Stratospheric Aerosol Loading after Pinatubo Studies Using a
Stratospheric Aerosol Loading after Pinatubo: Studies Using a Fully Coupled Sulfate Aerosol. Chemistry-Climate-Model (CCM) Thomas Peter 1, Jian-Xiong Sheng 1, Debra Weisenstein 2, Beiping Luo 1, Eugene Rozanov 1, 3 1 IAC ETHZ, Zurich; 2 Harvard University, Cambridge, MA; 3 PMOD/WRC, Davos SPARC ASAP (2006) The Challenge: Nucleation / Condensation / Evaporation / Coagulation / Transport / Sedimentation E. g. , sedimentation is a major revomal processes after large volcanic eruptions, depends on the particle size: 10% larger radius 60% more gravitational mass flux
SOCOLA: A fully coupled sulfate aerosol CCM This is a “spectral” or “sectional” or “size-resolving” or “bin” model. Not a “modal” model. Not a “bulk model. § CCM SOCOL v 3. 0: § MA-ECHAM 5 (GCM) and MEZON (CTM) § 39 sigma-pressure vertical levels (up to 0. 01 h. Pa); horizontal T 31 or T 42 § Sea surface temperature prescribed, QBO is nudged based on observations § AER aerosol module: § all relevant sulfur chemistry and microphysics § 40 size bins: 0. 39 nm - 3. 2 μm (volume doubling) § size-dependent composition (Kelvin effect) § Model coupling: § Each aerosol size bin is transported separately in ECHAM 5 incl. sedimentation § Interactive radiation scheme: online calculation of optical properties for 22 SOCOL wave bands Friday, October 2, 2020 2 2
Schematic diagram of the global sulfur budget: Burdens in Gg S, fluxes in Gg S/yr Red: SOCOLA Black: AER (as SPARC 2006 ASAP) and obs in brackets Green: GOCART model (Chin et al. , 2000) Purple: Chin and Davis (1995) Meteoritic and other non-S species not yet included 3
SOCOLA comparison with OPC measurements at Laramie, WY, 2000 -2010 Solid: OPC cumulative # density (cm-3) Dashed: SOCOLA Symbols: OPC volume density (μm 3/cm 3) Red line: SOCOLA Excellent agreement in volume density. Good agreement in size distributions, however, deviations for large particles at higher alts: - pvap is too low? Kelvin effect too weak? - Nucleation too slow? Coagulation is too fast? - Too fast meridional transport in the model responsible? 4
SAGE_4 Stenchikov et al. (1998) SAGE-4 dataset for CCMI • • 5 1024 nm extinction Zonal mean 5°S-5°N August 1991 bars: standard dev. of SAGE II (ASAP Report) Ø The ASAP Report pointed to the importance of the new gap-filling procedure. Ø However, we failed to highlight the massive shortcomings of the outdated datasets. Ø Furthermore, there is a general lack of accountability and tracability in the old literature (e. g. missing version numbers of SAGE data used) 5
Comparison of extinction Friday, October 2, 2020 6 6
Observations of stratospheric aerosol burden – what do we really know? HIRS: Baran and Foot (JGR, 1994): Did "cloud-clearing" work correctly? Also, figure could be read as H 2 SO 4 instead aerosol (Anthony Baran, pers. comm. ) Laramie: OPC from ASAP report: Used Mc. Cormick et al. (Nature, 1995) for scaling (who concluded 10 TG S injection and ~30 Tg aerosol peak loading) Laramie scaled: OPC from ASAP report: Scaling to only ~21 Tg aerosol peak loading) Laramie scaled: Agrees well with HIRS during early phase and with SAGE-4 from about 9 months after the eruption onward 7
Observations in comparison with Jason English’s modeling work 8
Coagulation efficiency ε Coagulation reduces the number density of small particles very efficiently and produces large particles rate of change of number density n due to coagulation ~ - K n 2 Coagulation kernal K is given by Fuch (1964) assuming a UNIT coagulation efficiency English et al. , (2013) explicitly introduce Van der Waals forces into K which lead to an increased coagulation efficiency in transition and free molecular regimes. English et al. , 2013 Friday, October 2, 2020 9 9
Low coagulation efficiency may prevail in free molecular regime! Narsimhan and Ruckenstein (1985), D’Alessio et al. (2005), Lindstedt and Waldheim (2013) • van-der-Waals force attractive • but molecular polarization needs time, otherwise force is repulsive Friday, October 2, 2020 10 10
