1. Quick Review of Remote Sensing Basic Theory Paolo Antonelli CIMSS University of Wisconsin-Madison South Africa, April 2006

2. Outline IR at High Spectral Resolution Basic Principles Limits in IR remote sensing

3. Visible (Reflective Bands) Infrared (Emissive Bands) High Spectral Resolution

4. High Vertical Resolution

5. Radiative Transfer Equation

6. Transmittance for an off-line Channels z  1  close to 1 a close to 0 z1z1 z2z2 z3z3  + a + r = 1 The molecular species in the atmosphere are not very active: most of the photons emitted by the surface make it to the Satellite if a is close to 0 in the atmosphere then  is close to 0, not much contribution from the atmospheric layers Surface emission Atmospheric emission T(z 1 ) T(z 2 ) T(z 3 )

7. Trasmittance on an Absorption Line z  1 z1z1 z2z2 z3z3 Absorption Channel:  close to 0 a close to 1 One or more molecular species in the atmosphere is/are very active: most of the photons emitted by the surface will not make it to the Satellite (they will be absorbed) if a is close to 1 in the atmosphere then  is close to 1, most of the observed energy comes from one or more of the uppermost atmospheric layers z4z4 T(z 4 ) T(z 3 ) T(z 2 ) T(z 1 )

8. Weighting Functions  1 z1z1 z2z2 zNzN d  /dz z1z1 z2z2 zNzN

9. What Causes Absorption? Molecules in the Atmosphere. For any layer of the atmosphere, molecular absorption determines the layer emissivity and trasmittivity

10. CO 2 Lines

11. H 2 O Lines

12. Vibrational Lines CO 2 O3O3 H2OH2O

13. Rotational Lines CO 2 O3O3 H2OH2O

14. HeIghtHeIght AbsorptionAbsorption Wavenumber Energy Contribution Weighting Functions

15. Line Shape: Pressure Dependence

16. Broad Band window CO 2 O3O3 H2OH2O window Samplig of vibrational bands Integration over rotational bands

17. … in Brightness Temperature

18. High Spectral Resolution Samplig over rotational bands

19. AIRS and MODIS (mt Etna, Sicily, 28 Oct 2002)

20. Broad Band vs High Spectral

21. Questions For a given water vapor line what happens to its weighting function When the amount of upper tropospheric water vapor increases?

22. Tau 100% 0 For a given water vapor spectral channel Wet Atm. Dry Atm. Dry Wet Moderate Weighting Function HEIGHTHEIGHT

23. Questions How does it look the observed spectrum for high thick water vapor cloud ?

24. Thick Cloud Opacity

25. Questions Moving deeper and deeper into an absorption line the observed BT tends do decrease, why? Is it always true that the BT decreases going deeper into the absorption band?

26. Temperature Inversions AIRS (20 July 2002) calculation Sub-Arctic Winter May 2001 S. Pole radiosonde profile used in calculation (0.365 mm H 2 O)

27. Conclusions High Spectral Resolution High Vertical Resolution Large Volume of Data (redundancy)

28. High Spectral Resolution: Products Sounding (T,WV) at high vertical resolution Surface Properties (Emissivity, T) Cloud properties (Top Pressure, Optical Thickness, Effective Emissivity) Wind Profiling Noise Filtered Observations Higher Spectral Resolution Higher vertical resolution retrieved variables

29. Airborne Instruments MAS: moderate spectral resolution, high spatial resolution NAST-I, S-HIS: high spectral resolution, lower spatial resolution

30. NAST on Proteus ER-2 Airborne FTS Instruments

31. NAST/FIRSC Payload on Proteus NPOESS Atmospheric Sounder Testbed Interferometer (NAST-I) Microwave (NAST-M) Far-Infrared Sensor for Cirrus (FIRSC) NAST-M NAST-I FIRSC

32. Scanning-HIS Aircraft Instrument

33. p Quicklooks

34. CrIS CO 2 O3O3 H2OH2O N 2 O CO 2 CO N 2 O CH 4 H2OH2O GIFTS-SWGIFTS-LW H2OH2O CO 2 O3O3 N2ON2O CH 4 CO N2ON2O CO 2 H2OH2O H2OH2O CrIS/AIRS NAST-I /IASI ( 0.25 cm -1 /0.5 cm -1 ) GIFTS (0.6 cm -1 ) 0.6 cm -1 1.2 cm -1 2.4 cm -1 Low Troposphere H 2 O Hole! Spectral Coverage

