Compendium

Atmospheric Chemistry and Radiative Transfer

14 min read

Atmospheric Chemistry and Radiative Transfer

A Tier 2 climate specialty covering the chemistry of Earth’s atmosphere — composition, reactive species, ozone, aerosols, cloud microphysics — and how electromagnetic radiation propagates through it (radiative transfer, the physics underlying climate sensitivity, satellite remote sensing, and proposed geoengineering interventions). The two are inseparable: the molecules that absorb and emit radiation are products of atmospheric chemistry, and the radiation field drives photochemistry.

Atmospheric structure

Vertical layering by temperature gradient:

  • Troposphere (0–12 km mid-latitudes; deeper in tropics ~17 km, shallower at poles ~8 km). Contains ~75% of atmospheric mass; site of weather; temperature decreases with altitude at lapse rate ~6.5 K/km (ICAO standard) under typical conditions.
  • Tropopause — temperature inversion boundary; cold trap that limits stratospheric H2O.
  • Stratosphere (~12–50 km). Temperature increases with altitude due to O3 absorption of UV (Hartley and Huggins bands). Ozone layer peaks ~25 km. Brewer–Dobson circulation: tropical upwelling + poleward transport + polar downwelling.
  • Mesosphere (~50–85 km). Coldest layer (~180 K at mesopause); noctilucent clouds.
  • Thermosphere (85–500 km). Temperature rises with altitude due to EUV / X-ray absorption by O, N2; auroral phenomena.
  • Exosphere (>500 km). Density so low molecules can escape Earth’s gravity (Jeans escape).

Inversion layers (e.g., subtropical subsidence inversions, nocturnal radiative inversions) trap pollutants → urban smog events (Los Angeles, Mexico City, Beijing, Delhi).

Composition

Major dry-air constituents (mole fractions):

  • N2 78.084 %
  • O2 20.946 %
  • Ar 0.934 %
  • CO2 — May 2026 Mauna Loa monthly mean ~424 ppm and rising at ~2.5 ppm/yr; pre-industrial 280 ppm.
  • Ne 18.18 ppm; He 5.24 ppm; CH4 ~1925 ppb (2024; up from ~720 ppb pre-industrial; record-fast growth 2020-22); Kr 1.14 ppm; H2 ~0.55 ppm; N2O ~335 ppb; CO 90 ppb tropospheric.

Trace and highly variable: H2O (0–4 %; highly variable spatially); O3 0–10 ppm in stratosphere, 10–100 ppb troposphere; NOx, SO2, NH3, NMVOCs, particulates. See carbon-cycle-and-greenhouse-gases for detailed greenhouse-gas budgets.

Stratospheric ozone

Chapman cycle

Sydney Chapman 1930 — pure-oxygen photochemistry of stratospheric ozone:

  • O2 + hν (λ < 240 nm) → 2 O.
  • O + O2 + M → O3 + M (three-body recombination; M = N2 or O2 as third body).
  • O3 + hν (200–320 nm; Hartley band) → O2 + O(¹D) or O2 + O.
  • O + O3 → 2 O2.

Steady-state predicts ~10 ppm peak O3; actual observations are lower because of additional catalytic cycles.

Catalytic ozone destruction

A catalyst X reacts with O3 → XO + O2; XO + O → X + O2 → net loss of O3 + O without consuming X. Paul Crutzen 1970 QJRMS — NOx (NO/NO2) catalytic cycle from N2O oxidation; concerns about SST stratospheric aircraft (Concorde, never deployed in feared numbers). Mario Molina + F. Sherwood Rowland 1974 Nature — “Stratospheric sink for chlorofluoromethanes: Chlorine atom-catalysed destruction of ozone” — CFCs (CCl2F2, CCl3F) photolyze in stratosphere releasing Cl atoms; ClOx catalysis. Br catalysis (from halons, methyl bromide) is ~50× more efficient than Cl per atom.

1995 Nobel Chemistry — Crutzen, Molina, Rowland “for their work in atmospheric chemistry, particularly concerning the formation and decomposition of ozone.”

