GERB/CERES Comparisons Update

Transcription

1 GERB/CERES Comparisons Update N. Clerbaux, S. Dewitte, A. Ipe, P.-J. Baeck, A. Velazquez, I. Decoster Royal Meteorological Institute of Belgium, Department of Observations, Section Remote Sensing from Space, Avenue Circulaire 3, B-80 Brussels, Belgium. Abstract The paper provides quantitative results of comparison between the GERB and CERES instruments in January and February The comparisons have been done for the radiance and the flux and for the shortwave and longwave radiation. We have followed the same methodology as in a previous GERB/CERES comparison exercise [Clerbaux et al., 2009]. Therefore we do not provide all the details here. In the SW, GERB- is closer to CERES than GERB-2. The results are consistent with a decrease of 3-4% of the GERB SW level between GERB-2 and GERB-. As it was already observed, the SW flux ratio is higher than the radiance ratio by about -2%. Finally for the SW, this work indicates a decrease of the GERB-2/CERES ratio by about.7% between 2004 and This decrease is under investigation. Concerning the LW radiation, GERB- is slightly higher than GERB-2 ( 0.4%) and closer to CERES. The radiance and the flux ratio are in good agreement. The GERB-2/CERES ratio are similar in 2007 and in We can conclude that GERB and CERES are both stable in the LW. Introduction Since February 2004 direct observation of the broadband radiant energy leaving the Top Of Atmosphere (TOA) is available from the Geostationary Earth Radiation Budget (GERB) [Harries et al., 2005] instrument on the Meteosat Second Generation satellites. The GERB-2 instrument, on Meteosat-8, was in charge of the operational GERB observation up to May At that time, the GERB- instrument on board Meteosat-9 becomes the operational one while GERB-2 mostly remains in SAFE mode. Simultaneous observations with both instruments are however regularly realized since 2006 (about twice per year). In this study the GERB observations from both GERB- and from GERB-2 are compared to the observations realized by the Cloud and Earth Radiant Energy System (CERES) [Wielicki et al., 996] instruments. The methodology is the same as the one followed during a previous GERB/CERES comparison for observations in June and December 2004 [Clerbaux et al., 2009]. 2 Data This study is based on GERB and CERES data acquired during 30 days, from 3 January 2007 to February Nicolas.Clerbaux@oma.be

2 GERB- Number of shortwave observation pairs INST α < 2 α < 5 α < 8 Flux FM FM FM FM GERB- Number of Longwave observation pairs INST α < 2 α < 5 α < 8 Flux FM FM FM FM GERB-2 Number of shortwave observation pairs INST α < 2 α < 5 α < 8 Flux FM FM FM FM GERB-2 Number of Longwave observation pairs INST α < 2 α < 5 α < 8 Flux FM FM FM FM Table : Numbers of coangular radiance pairs and colocated flux pairs for GERB- (top) and GERB-2 (bottom) used for the SW and LW comparisons. The GERB-2 data are the Edition- for the format and the Version 3 for the and formats. These data have been generated in near real time by the RMIB GERB Processing (RGP) system. Note that the Version 3 processing is similar to the Edition-. The GERB- data are Version 6 which is the latest processing version that is expected to become the Edition- in a close future. With respect to the near real time Version 5, the new version provides improved A values for the detectors and also fixes a bug that affected the cloud retrieval (use of a wrong LUT in Version 5). For CERES, the SSF Edition2F (FM and FM2) or Edition 2C (FM3 and FM4) have been used with application of the Revision-, using the scaling factors provided by the CERES team. Note that the SW detector of the FM4 instrument is broken since 30 March This instrument is therefore not useful for the comparisons in 2007, except for the nighttime longwave comparison, as in this case the LW is equal to the TOT measurement. Table gives the numbers of coangular radiance observations (for different maximum angles α between the GERB and CERES directions of observation) and the number of collocated flux between the 2 GERB instruments and the 4 CERES instruments. The higher number of matches for the FM2 is provided by operating this instrument in a scanning mode that optimizes the number of coangular observations with GERB. Compared to the previous comparison of 2004, the number of matches is smaller due to the use of less days for the comparison. As already said, the FM4 does not provide neither SW observation nor LW observations during day time. 2

