Spotlight #31 – Climate Science: Radiative Forcing, Feedback Loops, Carbon Cycle and Tipping Points

Climate science applies the same physical laws as the rest of this blog — radiative transfer, fluid dynamics, thermodynamics, geochemistry — to the system that most directly affects human civilisation. The atmosphere is a thin shell (99% of mass below 30 km on a 6 371 km radius planet) that traps infrared radiation, redistributes heat through giant convective cells, and exchanges carbon with the ocean and land biosphere on timescales ranging from days to millions of years.

1. Radiative Forcing and the Greenhouse Effect

The climate system is ultimately driven by the imbalance between incoming solar shortwave radiation and outgoing longwave radiation emitted by Earth. The greenhouse effect — absorption and re-emission of infrared radiation by atmospheric gases — raises Earth’s surface temperature from a frozen −18 °C (bare-planet blackbody equilibrium) to the observed +15 °C. Any external perturbation that alters this balance is a radiative forcing.

Zero-dimensional energy-balance model:
  C dT/dt = S_0(1 − α)/4 − εσT&sup4;
  S_0 = 1361 W/m² (solar constant)
  α = 0.30 (planetary albedo)
  ε ≈ 0.78 (effective emissivity; <1 due to greenhouse effect)
  σ = 5.67 × 10^−8 W m^−2 K^−4 (Stefan–Boltzmann)

Equilibrium temperature (set dT/dt = 0):
  T_eq = [S_0(1−α)/(4εσ)]^{1/4}
  Without greenhouse (ε=1): T_eq = 255 K = −18 °C
  With real atmosphere (ε=0.78): T_eq ≈ 288 K = +15 °C
  Greenhouse warming contribution: ~33 °C

Radiative forcing ΔF [W/m²]:
  An externally imposed W/m² of imbalance before feedbacks act.
  CO&sub2; doubling:   ΔF = 5.35 ln(C/C_0) ≈ 3.7 W/m²  (IPCC best estimate)
  Pre-industrial CO&sub2; = 280 ppm; 2024 value = 422 ppm → ΔF ≈ 2.2 W/m²
  Equilibrium climate sensitivity (ECS): ΔT = λ ΔF
  λ ≈ 0.8 K/(W/m²) (IPCC AR6 best estimate)
  ECS = λ · 3.7 ≈ 3 °C per CO&sub2; doubling (likely range 2.5–4 °C)

Other forcing agents (AR6 values, W/m²):
  CH&sub4;:              +0.54
  N&sub2;O:              +0.28
  Halocarbons:      +0.36
  Aerosols (total): −1.06 (cooling)
  Land-use change:  −0.20
  Total anthropogenic 1750–2019: +2.72 W/m²

2. Climate Feedbacks — Amplifiers and Stabilisers

Feedbacks amplify or dampen the initial forcing response. The climate system has both positive (amplifying) and negative (stabilising) feedbacks. The net feedback determines the equilibrium climate sensitivity. IPCC AR6 (2021) identified the Planck feedback as the dominant stabilising mechanism, ice-albedo and water-vapour feedbacks as the dominant amplifiers, and cloud feedbacks as the largest source of uncertainty in climate projections.

Feedback framework:
  ΔT = ΔF / (−λ^−1)  where λ = λ_Planck + Σ_i f_i
  Positive f_i → amplification; negative f_i → damping.

Planck feedback (blackbody stabiliser):
  f_Planck = −d(4εσT³)/dT ≈ −3.3 W m^−2 K^−1  (always negative — stabilising)
  Physical mechanism: warmer planet radiates more, restoring balance.

Water-vapour feedback:
  f_WV ≈ +1.8 W m^−2 K^−1   (amplifying)
  Mechanism: warmer air holds more water vapour (Clausius–Clapeyron);
  H&sub2;O is a greenhouse gas → more warming → more H&sub2;O (amplifying loop).
  Approximately doubles the Planck feedback sensitivity alone (ECS goes from ~1 °C to ~2 °C).

Ice–albedo feedback:
  f_ICE ≈ +0.3 W m^−2 K^−1   (amplifying)
  Mechanism: warming melts ice → darker ocean/land exposed (α from 0.9 to 0.06)
  → more absorption → more warming. Particularly strong in the Arctic.

Lapse-rate feedback:
  f_LR ≈ −0.4 W m^−2 K^−1   (weakly stabilising in tropics, amplifying at poles)
  Tropical upper troposphere warms faster than surface → increases OLR.
  Polar: little lapse-rate feedback → polar amplification (Arctic warms 3× global average).

Cloud feedbacks (largest uncertainty):
  IPCC AR6 likely range: −0.2 to +0.8 W m^−2 K^−1
  Low marine clouds (stratocumulus): decrease with warming → positive feedback
  High cirrus clouds: complex; net sign uncertain
  Cloud feedback uncertainty spans ~1 °C of ECS uncertainty.

Net feedback and ECS:
  Net λ = f_Planck + f_WV + f_LR + f_ICE + f_cloud ≈ −1.3 W m^−2 K^−1
  ECS = 3.7 / 1.3 ≈ 2.85 °C per CO&sub2; doubling (IPCC: likely 2.5–4.0 °C)

3. The Global Carbon Cycle

Carbon cycles between the atmosphere, ocean, land biosphere, and geological reservoirs on timescales from days (photosynthesis/respiration) to millions of years (rock weathering and volcanism). Anthropogenic emissions (~11 GtC/yr in 2023) are perturbing the fast cycle at a rate roughly 100 times faster than any geological carbon injection event in the past 66 million years.

