Stratospheric aerosol profile retrieval from SCIAMACHY limb observations Yang Jingmei Zong Xuemei Key Laboratory of Middle Atmosphere and Global Environment Observation (LAGEO), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, 129, China Introduction Data and Methodology Sensitivity Studies and Comparisons Conclusions
Introduction A long term measurement of stratospheric aerosol distributions is necessary for a better understanding of the stratospheric processes. Before the middle s, the satellite aerosol profile measurements were mainly carried out by the instruments using the solar occultation technique. 4 1 5-6 o N 4-5 o N -4 o N - o N 1- o N -1 o N high accuracy and high vertical resolution poor geographical coverage. The new-generation of satellite instruments perform limb measurements of the scattered solar radiation in the Earth s atmosphere. huge information of atmospheric trace gases and aerosols 1E-6 1E-5 1E-4 1E-3 Aerosol Extinction Coefficient (km -1 ) Annual averaged zonal mean SAGE II 1 nm aerosol extinction profiles over 1998 4. The observed heights of the tropopause are indicated by the vertical bars (one standard deviation. The aim of this study: Retrieve aerosol extinction profiles from SCIAMACHY Limb measurements
DATA AND METHODOLOGY SCIAMACHY on board ENVISAT: Channels: 8 Spectral range: 24 238 nm Spectral resolution:.2.5 nm Tangent heights: -3 1 km Vertical intervals: 3.3 km We use level 1 version 7.4 limb observation data SCIAMACHY limb radiance spectra at 12.,.2, 18.6, 21.8,.1, 28.4, 31.7 and. km tangent height on 21 January 4, when the tangent point passed over western China (36. N, 97.3 E). Limb radiance (1 13 photons s -1 cm -2 nm -1 sr -1 ) 6 5 4 3 2 1 4 5 6 7 8 9 1 Wavelength (nm) 12. km.2 km 18.6 km 21.8 km.1 km 28.4 km 31.7 km. km (Institute of Remote Sensing, University of Bremen). We use SCIATRAN for Limb scatter radiance simulation and weighting function calculations. Sensitivity of the measured radiances at 75 nm to variations of the aerosol extinction coefficients for the above observing conditions. 4 12. km.2 km 18.6 km 21.8 km.1 km 28.4 km 31.7 km. km 1..2.4.6.8.1.12.14 dln(i)/dln(β a )
The equation to be solved for the retrieval is a nonlinear inverse problem : y = F( x) + e y (1) F: the non-linear forward model; y: measurement vector; X: the state vector; e y :measurement errors of all kinds. In order to reduce the sensitivity to the assumed surface albedo and to minimize the impact of instrument calibration errors, the limb radiance at each tangent height is normalized to the radiance at a reference tangent height: To reduce the effect of the Rayleigh scattering, the ratio of two spectral channels is used in the retrieval. We choose the wavelength pare of λ l = 75nm and λ s = 47nm and construct the measurement vector y as : Retrieval Method In ( λ, THi ) = I( λ, TH I( λ, TH ref I (λ, TH i ): the limb radiance at wavelength λ and tangent height TH i, TH ref :reference tangent height ( km). The retrieval algorithm is established on the basis of the optimal estimation method described by Rodgers[4]. The calculation is performed in an iterative manner by employing the equation: i ) ) (2) I n( λl,thi ) y = ln( ) (3) I n( λs,thi ) I n (λ l,th) and I n (λs,th): the normalized measured radiance at the wavelengths of λ l and λs. λ l = 75 nm and λs =47 nm. x = x 1 T 1 1 T 1 + ( S + K S K ) K S [( y y ) K ( x x )] n + 1 n y n n y n n n (4) S is the a priori covariance matrix of the solution; S y is the error measurement covariace matrix, K n is the matrix of weighting functions, x is the a priori aerosol profile, and y n is the calculated y value. Subscripts n and n+1 denote the number of the iteration.
