisthenfol- lowedbyafittingstage,inwhichasmallsetoffitfunctions are derived that efficiently (w.r.t. speed and data requirements)canreproducethedatainthetables.Thesefunctions could e.g. representa multiplicative correction factorfunc- tion,withdomainintheabove6-dimensionalspaceandde- finedastheratioofthetotalscatteringdividedbythesingle scatteringcontribution(i.e.amultiplescatteringcorrection).

SuchacorrectionfactorcouldbeappliedtotheStokesvec- torthatisin-scatteringateachpointofthehalf-lineofsight.

For the trace gas retrieval, the RTM will represent the forward model,i.e.a singlescatteringalgorithmwithmultiple scat- teringcorrections. Interestingresultshavebeenobtainedrecently foraerosol[35] andtemperatureretrievals [14]inlimb geometry thatmaybeapplicableforALTIUSobservations.

In the infrared domain, we will consider the correlated-k method[18,21],atechniquethatisoftenusedtoreducethenum- ber of wavelength points for which monochromatic RT calcula- tions have to be done. The basic idea of the method is to sort the wavelengths according to the absorption coefficient at these wavelengths.Onthere-orderedwavelengths gridthespectrum is smoothandmonotonic,soveryfewquadraturepointsareneeded forthespectralconvolution.Theadvantageofthismethodarethat itis veryefficient andconceptually easy.Insteadof havingto do severalthousandsofmonochromaticRTcalculations,thiscannow bereducedtotypically10–40.Adisadvantageofthistechniqueis thattheexactsortingdependsonpressure,temperature,andtrace gasconcentration.Weintendtocombinethefineresolutionspec- traldata(computedinaline-by-lineapproach)andthecorrelated- kmethodtoproduceabroadbandmodel,foranumberofabsorp- tionlinecomplexesofanumberofchosentracegases(H₂O,CO₂, CH₄ andO₂). The outputof thisbroadband modelwill bein the form ofcross section tables, structured in such a way that they canserveasinputfilesfortheRTM.

Forbrightlimbretrievals,itisacommonpracticetoavoidthe need forabsoluteradiometric calibration by normalizing thesig- nalsfound atthetarget tangentaltitudes withrespect toarefer- encetaken at a higheraltitude. This renormalizationalso allows toremove an importantpart ofthegroundalbedo/cloudinessef- fectsto theextentthatall altitudesaresimilarlyaffected.The ef- ﬁciencyofthenormalization(f.i.withrespecttoareferenceradi- ance at40km)hasbeenestimatedby comparingthe relativedif- ference of normalizedradiances withand without albedo effects takenintoaccountintheRTmodel(seeFig.18).Above15km,the errormadebyneglectingthealbedo(A=0)whenusingnormalized radiancesisalwayssmallerthan6%,inthemostperturbativecase (A=1).Consideringtheerrorspectraldependence,afurthernor- malizationwithdataatarelativelyclosewavelengthwillvirtually suppressit.

ALTIUSmeasurementswillconsistofspectralimagescapturing a sceneof 100×100 km² atthe tangent pointwithout the need ofanyscanningmechanism.Thedetectorpixeldensityenablesan unmatchedspatialsamplingofthegeophysicalscene.Forinstance, the visible channel will provide independent measurements ev- ery200m, typically one orderof magnitudeﬁner than all previ- ousUV–vis–NIRlimb-scatter instrumentsormicrowaveemissions sounders.In addition, theabsence of scanningalso improves the knowledgeontheLOStangentheightwhichisacrucialparameter forconstituents exhibiting steep vertical concentration gradients.

Ontheother side,unlikegratingspectrometers,alimitednumber ofspectralmicro-windowshastobeselected.Thisisawell-known trade-off inspectral remote sensing: a full high-resolution spec- trumincreasesthe statistics (morespectral values) butdecreases the S/N ratio per spectral pixel while medium-resolution micro- windowsallow focus onspeciﬁc absorberswithhigherS/N ratio.

ForALTIUS,thenumberofusedwavelengthsisconditionedbythe sensitivityoftheinstrumentandthebrightnessofthelightsource, bothofwhichdrivethedetectorsexposuretime,andbythespec- tral information content of the target absorber optical thickness.

