MicroCarb Mission: A new space instrumental concept based on dispersive components for the measurement of CO2 concentration in the atmosphere

Transcription

1 International Conference on Space Optics 2012 MicroCarb Mission: A new space instrumental concept based on dispersive components for the measurement of CO2 concentration in the atmosphere Véronique PASCAL (CNES), Veronique.pascal@cnes.fr Christian BUIL, Elodie CANSOT, Jacques LOESEL, Laurie TAUZIEDE,Clémence PIERANGELO, Francois BERMUDO (CNES) Mathieu OLIVIER (ALTEN) Mickael DUBREUIL (Sophia Conseil)

2 MicroCarb science objectives and requirements OUTLINE MicroCarb Status Dispersive Spectrometer description Main performance requirement Zoom on non-linearity performance Zoom on dispersive component and polarisation performance Instrument simulator 2

3 MicroCarb Science Objectives Mission requirements are driven by the need to better constrain natural CO2 fluxes at the Earth surface through data assimilation. Carbon dioxide (CO2) is one of the most important Greenhouse Gases (GHGs) which are driving global climate change. MicroCarb will measure vertically integrated CO2 concentration to quantify CO2 surface fluxes at regional scales to identify and monitor global carbon sources and sinks In order to better quantify the CO2 fluxes at the surface, very high quality of CO2 concentration measurements are necessary Priority is given to precision on measurement (in ppm) rather than high spatial resolution or sampling 3

4 MicroCarb Mission Requirements The CO2 concentration will be retrieved by measurements of the absorption of reflected sunlight by CO2 at near infrared. The payload consists in a passive instrument. measurements of the absorption of reflected sunlight by CO2 Myriade Evolution platform with Myriade Flight Operation Center design shall be used. Mission design shall be based on technology with moderate development schedule and risks : a compact and low cost concept mission. Low cost launch providers shall be preferred (e.g. secondary passenger launch) 4

5 MicroCarb Mission Requirements The level 1 requirements are written such that: The goal gives the same level 2 performance as OCO The threshold is such that the level 2 performance is relaxed by 35% Spectral bands: measurement in SWIR CO2 and O2 Band 0,76µm O2 A-Band : Surface pressure, clouds/aerosols 1,61 µm CO2 band : Column CO2 2,06 µm CO2 band : Column CO2, clouds/aerosols reflectance reflectance O µm CO2-1.6 µm CO2-2 µm reflectance (cm-1) 5

6 MicroCarb status Phase A1 performed from January to September 2011 with trade off between 2 different instrument concepts ( both by the 2 chosen manufacturers TAS and ASTRIUM) : Static interferometer concept based on the phenomena of interference between two radiation waves. Grating spectrometer concept based on the use of dispersive optical components Selection of grating spectrometer instrumental concept based on comparative analysis covering different criteria (feasibility, performance level 1 to 4, scheduling and cost) Phase A2 performed from September 2011 to February 2012 with consolidated performance and satellite budgets Phase A3 performed from June 2012 to March 2013 with technological predevelopment, characterization of critical devices and specific study for mission performances The MicroCarb phase A will be concluded by mid 2013 with a Preliminary 6 Requirements Review (PRR)

7 Dispersive spectrometer principle Radiometric calibration Scan Spectral calibration Polarisation scrambler D - Entrance telescope Spectrometer Detection Power resolution (/d) is given by : k m d1 R D cos or R = typically 2 d1 tan R D With : = angular width of slit (radians) D = telescope diameter d1 = internal spectrometer pupil (linked to spectrometer size) k= diffraction order m = groove density A = incidence angle on the grating 7

8 Dispersive spectrometer principle Satellite speed 2D spectrum Each column is a monochromatic image of the slit Length of the slit (swath) Dispersion Spatial axis FOV 1 FOV 2 FOV 3 Nbin Nbin Nbin Width of the slit Spectral axis Binning is vital for distortion corrections, ILS stability 8

9 Main performance requirements Signal to Noise Ratio (SNR), spectral resolution (R) and bandwidth (BW) 9 They are the instrument driving parameter for CO2 retrieval accuracy As very different combinations of these parameters might give similarly good level 2 performances, we want to give such a freedom to the industry => trade-off based on instrument considerations for an optimal configuration Parametric relationship Calculated through linear error estimates for a clear scene (no aerosol) Search for the optimal values for α, β and γ on a set of ~50 instrument configurations k is fixed so that p=required performance in ppm Possibility to include the number of FOV across track (N) and along track (M) (ppm) L < 100 p BW p BW Linear error estimate k SNR R N M k SNR R