Flow Chart for Retrieval Process Input SCIAMACHY Observed Radiance Start Input Latitude, Longitude, Solar Angles, Viewing Angles, Tangent Heights, and a Priori Aerosol Profile SCIATRAN Simulated Radiance Radiance Normalization Forward Model y = F( x ) + e y S x n + 1 n + 1 = ( S 1 = x + S + K T 1 n y n + 1 S K ) 1 n K S [( y y ) K ( x T 1 n y n n x )] n Aerosol Profile Criterion for Iteration No Yes Output Retrieved Profile End
Sensitivity Studies Sensitivity analyses are performed to investigate the impact of the bias in the assumed surface albedo on the accuracy of the aerosol extinction retrieval. The average surface albedo is used in the calculation: Relative errors (Err) of the retrieved aerosol extinctions by using the assumed surface albedo in the retrieval when the actual surface albedo values are A =.1,.2,.4 and.5. Height (km) A =.1 Err (%) A =.2 Err (%) A =.4 Err (%) A =.5 Err (%) 3.4 1.2.8 1.5 3.7 1.8 1.2 2.5 4.9 2. 1.6 3.1 5.3 2.7 1.8 3.5 6.2 3.6 2.2 4.2
COMPARISONS With SAGE II retrieved SAGE II The retrieved SCIAMACHY 1 nm aerosol extinction profile (36. ºN, 97.3 ºE) on 21 January 4 at 3:42 UT compared to SAGE II profile (.2 ºN, 93.9 ºE) on the same day at 1:55 UT. Relative difference = (SCIAMACHY SAHE II) /SAGE II 1%. 1E-6 1E-5 1E-4 Aerosol extinction (km -1 ) --1-5 5 1 Relative difference (%) retrieved SAGE II Same as the above Figure, but for the average values of 12 individual measurements. The measurements are within 4 ºN latitude band, and were performed on 21 January 4. 1E-6 1E-5 1E-4 Aerosol extinction (km -1 ) -4-4 Relative difference (%)
CONCLUSIONS Stratospheric aerosol extinction profiles are retrieved from SCIAMACHY Limb observations Sensitivity analyses revealed that the errors caused by the bias of the assumed surface albedo in the retrieval are generally below 6% in the northern midlatitudes. Comparisons with the SAGE II measurements showed that the retrieved SCIAMACHY limb aerosol extinction profiles have good agreement with the SAGE II profiles. Based on the present reported study, we can conclude that our retrievals are reliable and accurate. The validation will be extended in the near future to include additional retrieval profiles and satellite measurements.
We wish to acknowledge the European Space Agency for SCIAMACHY data. We also wish to acknowledge the NASA Langley Research Center for the SAGE II aerosol data. We are thankful to the Institute of Remote Sensing, University of Bremen for the SCIATRAN software package. The research was funded by the National Natural Science Foundation of China (Grant No. 4127547, and 4117529), and the National Basic Research Program of China (Grant No. 11CB4341). [1] Bourassa, A. E., D. A. Degenstein, R. L. Gattinger, et al., 7: Stratospheric aerosol retrieval with optical spectrograph and infrared imaging system limb scatter measurements, J. Geophys. Res., 112, D1217, doi:1.129/6jd879. [2] Bovensmann, H., J. P. Burrows, M. Buchwitz, et al., 1999: SCIAMACHY: Mission objectives and measurement modes, J. Atmos. Sci., 56(2), 127. [3] Ovigneur, B., Landgraf, J., Snel, R., et al., 11: Retrieval of stratospheric aerosol density profiles from SCIAMACHY limb radiance measurements in the O2 A-band, Atmos. Meas. Tech., 4, 29 2373, doi:1.5194/amt-4-29-11. [4] Rodgers, C. D., : Inverse Methods for Atmospheric Sounding: Theory and Practice, World Scientific, Singapore. [5] Rozanov, V. V., M. Buchwitz, K.-U. Eichmann, et al., 2: SCIATRAN A new radiative transfer model for geophysical applications in the 24 24nm spectral region: The pseudo-spherical version, Adv. Space Res., 11, 1831 18. [6] Rozanov, A., V. Rozanov, M. Buchwitz, et al., 5: SCIATRAN 2. A new radiative transfer model for geophysical applications in the 175 24nm spectral region, Adv. Space Res., 36, 1 119. [7] Taha, G., D. F. Rault, R. P. Loughman, et al., 11: SCIAMACHY stratospheric aerosol extinction profile retrieval using the OMPS/LP algorithm, Atmos. Meas. Tech., 4, 547 556, doi:1.5194/amt-4-547-11.