(16)

Fig. 18. An upper bound of the impact of unknown albedo (through multiple scattering on renormalized radiances) can be assessed by comparing the relative difference between two extreme albedo cases (A = 0 and A = 1).

However,thetunabilityoftheAOTFandFPIwillpreservethefree- domofchoiceforthemeasurementwavelengths.

The approach here will consist of selecting the best set of wavelengths that maximizes the sensitivity to a particular constituent.First, the measured radiances, L1-corrected for the relative response non-uniformity across the FOV, are normalized by a higher altitude measurement at the same wavelength, and ﬁ- nallyassembled indoublets or triplets(multiplets) involvingtwo or more wavelengths. These multiplets will be compared to the corresponding output ofa radiative transfer (RT)model andwill triggeran iterativeoptimization of the trace gases concentration proﬁles.

As thefull RTof ALTIUSis still underdevelopment,the scalar commercial code MODTRAN 5 [1] has been used to assess the achievablemission performance. The retrieval of verticalconcen- tration proﬁles frommultispectral imagesis astandard inversion problem, for which we followed the widely used Bayesian approach[33,41].Also, an along-tracksamplingof200kmdisplaces theLOSabovetheprevioustangentpointbyafewkilometersand allowsfortomographicretrievals[47].

For occultations,a ray tracing computationwill be performed basedonECMWF air/temperaturedatainorderto determinethe physicaltangentaltitudeoftheopticalpathandtherefractivedi- lutionfactor to be applied [11].The occultation SNR depends on theselectedlightsource.Forsolaroccultation,successiveacquired sunimages partially overlap.Each pixel of the Sunimage is as- sociatedwitha speciﬁc tangentaltitudecomputed byray tracing andaradialdistancetothecenteroftheunrefractedSunthatde- termines the solar limb darkening distribution. Overlapping tan- gentaltitudesarerecombinedintoacommongrid.Thesourceal- waysdisplaysthesamecharacteristics:awealthofavailablelight, aﬁnite angularextension, andaspectral output that hasa max- imum at visible wavelengths. Stellar occultations use point-like light sources withpropertiesthat are characterized by star mag- nitudeandtemperature.Retrievalqualitychangesforeachindivid- ualstar:brighthot stars (strongUV output)will deliveraccurate

stratospheric andmesosphericozone retrievals, whilebright cold starswillfavorizeretrievalsforspeciesthatarespectrallyactivein theinfra-red(CH₄,H₂O).

The inversion algorithm proceeds through standard steps i.e.

Rayleigh scattering removal, dilution and residual scintillation (stars only) removal, spectral inversion in global or differential modes, vertical inversionwith regularization ora prioriinforma- tion. The forwardmodel is lesscomplex incomparison withthe limbcase:itisbasedontheBeer–Lambertlawforopticalextinc- tion.

Due to their intrinsic capacity to measure on both day and night sides, limb emission instruments can perform observations attwice thedailysampledensityofsensorsobserving solarlimb scattering. Furthermore, they are still operational in polar night.

On theother hand, the globalcoverage ofALTIUS isobtained by itsmulti-modenature:thesolar,stellarandplanetaryoccultations provide sparse data at virtually all latitudes on the dark side of the orbit, while the bright limb measurements provide a regu- laranddense samplingonthe dayside. Therefore,ontop ofthe L1/L2 dataset (the data levels usually released tothe user community),we areconsidering theaddedvalue ofmergingthedata obtainedfromdifferentobservationmodesthroughanassimilation scheme to produce L3 data.In addition to merging the different type ofALTIUS observations, this also offers gridded data at the globalcoverage to allow forfast validationor other applications.

L3 dataproductionwill be performedwith theBelgianAssimila- tionSystemforChemicalObsErvations(BASCOE)whichisfocussed onthestratosphere[7,8,16,36,37].ThissystemisbasedonaChem- istryTransportModel(CTM)thatincludes58chemicalspeciesin- teractingthrougharound 200chemicalreactions andisdriven by meteorologicaldynamics(typicallyECWMF).Thissystemhasbeen usedinthepre-operationalprojectonMonitoringtheAtmosphere CompositionandClimate(MACC)todelivernearrealtimeanalysis ofMLSobservations[20].