GOES-R Activities at CIMSS / SSEC

• Trade Studies Home
  • Instrument
  • Retrieval Algorithm
• Accomplishments
  • FY05
  • FY04
  • FY03
  • FY02
• Simulations
  • Based on AIRS Level 1B
  • Based on NAST-I
  • Temporal resolution study
• HES Trade-off studies
• ABI/HES synergism
• IR sounder spectra with various spectral resolutions study
• Scan Scenario study
• Dectector Optical Ensquare Energy study
• Publications
• HES/ABI Team
• Related Links
  • CIMSS ABI
  • CIMSS Realtime GOES
  • CIMSS
  • NAST-I
  • AIRS
  • ASPT
  • Current Sounder Tutorial
  • 3rd GOES-R User Conference (NOAA/OSD)
  • 4th NOAA Hyperspectral Workshop

Comments, questions on this web page? Send e-mail to:
Chian-Yi Liu

Accomplishments in FY02

 
Fast forward model development and retrieval algorithm development
Retrieval trade-off studies with updated TRD noise
Initial ABS/HES algorithm validation using AIRS measurements
Impact of band-to-band co-registration error on sounding retrieval from HES
HES simulation study using CUBE data from MM5
Cloud property retrieval from HES cloudy radiances
Conclusions from FY02 accomplishment
References

The HES will be an infrared Michelson interferometer or grating Atmospheric Infrared Sounder (AIRS) like sounder. Figure 1 shows the spectral coverage of the current GOES sounder bands, 3-waveband ABS, 2-waveband HES (ABS prime in the figure) and 2-waveband GIFTS spectra. HES radiance measurements, at varying spectral resolutions, can be converted into atmospheric temperature, moisture and ozone profiles. HES spectral resolution will resolve individual carbon dioxide absorption lines and should provide the high vertical resolution soundings with the required accuracy.
 

Figure 1. The spectral coverage of the current filter-wheel radiometer (GOES-8 Sounder Bands, 10km spatial resolution) and a number of high-spectral resolution instruments, including the 3-band ABS/HES (10km spatial resolution), 2-band ABS/HES (10km spatial resolution, ABS prime in the figure) and 2-band GIFTS (4km spatial resolution). In general, the entire portion of the infrared spectrum covered by the current sounder at low spectral resolutions (15 to 100 cm-1) will be covered by the ABS/HES (with spectral resolutions between 0.625 and 2.5 cm-1).

Fast forward model development and retrieval algorithm development

Figure 2. Improved water vapor line parameters in CO2 region, LBL RTM calculation versus observation with HIS instrument.

An efficient algorithm for retrieving atmospheric temperature and moisture profiles from HES two waveband radiances has been updated with the new fast forward model. This algorithm regression procedure with all spectral channels followed by a nonlinear iterative physical approach using optimally selected channels.

To perform the regression retrieval, HES radiances are calculated from the radiosonde observations of the atmospheric state, and the regression coefficients are generated from these calculated radiances/atmospheric profile pairs. The radiative transfer calculation of the ABS/HES two waveband spectral radiances is performed using the fast transmittance model mentioned above; this model has 101 pressure level vertical coordinates from 0.05 to 1100 hPa. The calculations take into account the satellite zenith angle, absorption by well-mixed gases (including nitrogen, oxygen, and carbon dioxide), water vapor (including the water vapor continuum), and ozone. The temperature and moisture sounding retrievals are produced when the regression coefficients are applied to the actual clear sky HES measurements. The advantage of this approach is that it does not need HES radiances collocated in time and space with atmospheric profile data, it requires only historical profile observations. This statistical regression algorithm is computationally efficient, numerically stable, and simple.

In order to account for the nonlinearity of atmospheric parameter to the HES radiances, a nonlinear iterative physical retrieval procedure name regularization method along with the discrepancy principle has been developed (Li and Huang 1999; Li et al. 2000, Zhou et al. 2002) for hyperspectral infrared sounder data process. This physical retrieval procedure uses the regression result as a first guess for further profile modification.Fast and accurate computation of Jacobian is very important for real time sounding retrieval from HES data or near real time assimilation of HES radiances in Numerical Weather Prediction (NWP). Two fast ways (analytical form and transmittance ratio form, Li 1996) to compute the Jacobian for HES forward model are tested and implemented in the physical retrieval procedure. The Jacobian computation is radative transfer model (RTM) model independent in the IR region; the computation time is less than 2 brightness temperature (BT) calculations. Figure 3 shows water vapor mixing ratio weighting functions for HES 2-band short midwave (SMW) band (1650 - 2250 cm^-1) and the current GOES sounder, along with the HES SMW brightness temperature spectra. HES has much more vertical water vapor information in troposphere than current GOES sounder. Because HES has two or three bands that contain more than 1600 channels; it is not necessary to use all channels in the physical retrieval processing. A subset of channels (optimal channels) can be selected for the physical retrieval process, the reasons for HES channel selection are: (a) There are information redundancies among all the HES channels; the other channels can represent some channels? information; (b) The weighting function and inverse computation is very time consuming in each iteration by use of all channels, a set of optimal channels will significantly reduce the retrieval processing time and meet the real time data processing requirement. Initial channel selection has been tested in the physical retrieval procedure.
 