35. NAST-I Spm RTVL Chnls CFCs N 2 O/CH 4 Low S/N Band Edge Noise Regression Retrieval: Channel Selection

36. NOV 30 DEC 05 Water Vapor and Temperature Retrievals

37. Altitude (km) Relative Humidity (%) Distance (75 km) 3km NAST Raob Andros Is. Bahamas, Sept 12,1998

38. H 2 O Sensitivity [dT B (,p)/dlnq(p)] Water Vapor Sensitivity

39. 75 km LW-side of H 2 O Band SW-side of H 2 O Band Water Vapor Sensitivity

40. This flight demonstrates the ability to observe the spatial moisture structure below a scattered and semi- transparent Cirrus cloud cover 14.9 13.8 16.0 UTC Sounding in presence of thin clouds

41. THORPEX 2003 THORPEX objectives are: to improve the use of various in situ and space-based observation systems in operational weather predictions to evaluate model sensitivity studies and test the impact of targeted observations to begin to develop intelligent observing systems along with forecast models capable of dynamical interaction through data assimilation systems

42. Soundings: 03 Mar 2003 23:49:0000:03:00

43. Temperature Sounding Validation

44. Temperature Sounding Validation: Dropsondes

45. Temperature Sounding Comparison: S-HIS / NAST-I

46. Relative Humidity Validation

47. Sounding: 22 Feb 2003

48. S-HIS sounding

49. Relative Humidity Validation

50. Noise Reduction High Spectral Resolution Observations are highly redundant Reducing the instrument noise without compromising the spatial and spectral resolutions is very expensive

51. Noise Reduction High Spectral Resolution means high redundancy Reducing the instrument noise without compromising the spatial and spectral resolutions is very expensive We can take full advantage of the high redundancy, to filter out part of the instrument noise PCA has been chosen to build a noise filter that takes advantage of the high redundancy

52. Noise Reduction Single Recons. Spectrum Averaged Spectrum Before Compression After Compression Uncorrelated Noise

53. Noise Filter Validation: AFWEX

54. S-HIS: Noise Filter

55. AIRS before Aqua integration, Aqua shake test & simulated AIRS spectrum Wavenumber (cm -1 )

56. AIRS: High Spectral Resolution, Global! From hot to cold 20-July-2002 Ascending LW_Window

57. To Coldest … Dome Concordia 20 July 2002 Obs-Calc AIRS (20 July 2002) calculation Sub-Arctic Winter May 2001 S. Pole radiosonde profile used in calculation (0.365 mm H 2 O)

58. Combining AIRS and MODIS for cloud mask Not Determined CloudyUncertain Probably Clear Confident Clear # confident clear # total = 1 # cloudy # total = 1

59. Not Cloudy Uncertain Probably Confident Determined Clear Clear MODIS cloud mask 20-July-2002 cMODIS fractions cloudyuncertain probably clearconfident clear AIRS cloud flag. All confident clear clear not clear AIRS clear flag from MODIS cloud mask

60. GIFTS The Next Major Advance in Observing from Geo 16,000 Temperature, Humidity & Trace Gas Profiles in 10 sec Global Sounding in < 10 min High resolution Sounding of 6000 x 6000 km in < 30 min Dense Wind Observations, tracked from Water Vapor Sdgs For:Hurricane Landfall Tornadic Storms Nowcasting Numerical Prediction Air Quality Forecasts

61. Noise Characterization using PNF

62. Granule 016 channels 572, 573 AB_State = [0,0], Rad_Qual = [0,0], Bad_Flag = [0,0] AIRS: Popping Noise

63. Original Filtered Channel 573 - Channel 572 Brightness Temperature Differences AIRS: PCA Noise Filter

64. Main Challenges of High Spectral Resolution Data volume –data transmission –data storage –data processing (calibration, inversion)

65. Data Volume The increased volume of data transmitted from the satellite to the ground, if not properly processed, will: –exceed the capacity of the current operational downlink technology –require expensive data systems to process the data on the ground

66. Data Volume: Transmission Rate GOES Data Rate GIFTS  GOES*30

67. GIFTS Data Volume To download GIFTS data from home you would need 1000 modems This CD can contain less than 1.5 minutes of GIFTS data …24 hours data of GIFTS data would fill 925 CDs