![NIES Lidar Network Observation Compact 2 (532, 1064nm) + 1 (532nm) Mie lidar with automatically measurement capability 20 sites ground based network observation in East Asia (2001~) Ship-borne measurements (1999~, vessel “MIRAI” (JAMSTEC)) [Sugimoto et al., 2001; 2005] Data analysis Classify aerosol components and Retrieve their extinctions at each layer (assuming external mixture of each aerosol component) 1 (532)+1 data Dust ( nonSpherical ) + non-Dust ( Spherical ) [Sugimoto et al., 2003; Shimizu et al., 2004] 2 data Air-pollution aerosol*( Small ) + Sea-salt or Dust( Large ) [Nishizawa et al., 2007; 2008] 2 +1 data Air-pollution aerosol* ( Spherical / Small ) + Sea-salt ( Spherical / Large ) + Dust ( nonSpherical / Large ) [Nishizawa et al., 2010] Polarization Spectral Polarization + Spectral *Air-pollution aerosol is defined as mixture of Sulfate, Nitrate, Organic carbon, and Black carbon](https://image.slidesharecdn.com/fr2-110729115951-phpapp02/85/AEROSOL-CLASSIFICATION-RETRIEVAL-ALGORITHMS-FOR-EARTHCARE-ATLID-CALIPSO-CALIOP-AND-GROUND-BASED-LIDARS-3-320.jpg)
![Comparison with a global aerosol transport model “SPRINTARS” [Nishizawa et al. JGR 2008] *SPRINTARS is a global, three-dimensional aerosol transport model [Takemura et al. 2005]. The simulation data by the SPRINTARS was provided by Takemura of Kyusyu Univ. Mean values (Obs.)=0.0006 km -1 sr -1 (Sim.)=0.0003 km -1 sr -1 Mean values (Obs.)=0.0027 km -1 sr -1 (Sim.)=0.0017 km -1 sr -1 532 1064 532 SPRINTARS Lidar Mean values (Obs.)=0.044 km -1 (Sim.)=0.009 km -1 Mean values (Obs.)=0.005 km -1 (Sim.)=0.014 km -1 AP SS AP SPRINTARS Lidar](https://image.slidesharecdn.com/fr2-110729115951-phpapp02/85/AEROSOL-CLASSIFICATION-RETRIEVAL-ALGORITHMS-FOR-EARTHCARE-ATLID-CALIPSO-CALIOP-AND-GROUND-BASED-LIDARS-9-320.jpg)
![Application to satellite-borne 2 +1 lidar [CALIOP/NASA 2006~] Saharan Dust transport to the Atlantic Ocean 2006.8/1, 2:36UTC Aerosol Mask Scheme ● Remove cloud area CloudSat + CALIOP [Hagihara et al. 2009] ● Remove molecule scat. area CALIOP (β 1064 ) ＊ β 1064 was re-calibrated by using water-cloud signals β 1064 β 532 δ 532 Air pollution Sea-salt Dust Altitude [km] Altitude [km] Latitude [deg] Cited from NASA/CALIOP website](https://image.slidesharecdn.com/fr2-110729115951-phpapp02/85/AEROSOL-CLASSIFICATION-RETRIEVAL-ALGORITHMS-FOR-EARTHCARE-ATLID-CALIPSO-CALIOP-AND-GROUND-BASED-LIDARS-10-320.jpg)
![Aerosol classification algorithms using 1α+2β+1δ data Classify aerosol components and Retrieve their extinctions at each layer (assuming external mixture of each aerosol component) 1α+2 data SF-NT-OC ( Weak / Small ) + BC ( Strong / Small ) + Dust ( Weak / Large ) [Nishizawa et al., 2008] 1α+1 +1 data SF-NT-OC ( Weak / Spherical ) + BC ( Strong / Spherical ) + Dust ( Weak / Non-spherical ) *Air-pollution aerosol is defined as mixture of Sulfate ( SF ), Nitrate ( NT ), Organic carbon ( OC ), and Black carbon ( BC ) Light absorption + Spectral Light absorption + Polarization](https://image.slidesharecdn.com/fr2-110729115951-phpapp02/85/AEROSOL-CLASSIFICATION-RETRIEVAL-ALGORITHMS-FOR-EARTHCARE-ATLID-CALIPSO-CALIOP-AND-GROUND-BASED-LIDARS-12-320.jpg)