Figure 3. water vapor mixing ratio weighting functions (upper left and upper right panels) for ABS/HES SMW band (1650 - 2250 cm**-1). The lower left panel is the current GOES sounder water vapor weighting functions, the lower right panel shows the ABS/HES SMW brightness temperature spectra. The U.S. standard atmosphere is used in the calculation.

Figure 4, 1km vertical temperature retrieval RMSE (left panel) and 2km vertical water vapor RH retrieval RMSE (right panel) from regression and physical retrieval procedures, the physical retrieval uses the regression result as a first guess. ABS/HES 2-band TRD noise is used in this simulation.

Retrieval trade-off studies with updated TRD noise

Figure 5, the simulated 1km temperature retrieval RMSE from ABS/HES longwave 650cm-1 cutoff and 685 cm-1 cutoff with the same TRD instrument noise, along with simulated temperature retrieval RMSE from current GOES sounder radiances. 463 global independent profiles are included in the statistics.

Figure 5 shows the simulated 1km temperature retrieval RMSE from HES longwave 650cm^-1 cutoff and 685cm^-1 cutoff with the same TRD instrument noise, along with the simulated temperature retrieval RMSE from the current GOES sounder radiances. About 0.2 K improvement for upper tropospheric temperature is obtained with 650cm^-1 cutoff over the 685cm^-1 cutoff which GIFTS LW band starts.

Figure 6 shows the 1km vertical temperature retrieval RMSE (left panel) and 2km vertical water vapor (RH) retrieval RMSE (right panel) from HES LW (650 - 1200 cm^-1) only, SMW (1650 - 2250 cm^-1) only, LW + SMW, and current GOES sounder. It can be seen that temperature information is mostly provided by LW only. For moisture information, LW provides very useful boundary layer information while SMW provides most water vapor information above 700 hPa, LW + SMW gives the best moisture vertical information. HES two band TRD noise is added in the ABS/HES simulation. Figure 7 shows 1km vertical temperature retrieval RMSE (left panel) and 2km vertical water vapor (RH) retrieval RMSE (right panel) from HES LW+ SMW with noise factor equal to 0.5 (half noise), 1.0 (TRD noise) and 2.0 (double noise). It can be seen that the retrieval error is almost linearly amplified by increased instrument noise.This demonstrates that pixels with good noise performance will have good sounding retrieval.
 

Figure 6, the 1km vertical temperature retrieval RMSE (left panel) and 2km vertical water vapor (RH) retrieval RMSE (right panel) from HES LW only, SMW only, LW + SMW, and current GOES sounder. 463 global independent profiles are included in the retrieval statistics; TRD noise is used in the simulation.

Figure 7, the 1km vertical temperature retrieval RMSE (left panel) and 2km vertical water vapor (RH) retrieval RMSE (right panel) from HES LW+ SMW with noise factor equal to 0.5 (half noise), 1.0 (TRD noise) and 2.0 (double noise).

Initial ABS/HES algorithm validation using AIRS measurements

Impact of band-to-band co-registration error on sounding retrieval from HES

Figure 8, HES TRD noise and the RMS errors with 10%, 20% and 50% LW band-to-band mis-registration.

The upper panel of Figure 8 shows HES two wavebands' SMW TRD noise and the errors with 10%, 20% and 50% band-band mis-alignment, all the noises are scaled to NeDT@250K. The lower panel shows the brightness temperature spectra calculated from standard atmosphere.
 

Figure 9, HES TRD noise and RMS errors with 10%, 20% and 50% SMW band-to-band mis-registration.

Figure 9 is the same as Figure 8 but for SMW band.
 

Figure 10, the 1km vertical temperature and 2km RH retrieval RMSE from ABS/HES two waveband simulation with 0% (no mis-alignment error), 10%, 20% and 50% band-band mis-alignment errors from LW band only.

Figure 11, the 1km vertical temperature and 2km RH retrieval RMSE from HES two waveband simulation with 0% (no mis-alignment error), 10%, 20% and 50% band-band mis-alignment errors from SMW band only.