Antarctic ozone hole

Farman, Gardiner, Shanklin 1985 Nature “Large losses of total ozone in Antarctica reveal seasonal ClOx/NOx interaction” — British Antarctic Survey Halley station documented dramatic springtime O3 column declines beginning ~1979. Mechanism:

  1. Polar vortex isolates Antarctic stratosphere through winter darkness.
  2. Polar stratospheric clouds (PSCs) form at < 195 K — type I nitric acid trihydrate NAT, type II water ice.
  3. Heterogeneous surface reactions activate chlorine reservoirs: ClONO2 + HCl → Cl2 + HNO3; ClONO2 + H2O → HOCl + HNO3.
  4. Polar sunrise (Aug-Sep) photolyzes Cl2, HOCl → reactive Cl.
  5. ClO + ClO + M → Cl2O2 + M → Cl2O2 + hν → 2 Cl + O2 → catalytic O3 loss.

Peak depletion late September to mid-October; recovery begins as vortex breaks down and warmer air mixes in. Vostok and Halley continue routine balloon ozone soundings.

Montreal Protocol

1987 Montreal Protocol plus amendments — London 1990, Copenhagen 1992, Vienna 1995, Montreal 1997, Beijing 1999, Kigali 2016 (HFCs). CFC phase-out completed in developed countries by 1996, developing countries by 2010. HCFCs phase-out 2030 developed / 2040 developing. Kigali Amendment phases down HFCs (high-GWP refrigerants, not ozone-depleting but climate-relevant). Often cited as the most successful international environmental treaty.

WMO/UNEP Scientific Assessment of Ozone Depletion 2022 (quadrennial) — ozone layer on track to recover to 1980 baseline: midlatitudes ~2040, Arctic ~2045, Antarctic ~2066. Recent 2020 and 2023 large ozone holes attributed to dynamical variability (vortex strength) and possibly Hunga Tonga eruption Jan 2022 stratospheric water vapor injection.

Tropospheric chemistry

OH — the atmospheric detergent

Hydroxyl radical OH is the primary oxidant initiating degradation of most reduced trace gases (CH4, CO, NMVOCs). Daytime steady-state concentration ~10⁶ molecules/cm³; lifetime ~1 second.

Primary OH production: O3 + hν (λ < 320 nm) → O(¹D) + O2; O(¹D) + H2O → 2 OH.

Photochemical smog

Haagen-Smit 1952 Ind Eng Chem elucidated Los Angeles smog as photochemical O3 formation from NOx + VOC + UV:

  • NO2 + hν (λ < 420 nm) → NO + O.
  • O + O2 + M → O3 + M.
  • NO + O3 → NO2 + O2 (the cycle alone doesn’t yield net O3).
  • RH + OH → R + H2O.
  • R + O2 → RO2.
  • RO2 + NO → RO + NO2 (the key step — converts NO to NO2 without consuming O3, enabling net O3 buildup).
  • HO2 + NO → OH + NO2 (regenerates OH).

EPA NAAQS 8-hour O3 standard 70 ppb (revised 2015). NOx control + VOC control depending on local chemistry regime (NOx-limited vs VOC-limited; determined by HCHO/NO2 ratio observable from satellite — TROPOMI, GEMS, TEMPO).

Methane oxidation

CH4 + OH → CH3 + H2O; CH3 + O2 + M → CH3O2; CH3O2 + NO → CH3O + NO2; CH3O + O2 → HCHO + HO2; HCHO + hν → HCO + H or H2 + CO; HCO + O2 → CO + HO2. Net global CH4 lifetime ~ 9 years (tropospheric ~ 12 years if you include stratospheric and soil sinks). Cl-radical sink in marine boundary layer (Hossaini, Atlas).

Mercury cycle

Hg⁰ atmospheric (~1 year residence) → oxidized to Hg²⁺ by O3, OH, Br radicals (Mercury Atmospheric Depletion Events MDEs in polar spring) → wet/dry deposition → methylated to MeHg in sediments by SRB sulfate-reducing bacteria → bioaccumulation in fish. Minamata Convention 2013 (entered force 2017) regulates Hg emissions globally.