AEROSOL CLASSIFICATION RETRIEVAL ALGORITHMS FOR EARTHCARE/ATLID, CALIPSO/CALIOP, AND GROUND-BASED LIDARS
- 1. Sugimoto, N., T. Nishizawa, I. Matsui, National Institute for Environmental Studies (NIES), Tsukuba, Japan H. Okamoto Kyushu Univ., Fukuoka, Japan AEROSOL CLASSIFICATION RETRIEVAL ALGORITHMS FOR EARTHCARE/ATLID, CALIPSO/CALIOP, AND GROUND-BASED LIDARS IGARSS 2011, 29/Jul/2011 FR2T03
- 2. NIES Lidar Network 20 observation sites in East-Asia using 2 + 1 Mie lidar 532nm attenuated Backscatter ( 532 ) 532nm total depolarization ( 532 ) 1064nm attenuated backscatter ( 1064 ) Measured data APD (1064nm) PMTs (532nm) 2 + 1 Mie lidar China Japan Thai Mongol Korea NIES Lidar network Lidar at “Hedo” site The lidars measure aerosols (& clouds) 24-hour- automatically and we provide 2 + 1 data in semi-real-time (http://www-lidar.nies.go.jp/)
- 3. NIES Lidar Network Observation Compact 2 (532, 1064nm) + 1 (532nm) Mie lidar with automatically measurement capability 20 sites ground based network observation in East Asia (2001~) Ship-borne measurements (1999~, vessel “MIRAI” (JAMSTEC)) [Sugimoto et al., 2001; 2005] Data analysis Classify aerosol components and Retrieve their extinctions at each layer (assuming external mixture of each aerosol component) 1 (532)+1 data Dust ( nonSpherical ) + non-Dust ( Spherical ) [Sugimoto et al., 2003; Shimizu et al., 2004] 2 data Air-pollution aerosol*( Small ) + Sea-salt or Dust( Large ) [Nishizawa et al., 2007; 2008] 2 +1 data Air-pollution aerosol* ( Spherical / Small ) + Sea-salt ( Spherical / Large ) + Dust ( nonSpherical / Large ) [Nishizawa et al., 2010] Polarization Spectral Polarization + Spectral *Air-pollution aerosol is defined as mixture of Sulfate, Nitrate, Organic carbon, and Black carbon
- 4. 2 +1 algorithm Spheroidal for dust (Spherical for the other components) Assumptions Log-normal size distribution Mode radius, standard deviation, refractive indexes 3 components in each layer AP : Air-pollution SS : Sea-salt DS : Dust r m : Mode radius S : Lidar ratio (Extinction-to-Backscatter ratio) δ : Particle depolarization ratio AP SS DS r m 0.13 3.0 2.0 S 55 20 48 0 0 0.3 532, || 1064 532, SS AP DS
- 5. Application to shipborne lidar data I Pacific Ocean near Japan Observed data ( 2 + 1 Mie lidar ) Tohoku Univ. HP, http://caos-a.geophys.tohoku.ac.jp MIRAI/JAMSTEC 14 days 6 km 0 km 532 1064 6 km 0 km 532 6 km 0 km
- 6. Retrieved aerosol component data Air-pollution aerosols Dust Sea-salt Agreement within 5% AOT (532) Angstrom Total
- 7. Application to shipborne lidar data II Tropical Pacific Ocean Mirai Cruises MR01K05: 9.21 ~ 12.17, 2001 MR04K07: 11.18 ~ 12.9, 2004 MR04K08: 12.16 ~ 2.17, 2005 MR06K05: 10.16 ~ 11.25, 2006 7-month data in total
- 8. Total Air-Pollution SS DS 12-hour average Horizontal distribution (Optical thickness) The total optical thicknesses were larger from the Japan to the New Guinea and in the western region off Sumatra Island than in the other regions. AP was the major contributor to the total optical thickness of aerosols.