3 G/CERES FM SW Radiance Ratio G/CERES FM G/CERES FM2 SW Radiance Ratio G/CERES FM2 G/CERES FM3 SW Radiance Ratio G/CERES FM G2/CERES FM SW Radiance Ratio G2/CERES FM G2/CERES FM2 SW Radiance Ratio G2/CERES FM2 G2/CERES FM3 SW Radiance Ratio G2/CERES FM Figure : GERB/CERES SW radiance ratio and uncertainty in reflectance bins for α < 5. 3 Shortwave radiance comparison Table 2 provides the shortwave radiance comparison results for the α < 5 criteria (similar results, not shown, are obtained with the α < 2 and α < 8 criteria). The results for the format are not given here, as they are significantly affected by the non correction of the GERB PSF in the matching with CERES [Clerbaux et al., 2009]. In January 2007, the GERB- SW radiances are very close to the CERES ones. The GERB/CERES ratio are around 95,.002 and.007 for the FM, FM2 and FM3 respectively. The uncertainty on the ratio is evaluated to about 0.005, thus half a percent. Compare to the same CERES instruments, the ratio for GERB-2 are about 3% 4 % higher than for GERB-. This result is consistent with the GERB-/GERB-2 comparisons performed at Imperial College London and presented during the GIST meetings. It is interesting to quantify the scene type dependency of the GERB/CERES ratio as it is an indirect validation of the instrument spectral response and unfiltering strategy. Figure shows the variation of the ratio as a function of the scene reflectance (i.e. radiance normalized by the incoming solar irradiance). The ratio is very stable for the and formats (the scene type dependency that seems to affect the is due to the PSF). 4 SShortwave flux comparison Table 3 summarizes the SW flux comparison in a similar form to that given in Table 2 for the SW radiance. All together, the flux ratio are about.5% higher than the ratio observed in radiance. This ratio increase is observed for GERB- and GERB-2, and was already observed for GERB-2 in the 2004 comparison. 3

4 GERB- Binned Averaged Rectified Geolocated () Allsky 96 ± ( 95 ).00 ± (.002 ).008 ± (.007 ) Overcast 92 ± 0.0 ( 87 ) 96 ± 0.04 ( 98 ).005 ± 0.06 (.007 ) Clearsky.07 ± (.03 ).0 ± (.007 ).0 ± (.08 ) ocean.038 ± (.027 ).08 ± (.032 ) 82 ± (.003 ) dark veg. 97 ± 0.02 ( 98 ).004 ± (.005 ).023 ± (.09 ) bright veg. 92 ± 0.00 ( 95 ).000 ± ( 99 ).034 ± (.040 ) dark desert - 97 ± 0.05 (.000 ).022 ± 0.00 (.08 ) bright desert -.0 ± (.009 ).027 ± (.023 ) GERB- High Resolution () Allsky 95 ± ( 95 ).002 ± (.004 ).009 ± (.007 ) Overcast 83 ± ( 8 ) 94 ± 0.03 ( 94 ) 98 ± 0.03 ( 99 ) Clearsky.07 ± 0.04 (.006 ).00 ± (.008 ).06 ± 0.03 (.020 ) ocean.032 ± (.06 ).026 ± 0.09 (.037 ).00 ± (.07 ) dark veg..000 ± (.000 ).006 ± (.007 ).022 ± 0.0 (.09 ) bright veg. 89 ± 0.00 ( 9 ).000 ± (.00 ).025 ± 0.06 (.034 ) dark desert ± (.006 ).07 ± (.04 ) bright desert -.02 ± (.009 ).028 ± (.029 ) GERB-2 Binned Averaged Rectified Geolocated () Allsky.026 ± (.025 ).039 ± (.04 ).048 ± (.047 ) Overcast.038 ± 0.02 (.032 ).039 ± (.039 ).047 ± 0.09 (.050 ) Clearsky.042 ± (.042 ).05 ± (.05 ).055 ± 0.05 (.067 ) ocean.046 ± (.039 ).058 ± (.049 ) 70 ± ( 52 ) dark veg..043 ± 0.05 (.039 ).056 ± 0.0 (.053 ).079 ± 0.02 (.072 ) bright veg..066 ± (.067 ).058 ± 0.03 (.045 ).094 ± (.093 ) dark desert ± 0.05 (.04 ).074 ± 0.06 (.069 ) bright desert ± (.055 ).079 ± (.085 ) GERB-2 High Resolution () Allsky.025 ± (.025 ).039 ± (.042 ).045 ± (.048 ) Overcast.03 ± (.05 ).027 ± 0.02 (.03 ).046 ± 0.02 (.037 ) Clearsky.03 ± 0.05 (.026 ).049 ± (.052 ).059 ± (.070 ) ocean.024 ± 0.08 (.020 ).046 ± 0.06 (.049 ) 97 ± (.009 ) dark veg..04 ± 0.00 (.035 ).055 ± (.055 ).076 ± 0.0 (.074 ) bright veg..044 ± (.045 ).047 ± (.045 ).082 ± 0.07 (.09 ) dark desert ± (.048 ).068 ± 0.02 (.067 ) bright desert ± (.055 ).079 ± (.079 ) Table 2: GERB /CERES SW radiance ratio m and uncertainty for α < 5. 4

5 GERB- - Binned Averaged Rectified Geolocated () Allsky.0 ± ± ± Overcast.007 ± ± ± Clearsky.05 ± ± ± ocean.022 ± ± ± dark veg. 9 ± ± ± bright veg..023 ± ± ± dark desert.002 ± ± ± bright desert.06 ± ± ± GERB- - High Resolution () Allsky.00 ± ± ± Overcast.000 ± ± ± Clearsky.06 ± ± ± ocean.028 ± ± ± dark veg. 94 ± ± ± bright veg..09 ± ± ± dark desert.02 ± ± ± bright desert.07 ± ± ± GERB-2 - Binned Averaged Rectified Geolocated () Allsky.047 ± ± ± Overcast.042 ± ± ± Clearsky.066 ± ± ± ocean.02 ± ± ± dark veg..053 ± ± ± bright veg..076 ± ± ± dark desert.053 ± ± ± bright desert.070 ± ± ± GERB-2 - High Resolution () Allsky.047 ± ± ± Overcast.037 ± ± ± Clearsky.066 ± ± ± ocean.027 ± ± ± dark veg..058 ± ± ± bright veg..072 ± ± ± dark desert.065 ± ± ± bright desert.07 ± ± ± Table 3: GERB /CERES SW flux ratio and uncertainty. 5