Sulfur cycle

SO2 + OH + M → HSO3 + M → H2SO4; rapid oxidation in cloud water via H2O2 and O3 (pH-dependent). H2SO4 + H2O → sulfate aerosol → CCN, acid rain. DMS (dimethyl sulfide) from marine plankton → ocean-atmosphere sulfur cycle (CLAW hypothesis Charlson-Lovelock-Andreae-Warren 1987 — partially supported, contested).

Acid rain history

Coal combustion → SO2 + NOx → H2SO4 + HNO3 → acid deposition. Adirondacks, Black Forest, Scandinavia 1960s-80s damage; US Clean Air Act 1990 SO2 cap-and-trade; rapid decline; current attention on China where SO2 also declining rapidly since 2010s.

Aerosols and clouds

Sources

  • Natural — sea salt (marine boundary layer); mineral dust (Sahara, Gobi, dust storms); biomass burning (forests, savannas, peatlands; record 2023 Canadian fires emitted ~640 Mt CO2); volcanic SO2 and ash; biogenic SOA from terpenes / monoterpenes / sesquiterpenes / isoprene.
  • Anthropogenic — SO2 (coal, oil), NOx (combustion), NH3 (livestock, fertilizer), VOCs (industrial, transport), primary organic carbon, black carbon (incomplete combustion, especially diesel and biomass).

Size distribution

  • Nucleation mode 1-10 nm (newly formed clusters from H2SO4 + NH3 + amines; Kulmala SMEAR-II Hyytiälä Finland; CLOUD CERN chamber experiments).
  • Aitken mode 10-100 nm.
  • Accumulation mode 100-1000 nm — dominates aerosol number and mass; primary CCN; main optical scatterer at solar wavelengths.
  • Coarse mode > 1 µm — sea salt, dust.

Optical properties

  • Optical depth τ_aer — column integral of extinction; AERONET sun-photometer network (Holben et al. 1998; ~600 stations globally) ground-truths satellite retrievals (MODIS Terra/Aqua, VIIRS, MISR, SLSTR Sentinel-3).
  • Single-scattering albedo SSA = scattering / (scattering + absorption). SSA ~ 0.95 for sulfate; ~0.2-0.3 for fresh BC; ~0.85-0.95 for dust.
  • Asymmetry parameter g ~ 0.6-0.7 forward-peaked for typical accumulation mode.
  • Phase function — angular distribution of scattered intensity; Henyey-Greenstein or full Mie computation.

Radiative forcing of aerosols

  • Direct effect — scattering cools, absorption warms. Sulfate aerosols globally cooling; BC absorbing → warming aloft.
  • First indirect (Twomey 1974) — more CCN → more droplets → smaller droplets at fixed liquid water → higher cloud albedo → cooling.
  • Second indirect (Albrecht 1989) — more CCN → smaller droplets → suppressed precipitation → longer cloud lifetime → more cooling.
  • Net total aerosol forcing ~ −1.1 W/m² (IPCC AR6; large uncertainty −0.4 to −2.0); offsets ~30 % of CO2 forcing currently. Aerosol cleanup (e.g., 2020 IMO shipping fuel sulfur cut from 3.5% to 0.5%) measurably reduced cooling — “unmasking” warming visible in 2023-24 SST anomalies (Hansen 2023; Quaas 2024).

Cloud microphysics

  • Warm-cloud activation — Köhler theory; CCN spectra.
  • Mixed-phase clouds — Wegener-Bergeron-Findeisen process: at T < 0 °C ice + supercooled droplets coexist; saturation vapor pressure over ice < over liquid → ice grows at droplet expense.
  • Ice nucleation — homogeneous freezing < −38 °C; heterogeneous via INPs (ice-nucleating particles) — dust, biological (Pseudomonas syringae INP; Brennan-Boucher review); important for mixed-phase clouds, cirrus, polar.
  • Cirrus — high ice clouds; radiative effect net warming (greenhouse > albedo).
  • Contrails — aviation; line-shaped ice clouds → cirrus; ~50% of aviation’s effective climate forcing per Lee et al. 2021. CCT (cirrus cloud thinning) proposed as targeted climate intervention.
  • Convection — deep cumulonimbus + tropical convective systems; key control on tropical hydrological cycle.