- 9. Comparison with a global aerosol transport model “SPRINTARS” [Nishizawa et al. JGR 2008] *SPRINTARS is a global, three-dimensional aerosol transport model [Takemura et al. 2005]. The simulation data by the SPRINTARS was provided by Takemura of Kyusyu Univ. Mean values (Obs.)=0.0006 km -1 sr -1 (Sim.)=0.0003 km -1 sr -1 Mean values (Obs.)=0.0027 km -1 sr -1 (Sim.)=0.0017 km -1 sr -1 532 1064 532 SPRINTARS Lidar Mean values (Obs.)=0.044 km -1 (Sim.)=0.009 km -1 Mean values (Obs.)=0.005 km -1 (Sim.)=0.014 km -1 AP SS AP SPRINTARS Lidar
- 10. Application to satellite-borne 2 +1 lidar [CALIOP/NASA 2006~] Saharan Dust transport to the Atlantic Ocean 2006.8/1, 2:36UTC Aerosol Mask Scheme ● Remove cloud area CloudSat + CALIOP [Hagihara et al. 2009] ● Remove molecule scat. area CALIOP (β 1064 ) ＊ β 1064 was re-calibrated by using water-cloud signals β 1064 β 532 δ 532 Air pollution Sea-salt Dust Altitude [km] Altitude [km] Latitude [deg] Cited from NASA/CALIOP website
- 11. Mie-lidar and High-Spectral-Resolution-Lidar (HSRL) measurements (1α+2β+1δ) Observed data April 8 2005, 0~10 UTC 532 S 532 = 532 / 532 1064 Observation Site NIES, Tsukuba HSRL Extinction (532nm ） Backscatter (532nm ） Mie lidar Backscatter (1064nm ） Depolarization (532nm) 532 HSRL Mie lidar
- 12. Aerosol classification algorithms using 1α+2β+1δ data Classify aerosol components and Retrieve their extinctions at each layer (assuming external mixture of each aerosol component) 1α+2 data SF-NT-OC ( Weak / Small ) + BC ( Strong / Small ) + Dust ( Weak / Large ) [Nishizawa et al., 2008] 1α+1 +1 data SF-NT-OC ( Weak / Spherical ) + BC ( Strong / Spherical ) + Dust ( Weak / Non-spherical ) *Air-pollution aerosol is defined as mixture of Sulfate ( SF ), Nitrate ( NT ), Organic carbon ( OC ), and Black carbon ( BC ) Light absorption + Spectral Light absorption + Polarization
- 13. 1 + 1 + 1 algorithm Dust + BC + SF-NT-OC Spheroidal for dust (Spherical for the other components) Assumptions Log-normal size distribution Mode radius, standard deviation, refractive indexes r m : Mode radius S : Lidar ratio (Extinction-to-Backscatter ratio) δ : Particle depolarization ratio SF-NT-OC BC DS r m 0.13 0.05 2.0 S 55 101 48 0 0 0.3 α 532 532, || 532, SF-NT-OC BC DS
- 14. Application to ground-based Mie/HSRL data ( NIES, Tsukuba, Japan) Estimates SF-NT-OC Dust BC Sulfate originated from Coastal area of China Dust originated from Gobi desert BC+OC originated from Coastal area of China and Indochina peninsula SPRINTARS Provided by Dr. Takemura (Kyusyu Univ.) Sulfate Dust BC+OC Observation site
- 15. ATLID / EarthCARE (2015 ~) 355 nm High Spectral Resolution lidar (HSRL) 3 channels: 1α+1 +1 Extinction coefficient ( ) Backscattering coefficient ( ) Depolarization ratio ( )
- 16. Summary We developed several aerosol classification and retrieval algorithms. => The algorithms can be used to understand aerosol component distributions in regional and global scales by applying to the network lidar data and the satellite-borne lidar data. We are going on developing (or improving) aerosol classification and retrieval algorithms using more channels. NIES 2 +1 Mie lidar + Raman (or HSRL) 1α+2 +1 data SF-NT-OC ( Weak / Small / Spherical ) + BC ( Strong / Small / Spherical ) + Dust ( Weak / Large / Non-spherical ) + Sea-salt ( Weak / Large / Spherical ) NIES 2α+3 +2 HSRL (Under development : Nishizawa et al. FR2T07 ) 2α+3 +2 data SF-NT-OC ( Weak / Small / Spherical ) + BC ( Strong / Small / Spherical ) + Dust ( Weak / Large / Non-spherical ) + Sea-salt ( Weak / Large / Spherical ) + Size information for SF-NT-OC, Dust, Sea-salt Light absorption + Spectral + Polarization Light absorption + Spectral + Polarization