6 GERB- - Binned Averaged Rectified Geolocated () Allsky 94 ± ± ± ± Day 99 ± ± ± Night 87 ± ± ± ± Clearsky 88 ± ± ± ± Cloudy 97 ± ± ± ± GERB- - High Resolution () Allsky 93 ± ± ± ± Day 99 ± ± ± Night 87 ± ± ± ± Clearsky 90 ± ± ± ± Cloudy 95 ± ± ± ± GERB-2 - Binned Averaged Rectified Geolocated () Allsky 89 ± ± ± ± Day 95 ± ± ± Night 83 ± ± ± ± Clearsky 85 ± ± ± ± Cloudy 89 ± ± ± ± GERB-2 - High Resolution () Allsky 88 ± ± ± ± Day 95 ± ± ± Night 82 ± ± ± ± Clearsky 86 ± ± ± ± Cloudy 88 ± ± ± ± Table 4: GERB /CERES LW radiance ratio and uncertainty for α < 5. 5 Longwave radiance comparison Table 4 displays the GERB/CERES LW radiance comparison results. The GERB LW radiances are lower than CERES by about % for GERB- and.3% for GERB-2. However, the longwave radiance ratio differs significantly between the 4 CERES instruments. 6 Longwave flux comparison Table 5 reports the LW flux intercomparisons in a similar form to Tables 4. The GERB-/CERES flux ratio in all sky conditions lies between 88 (FM4 - nighttime only) and 95 (FM2). The average across the 4 CERES instruments is 9 (GERB-) and 95 (GERB-2) which is in agreement with the radiance comparison. Slightly lower (about 0.2%) ratio are observed for the format than for the format. 7 Conclusions The SW comparisons are summarized in the following table of GERB/CERES ratio values: 6

7 GERB- Binned Averaged Rectified Geolocated () Allsky 92 ± ± ± ± Day 97 ± ± ± Night 87 ± ± ± ± Clearsky 85 ± ± ± ± Cloudy 99 ± ± ± ± GERB- High Resolution () Allsky 89 ± ± ± ± Day 94 ± ± ± Night 84 ± ± ± ± Clearsky 83 ± ± ± ± Cloudy 93 ± ± ± ± GERB-2 - Binned Averaged Rectified Geolocated () Allsky 88 ± ± ± ± Day 93 ± ± ± Night 83 ± ± ± ± Clearsky 82 ± ± ± ± Cloudy 93 ± ± ± ± GERB-2 - High Resolution () Allsky 85 ± ± ± ± Day 89 ± ± ± Night 8 ± ± ± ± Clearsky 80 ± ± ± ± Cloudy 87 ± ± ± ± Table 5: GERB/CERES LW flux ratio m and uncertainty. FM FM2 FM3 FM4 GERB- SW rad GERB-2 SW rad GERB-2 SW rad GERB- SW flux GERB-2 SW flux GERB-2 SW flux We can conclude that: GERB- is lower than GERB-2 by about 3% 4%. The GERB/CERES radiance ratio is lower than the flux ratio by about % 2%. The GERB-2/CERES ratio has decreased by about.7% between 2004 and The magnitude of this decrease is similar for the different CERES instruments. An aging of GERB-2 could be an explanation of this. However, the SW calibration monitoring device (an integrating sphere) does not provide evidence of SW aging. The LW comparisons are summarized in the following table of GERB/CERES ratio values: 7

8 FM FM2 FM3 FM4 GERB- LW rad GERB-2 LW rad GERB-2 LW rad GERB- LW flux GERB-2 LW flux GERB-2 LW flux We can conclude that: GERB- is slightly higher than GERB-2, by about 0.4%. The radiance and the flux ratio are in good agreement. The GERB-2/CERES ratio are similar in 2007 and in CERES are both stable in the LW. We can conclude that GERB and References [] N. Clerbaux, J.E. Russell, S. Dewitte, C. Bertrand, D. Caprion, B. De Paepe, L. Gonzalez Sotelino, A. Ipe, R. Bantges, H.E. Brindley, 2009: Comparison of GERB instantaneous radiance and flux products 2 with CERES Edition-2 data, Rem. Sens. Environ., 3, [2] Harries, J.E., et al., 2005: The Geostationary Earth Radiation Budget Experiment (GERB), Bull. Amer. Meteorol. Soc., 86(7): [3] Wielicki, B. A., et al., 996: Clouds and the Earth s Radiant Energy System (CERES): An Earth Observing System Experiment, Bull. Amer. Meteorol. Soc., 77, ,. 8