Major field campaigns and networks

  • AGAGE — Advanced Global Atmospheric Gases Experiment (Prinn MIT 1978+) — high-frequency measurements of halocarbons + N2O + CH4 at Mace Head Ireland, Trinidad Head California, Cape Grim Tasmania, Ragged Point Barbados, Cape Matatula Samoa, Mt Cimone Italy, Zeppelin Svalbard, Jungfraujoch.
  • NOAA GML — Global Monitoring Laboratory (Boulder) — Mauna Loa CO2 since 1958 (Charles Keeling).
  • NCAR — National Center for Atmospheric Research (Boulder).
  • CARES, IMPROVE US air-quality and visibility networks.
  • EBAS European aerosol network.
  • TROPOMI ESA Sentinel-5P (since 2017) — daily global NO2, HCHO, SO2, CH4 at ~5.5 km.
  • TEMPO NASA (geostationary 2023) — hourly North America NO2/HCHO/O3.
  • GEMS South Korea (geo Asian).

Radiative transfer

Solar input

Total solar irradiance TSI = 1361 W/m² (SORCE, TIM Greg Kopp 2008 — earlier values were overestimates); 11-year cycle amplitude ±0.1 % (~1 W/m² in TSI; ~0.2 W/m² globally averaged).

Earth’s planetary albedo ~0.3; absorbed shortwave ~240 W/m² (= 1361/4 × 0.7). Equilibrium blackbody temperature T_eff = (240/σ)^{1/4} = 255 K; observed surface mean ~288 K → 33 K natural greenhouse effect; ~0.8-1.2 K of which is anthropogenic warming since pre-industrial.

Beer–Lambert–Bouguer law

I(λ, s) = I₀(λ) exp(−τ(λ, s)) where τ is optical depth. Differential absorption: dτ = σ(λ) n ds where σ is cross-section, n number density.

Multi-stream solutions

  • Two-stream Eddington approximation — upward and downward azimuthally-averaged streams; computationally cheap; standard for global models.
  • DISORT — Discrete Ordinates Radiative Transfer (Stamnes, Tsay, Wiscombe, Jayaweera 1988) — n-stream; gold standard for benchmarking.
  • Four-stream + eight-stream for clouds + aerosols.
  • Monte Carlo for 3D inhomogeneous (cloud horizontal photon transport; Marshak, Davis).

Spectral databases

  • HITRAN (Rothman, Gordon; Harvard-Smithsonian + IAO Tomsk) — ~5 M lines for atmospheric molecules + line shapes (Voigt, Lorentz, Doppler, line mixing). 2020 release standard reference. HITEMP extends to high temperatures (planetary, combustion).
  • GEISA — French equivalent.
  • ExoMol — Tennyson UCL; exoplanet atmospheres, broader T range, water hot line lists, high-temperature molecules.
  • MODTRAN (Spectral Sciences / AFRL) — moderate-resolution atmospheric transmission; widely used for sensor mission planning.
  • LBLRTM (AER / Clough) — line-by-line; reference for satellite forward models.
  • RRTMG / RRTM (Mlawer AER) — Rapid Radiative Transfer Model GCM-grade broadband.

Cloud and aerosol RT

  • Mie theory (Mie 1908) — exact for homogeneous spheres; tabulated for aerosol and cloud droplet sizes.
  • T-matrix (Mishchenko) — non-spherical particles; ice crystals, dust.
  • Ray-tracing for ice habits (Yang, Liou) — plates, columns, bullet rosettes, aggregates.

Active and passive remote sensing

  • Sun photometers — AERONET, MAN (Maritime Aerosol Network).
  • Ground LIDAR — MPLNET, EARLINET; vertical aerosol profiles.
  • Spaceborne LIDAR — CALIPSO (2006-2023; CALIOP); CATS (ISS 2015-17); ICESat-2 ATLAS (primarily ice/topography). EarthCARE (ESA/JAXA, launched May 2024) carries ATLID lidar + CPR cloud radar + MSI imager + BBR broadband radiometer — first joint cloud-aerosol space mission with all four instruments.
  • Spaceborne radar — CloudSat (2006-2023; 94 GHz); EarthCARE CPR Doppler.