Figure 10 shows the 1km vertical temperature and 2km RH retrieval RMSE from HES two-band simulation with 0% (no mis-alignment error), 10%, 20% and 50% band-band mis-alignment errors from LW band only. It can be seen that LW mis-alignment error will affect temperature and boundary layer moisture retrievals.

Figure 11 shows the 1km vertical temperature and 2km RH retrieval RMSE from HES 2-band simulation with 0% (no mis-alignment error), 10%, 20% and 50% band-band mis-alignment errors from SMW band only. It can be seen that SMW mis-alignment error will only slightly affect temperature and boundary layer moisture retrievals.

HES simulation study using CUBE data from MM5

Figure 12. Simulated HES clear 700 hPa water vapor mixing ratio retrieval image (upper panel), true image (middle panel) and the difference between retrieval and true (lower panel) at time 1.

Figure 12 shows simulated HES clear 700hPa water vapor mixing ratio retrieval image (upper panel), true image (middle panel) and the difference between retrieval and true (lower panel) at time 1. Figure 13 is the same as Figure 12 but for time 2. The HES TRD noise is added in the simulated radiances, only regression approach is used in this cube study.
 

Figure 13, simulated HES clear 700hPa water vapor mixing ratio retrieval image (upper panel), true image (middle panel) and the difference between retrieval and true (lower panel) at time 2.

Cloud property retrieval from HES cloudy radiances

Figure 14, CTP sensitivity spectra with 200, 300, 500, 700 and 850 hPa CTPs, four atmospheric conditions are used in the calculations (see the 4 panels).

Figure 15, simulated HES and current GOES sounder CTP retrieval RMSE as a function of ECA.

To compare the HES and the current GOES sounder cloud parameter retrievals, a simulation study was conducted. A set of 75 CONUS radiosonde profiles, that represent different atmospheric conditions. Twenty combinations were formed from each profile by assigning two CTPs plus 50 hPa random variation (200 hPa and 300 hPa corresponding to very high-, high-level clouds) and 10 ECAs (0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 and 1.0). The HES and GOES-8 longwave spectral band cloudy radiances were simulated for all combinations of each profile. An infrared surface emissivity of 0.98 for each longwave spectral band was assumed in the simulation. The HES TRD noise and GOES-8 instrument noise were added into cloudy radiance. Cloud retrievals from both ABS/HES with local minimum emissivity variance (LMEV, Huang et al. 2002) and GOES-8 sounder with 1DVAR (Li et al. 2001) are compared with true CTPs for RMSE calculations. In order to account for the surface uncertainties and the atmospheric profile error in the LMEV and 1DVAR cloud retrievals, nominal errors were considered in the simulation. For atmospheric temperature, a 1.5 K random error was assumed at each pressure level, which is close to the accuracy of the forecast analysis. For surface skin temperature a nominal random error of 2.5 K was assumed in the simulation. In addition, 1.0% error was included for infrared (IR) surface emissivity and 15% error was included for water vapor mixing ratio at each pressure level.

Figure 15 shows the simulated HES and current GOES sounder CTP retrieval RMSE as a function of effective cloud amount (ECA) or effective cloud emissivity. For thin clouds, high spectral is very important for more accurate clouds.

Conclusions from FY02 accomplishment

• ABS/HES has much better spatial coverage, vertical resolution for sounding and accuracy for products than current GOES sounder.
• The TRD noise meets the temperature and moisture retrieval accuracy (1K for temperature and 10% for moisture).
• LW band co-registration error has impact on both boundary layer temperature and moisture retrieval, 10% co-registration error will double the system noise in longwave window region; LW band co-registration error has less impact on the upper tropospheric temperature and moisture retrieval, however, surface skin temperature and emissivity retrieval is expected to be impacted significantly by the LW band co-registration error.
• SMW band co-registration error of less than 50% won't create significantly additional error to both temperature and moisture retrieval in this particular case.
• Cube study shows that both spatial and temporal features of the atmosphere can be detected from the ABS/HES retrievals.
• ABS/HES provides much better CTP retrieval accuracy over the current GOES sounder, especially for thin high clouds.
• About 0.2 K improvement for upper tropospheric temperature is obtained with 650cm^-1 cutoff over the 685cm^-1 cutoff which GIFTS LW band starts.
• Temperature information is mostly provided by LW (650 - 1200 cm^-1) only. For moisture information, LW provides very useful boundary layer information while SMW (1650 - 2250 cm^-1) provides most water vapor information above 700 hPa, LW + SMW gives the best moisture vertical information.
• The retrieval error is almost linearly amplified by increased instrument noise, pixels with good noise performance will have good sounding retrieval accuracy. Double of TRD noise won't create unacceptable sounding retrieval.