Experiments – Coagulation Efficiency a Pinatubo Simulation: 20 Mt SO 2 injected into 16 – 30 km with maximum at 20 - 22 km § ε = 1 reference run with QBO and interactive radiation § ε = 2 double coagulation efficiency in transition and free molecular regimes (simplifying VDW) ε Cont. Transition free molec. § ε = step § ε = smooth based on Narsimhan and Ruckenstein (1985) with Hamaker constant 5 x 10 -13 erg Friday, October 2, 2020 11
Time Evolution of Aerosol Burden in SOCOLA Friday, October 2, 2020 12 12
Time Evolution of Aerosol Opt. Depth in SOCOLA Number of SAGE-II observations available and gap-filled surface area density (SAD) Friday, October 2, 2020 13 13
SOCOLA Particle Sizes Compared with SAGE-II rmode = 0. 205 μm for ε = 2 rmode = 0. 193 μm for ε = Lennard. Jones • 6% in size • 33% sedimentation flux Friday, October 2, 2020 14 14
Results – Particle Size Comparison with OPC measurement December 1991 R > 0. 25 um Friday, October 2, 2020 R > 0. 01 um 15 15
16
Conclusions § SAGE-4λ homogenizes different aerosol data sets (photometers, SAGE I, SAM II, SAGE II, CALIOP) – probably the best we presently have in terms of observations. § Uncertainties in initial S loading remain high ( 10%), despite careful gap-filling in SAGE-4λ § Aerosol-CCM SOCOLA successfully simulates the time-space development of the main Pinatubo cloud § Tropopause heating is realistic with the observational dataset SAGE 4λ and with SOCOLA § Middle stratospheric aerosols in SOCOLA are too abundant after Pinatubo and under background conditions – for unknown reasons. § Coagulation efficiencies appear to play only a smaller role – if anything ε < 1 might help, not ε > 1 Friday, October 2, 2020 17 17
Stratospheric Aerosol Loading after Pinatubo: Studies Using a Fully Coupled Sulfate Aerosol-Chemistry-Climate-Model Thomas Peter, Jianxiong Sheng, Debra Weisenstein, Beiping Luo, Eugene Rozanov The eruption of Mt. Pinatubo in June 1991 emitted 7– 11 Mt sulfur into the lower stratosphere. Previous modeling approaches have been unsuccessful to accurately predict the evolution of the stratospheric aerosol burden, and thus the climatic effects. Most modeling attempts yield shorter residence times of the volcanic aerosols in the stratosphere compared to satellite observations. In this study, we coupled a sulfate aerosol module to the global chemistry-climate model (CCM) SOCOL. The AER aerosol module includes comprehensive sulfur chemistry and microphysics, in which the particles are size-resolved by 40 size bins spanning radii from 0. 39 nm to 3. 2 μm. Radiative forcing is calculated online from the aerosol module using Mie theory. The eruption is modelled assuming various magnitudes of SO 2 injections, as well as different initial vertical and meriodional distributions. We put particular emphasis on investigating the aerosol particle coagulation efficiency. By means of a Lennard-Jones potential in the free molecular regime, where very low efficiency values may prevail, the results show that after an eruption the simulated stratospheric aerosol burden agrees well with the high-resolution infrared radiation sounder (HIRS) measurements and the Laramie in situ measurements. Although the coagulation efficiency is largely uncertain, this suggests that the sedimenting large particles form more slowly due to the reduced coagulation process under stratospheric conditions, and therefore stratospheric aerosols may have a longer residence time than derived in previous studies. This study represents the first steps to properly predict large volcanic eruptions and also potential geoengineering measures against dangerous climate change. 18
SAGE better (HIRS also has troposphere) HIRS better (SAGE is blind) English VDW vander Waa x 1. 3 19
Stratosphere: free molecular regime for ‘small’ particles Knudsen number Kn = mean free path / particle radius Kn < 0. 1 continuum regime 0. 1< Kn < 10 transition regime Kn > 10 free molecular regime In stratosphere: particles (<50 nm) are generally in free molecular regime! Friday, October 2, 2020 20 20
- Slides: 20