Climate sensitivity and feedbacks

Sherwood, Webb, Annan et al. 2020 Rev Geophys “An assessment of Earth’s climate sensitivity using multiple lines of evidence” — narrowed equilibrium climate sensitivity (ECS) to 2.6–3.9 K (66 % range) by combining historical, paleoclimate, and process evidence. IPCC AR6 (2021) WG1 adopted ECS very likely 2–5 K, likely 2.5–4 K, best estimate 3 K.

Feedback partition (linear approximation; W/m²/K):

  • Planck −3.2 (warmer Earth radiates more).
  • Water vapor +1.5 (Clausius-Clapeyron; warmer atmosphere holds more H2O which is a GHG).
  • Lapse rate −0.5 (tropics warm aloft more than surface → enhanced OLR).
  • Cloud +0.4 ± 0.5 — largest uncertainty; high clouds + warming (greenhouse), low clouds − warming (albedo); shallow boundary-layer cumulus / stratocumulus over E ocean basins (Peru-Chile, California, Namibia, Canary) is the critical region.
  • Surface albedo +0.4 — ice + snow loss.

Cloud feedback uncertainty

Sc (stratocumulus) decks are particularly contested — high-resolution LES (large-eddy simulations) at ~25-100 m grid spacing resolve marine boundary layer turbulence and cloud-top entrainment instability (Stevens, Bretherton). LES-derived emergent constraints (Sherwood et al. 2014 Nature; Brient-Schneider 2016) suggest positive Sc feedback. Schneider’s CliMA project (Caltech) targets ML-augmented Earth system models trained on LES.

Geoengineering chemistry

Stratospheric aerosol injection (SAI)

Crutzen 2006 Climatic Change “Albedo enhancement by stratospheric sulfur injections: A contribution to resolve a policy dilemma” — proposed mimicking Mt Pinatubo 1991 eruption (20 Mt SO2 → ~0.5 °C cooling for 2 years; Hansen 1992). Sustained 8-15 Mt SO2/year via aircraft or balloon to lower stratosphere offsets ~1-1.5 W/m² forcing.

Side effects: termination shock (rapid warming if injection stops); ozone chemistry (heterogeneous reactions on sulfate enhance ClO/Br release in cold regions); monsoon disruption (especially African and Asian); acid deposition (small fraction of natural rainfall acidity); regional precipitation effects.

Programs: SCoPEx Harvard (Keutsch, Keith) — outdoor balloon test, cancelled in 2024 after governance and stakeholder concerns; GeoMIP (Geoengineering Model Intercomparison; Kravitz, Robock) — CMIP modeling experiments.

Commercial / activist actors: Make Sunsets (Luke Iseman 2022-) — sulfur-balloon releases ~ USD 1 M raised; controversial; lacks scientific protocol. Stardust Solutions (Israel) — outdoor SAI experiments planned.

Marine cloud brightening (MCB)

Sea-spray nozzles increase low-cloud CCN → albedo increase. John Latham 1990 original proposal. Field experiments off Great Barrier Reef (Reef Restoration and Adaptation Program) and Bay Area pilot 2024 (UW MCB).

Cirrus cloud thinning (CCT)

Mitchell 2009 — seeding cirrus with INPs (dust, BiI3) to reduce ice number, larger ice → faster fall + sublimation → reduced cirrus coverage → reduced warming. Modeling suggests potential ~−1 W/m² but ice nucleation parameterizations highly uncertain.

Carbon dioxide removal (CDR)

Distinct from solar radiation management — DAC (direct air capture; Climeworks, Carbon Engineering / 1PointFive), enhanced weathering (basalt; Project Vesta, Lithos Carbon), ocean alkalinity enhancement, BECCS, biochar. Cross-link climate-mitigation-and-adaptation for the broader CDR landscape.

Adjacent

Graph View