10 Main performance requirements Min and max values of BW, SNR, R, M, N are specified, together with inter-band variations Spectral bands B1 B2 B3 Bandwidth BW cm cm-1 SNR per FWHM Resolving power R M > 4 along 50 km track trace N Between 1 to 5 Footprint on the ground of the sounding elementary pixel between 9 km2 and 50 km2 Beyond radiometric noise, instrument defects will affect the mesured spectrum and the precision of the estimated CO2 concentration. Effect similar to radiometric noise «pseudo noise» Different kind of pseudo noise:» Geometric defects» Spectral defects (ILS knowledge, keystone, smile..)» Polarisation defects» Calibration residue» Non linearity and non uniformity defects 10

11 Non linearity pseudo noise : Zoom on non-linearity performance Main contributor is detector Non linearity affects the radiometric performances To reduce development and qualification costs: use detector off the shelf Potential solution identified: HgCdTe detector from Sofradir for B2 and B3 Bands 1000 x 256 or 500 x 256 pixels with a 30 µm pitch 2,2µm cutt off (customized in order to reduce dark current) Capacitive TransImpedance Amplifier (CTIA) electrons capacity handling Existing data concerning detector response for signal under 10% full well shows : Pixel reponse dispersion increases Non linearity shape is different from one pixel to another Accuracy of measurement is limited New measurements scheduled with better accuracy at low signal level Illustration of non linearity shape for 5 different pixels at low signal 11

12 Zoom on dispersive component and polarisation performance Main component of the spectrometer studied at CNES is based on echelle grating To assess the spectral requirements, we choose: High incidence angle High diffraction order Low groove density Operated in near littrow conditions The diffraction orders are separated by a set of dichroic filters and spectral bandpass filters centred on the 3 bands Spectral bands B1 B2 B3 Diffraction order size Groove density 48 mm along the lines and 140mm perpendicularly to the lines lines/mm Blaze angle

13 Zoom on dispersive component and polarisation performance 3 possible strategies for polarisation Polarisation measurement (2 components, ex: GOSAT) One polarised component measurement (ex: OCO-2) Scrambling the polarisation choice for MicroCarb» Less complex and expensive than polarisation measurement» Less sensitive to aerosol and cirrus scattering than one polarised component measurement MicroCarb polarisation requirement : Glint mode: Tp<0.1% (G), 0.7% (T) Nadir mode: Tp<0.25% (G), 5% (T) With these values the pseudo noise level is below 1/1000 The main contributor of polarisation instrument budget is the grating. Special care is taken to ensure good grating polarisation performance 13

14 Zoom on dispersive component and polarisation performance Modelling polarisation Development of model with Institut Fresnel Grating model to predict grating performances Comparison with real measurements on breadboard Simulation using real groove profile 31.6 grooves/mm, blaze angle 71.45, diffraction order Measuring efficiency and polarisation of grating with CNES breadboard Measurement of Richardson grating (ref 53_453 E ) in B1 and B2 bands Collimated beam in quasi littrow with a constant angular deviation of 0.6 for measurement with echelle grating and measurement with reference mirror efficacité (%) Measured efficiency of Richardson echelle Grating ref 53_*_453E, 31.6 grooves/mm, Blaze = 71.45, diffraction order 78, angular deviation 0.6, unpolarized light longueur d'onde (nm)

15 Instrument Simulator Model of dispersive spectrometer used to obtain signals measured on detector Use sun spectrum, atmospheric spectrum or specific sources (black body, gas cell, laser ) Add instrumental defects from optics, detector, electronic model of level 0 (non calibrated raw data) and level 1 (calibrated spectral and radiometric data) Model usefulness in establishing and verifying the validity of the instrument specification and concept Model of breadboard : Estimation of misalignment, distortion of breadboard optics After correction of the distortion, modelisation of the laser shape from measurements with laser line spectrum at 770nm 15

16 Conclusion To fulfill the objective performances, the pseudo noise must be reduced in particular non linearity defects and the polarisation defect The detector is the main contributor of non linearity performance. Using the existing HgCdTe detectors above 10% of charge handling capacity can order to guaranty linearity The main contributor of polarisation instrument budget is the grating. Special care is taken to ensure good grating polarisation performance. The integration of a scrambler in the design is necessary. Linking the instrument simulator with the breadboard is very interesting to validate the concept performances. The compact design approach combined with the Myriade Evolution product line will offer a low cost solution for the measurement of CO2 concentration in the atmosphere. More information can be found on the CNES website 16