References

Gurka J. J., and G. J. Dittberner, 2001: The next generation GOES instruments: status and potential impact?. Preprint Volume. 5^th Symposium on Integrated Observing Systems. 14-18 January, 2001, Albuquerque, NM., Amer. Meteor. Soc., Boston.

Li, J., 1996: Infrared remote sensing of atmosphere and its inversion problem studies, Ph.D. dissertation, Institute of Atmospheric Physics, Chinese Academy of Science, Beijing.

Li, J., T. J. Schmit, F. Sun, and W. P. Menzel, 2003: Surface and atmospheric retrievals from future GOES sounder - Advanced Baseline Sounder (ABS), Proceeding, 12^th Satellite and Oceanography Conference,Feb. 9 - 13, 2003, Long Beach, American Meteorological Society.

Li, J., T. J. Schmit, and W. P. Menzel, 2002: Advanced Baseline Ssounder (ABS) for future Geostationary Operational Environmental Satellites (GOES-R and beyond), SPIE proceeding, Applications with Weather Satellites, Oct.23 - 27, 2002, Hangzhou, China.

Li, J., W. P. Menzel, Z. Yang, R. A. Frey, and S. A. Ackerman, 2002: High spatial resolution surface and cloud type classification from MODIS multi-spectral band measurements, J. Appl. Meteorol.,(Accepted)

Li, J. and H.-L. Huang, 1999: Retrieval of atmospheric profiles from satellite sounder measurements by use of the discrepancy principle.? Appl. Optics. 38: 916-923.

Li, J., and S. Huang, 2001: Application of improved discrepancy principle in inversion of atmosphere infrared remote sensing, Science in China (series D), 40, 847 - 857.

Li, J., Menzel, W. P., and A. J. Schreiner, 2001: Variational retrieval of cloud parameters from GOES sounder longwave cloudy radiance measurements. J. Appl. Meteorol. 40, 312-330.

Li, J., T. J. Schmit, H. L. Huang, and H. M. woolf, 2001: Retrieval of atmospheric inversions using geostationary high-spectral-resolution sounder radiance information. 11^th Conference on Satellite Meteorology and Oceanography, 15 - 18 October 2001, Madison, WI.

Li, J., C. C. Schmidt, J. P. Nelson, T. J. Schmit, and W. P. Menzel, 2001: Estimation of total ozone from GOES sounder radiances with high temporal resolution. Journal of Atmospheric and Oceanic Technology. 157 - 168.

Li, J., W. P. Menzel, Z. Yang, R. Frey, and S. Ackerman, 2002: High spacial resolution surface and cloud type classification from MODIS multi-spectral band measurements, J. Appl. Meteor.: (in press).

Li, J., S. W. Seemann, W. P. Menzel, and L. Gumley, 2002. Operational retrieval of atmospheric temperature, moisture, and ozone from MODIS, J. Appl. Meteor.: (accepted).

Li, J., W. Wolf, W. P. Menzel, W. Zhang, H. Haung, and T. Achtor, 2000: Global soundings of the atmosphere from ATOVS measurements: The algorithm and validation, J. Appl. Meteor.39: 1248-1268.

Ma, X. L., T. J. Schmit, and W. L. Smith, 1999: A non-linear physical retrieval algorithm-its application to the GOES-8/9 sounder, J. Appl. Meteor.38: 501-513.

Plokhenko, Youri and W. Paul Menzel, 2002: Mathematical aspects in meteorological processing of infrared spectral measurements from the GOES Sounder. III. Emissivity estimation in solving the inverse problem of atmospheric remote sensing. J. Appl. Meteor ( in press)

Schmit, T. J., J. Li, T. J. Schmit, and W. P. Menzel, 2002: Advanced Baseline Imager (ABI) for future Geostationary Operational Environmental Satellites (GOES-R and beyond), SPIE proceeding, Applications with Weather Satellites, Oct. 23 - 27, 2002, Hangzhou, China.

Zhou, D. K., W. L. Smith, J. Li, G. W. Cantwell, A. M. Larar, J.-J. Tsou, and S. Mango, 2002: Geophysical product retrieval methodology for NAST-I and validation, Applied Optics (in press)

Zhou, D. K., W. L. Smith, A. M. Larar, M. A. Avery, J. Li, X. Liu, J.-L. Moncet, and N. S. Pougatchev, "NAST-I remote sensing and carbon monoxide", in Proceedings, SPIE Optical Remote Sensing of the Atmosphere and Clouds III, A. H-L. Huang, D. Lu, and Y. Sasano, Eds., 4891, in print, Oct. 23 - 27, 2002, Hangzhou, China.