Major reservoirs (GtC = gigatonnes of carbon):
  Atmosphere:                    880 GtC   (2023; pre-industrial 600 GtC)
  Terrestrial biosphere (live):  600 GtC
  Soil organic carbon:         1 500 GtC
  Permafrost (CH&sub4;+CO&sub2;):       1 700 GtC   (vulnerble under warming)
  Surface ocean:                 900 GtC
  Deep ocean:                 37 000 GtC
  Lithosphere (sediments):   5 000 000 GtC  (extremely slow exchange)

Current annual fluxes (GtC/yr, approximate 2020s):
  Fossil fuel + cement emissions:    +10.5
  Land-use change:                   +1.2
  Terrestrial uptake (net):          −3.1
  Ocean uptake (net):                −3.0
  Atmospheric accumulation:          +5.6  (about 51% of emissions remain airborne)

Ocean carbon uptake mechanisms:
  1. Solubility pump: CO&sub2; dissolves in cold surface water (Henry’s law: s ∝ 1/T)
  2. Biological pump: phytoplankton fix CO&sub2; → organic carbon sinks as dead cells
     Efficiency < 25% (most organic matter remineralised before reaching seafloor)
  Ocean acidification: pH fell from 8.25 (pre-industrial) to 8.06 (2023)
  ΔpH = −0.19 → [H&sup+;] increased by 55%  (logarithmic pH scale)

Atmospheric CO&sub2; lifetime:
  Pulse fraction still airborne after:
    10 yr:     ~64%
   100 yr:     ~40%
  1 000 yr:    ~15–25%
  10 000+ yr:  ~5–10%
  No single timescale: ocean equilibration (centuries), weathering (tens of millennia).

4. Tipping Elements — Non-Linear Transitions

Some components of the climate system have tipping points: thresholds beyond which a small additional forcing triggers a self-reinforcing transition to a qualitatively different state. Unlike linear feedbacks, tipping elements may be irreversible on human timescales — once crossed, even reducing emissions does not immediately restore the previous state. Tipping cascades, where one tipping element triggering another, are a major concern in Earth system science.

McKay et al. (2022) assessment of tipping element thresholds (global warming above pre-industrial):

Cryosphere:
  West Antarctic Ice Sheet (WAIS): 1.5–3 °C  (collapse over centuries; 3–5 m sea level rise)
  Greenland Ice Sheet:             0.8–3 °C  (nearly irreversible beyond 1.5 °C)
  Arctic sea-ice perennial loss:   ~1.5–2 °C  (seasonal first, perennial summer loss next)
  Mountain glacier TP:             ~2 °C        (commitment to near-total loss; water security)

Ocean circulation:
  AMOC collapse (Atlantic thermohaline):  3–5 °C  (may already be destabilising)
  Abrupt AMOC weakening brings cooling to N. Europe, strengthens Sahel drought.

Ecosystems:
  Amazon dieback:                  3.5 °C  (drought + deforestation interaction; 17% reached)
  Boreal forest northward shift:   4 °C
  Coral reef bleaching (>90% loss): ~1.5–2 °C  (already 67% experienced mass bleaching 2023)

Permafrost methane:
  Boundary poorly constrained; gradual permafrost thaw may not be a hard tipping point
  but adds significant (0.1–0.4 GtC/yr additional by 2100 under high scenarios)

Tipping cascade risk:
  Lenton et al. (2019); Armstrong McKay et al. (2022): crossing one TP at 2 °C
  roughly doubles probability of triggering another within decades.
  At 2 °C: ~9 TPs with thresholds below 2 °C may be approaching or exceeded.

5. Atmospheric Circulation — Hadley, Ferrel and Polar Cells

The differential solar heating between equator and poles drives a global atmospheric circulation that redistributes heat poleward. The Coriolis effect from Earth’s rotation breaks the single-cell convection expected from pure thermal drive and creates three meridional circulation cells in each hemisphere: Hadley, Ferrel, and Polar cells. These cells set the large-scale precipitation patterns — tropical rain belts, subtropical deserts, temperate storm tracks, and polar high-pressure zones.

Meridional circulation cells:
  Hadley cell (0–30°):
    Rising branch: ITCZ (Inter-Tropical Convergence Zone) near equator
    Descending branch: subtropical high-pressure belts (Sahara, Arabian, Atacama deserts)
    Trade winds: easterlies at surface, driven by poleward-deflected air (Coriolis)
  Ferrel cell (30–60°):  indirect thermally-driven cell; westerlies at surface
  Polar cell (60–90°):  cold descending air at poles; easterlies at surface

Potential temperature and CAPE:
  Hadley rising branch: CAPE ~ 1 000–3 000 J/kg (tropical convection energy)
  Moist Hadley cell: Held–Suarez scaling → cell poleward boundary ∝ T_s / Ω  (Ω = Earth’s rotation rate)

Jet streams:
  Subtropical jet: ~12 km altitude, ~30–35° latitude, v ~ 60 m/s
  Polar front jet: ~9 km, ~55–65°, v ~ 30–100 m/s (westerlies, storm tracks)
  Thermal wind balance: ∂u/∂z = −(g/fT)(∂T/∂y)   (f = Coriolis factor)
  Warming stratigy (Arctic amplification) → reduced ∂T/∂y → weaker jet → wavier, slower meandering

Hadley cell expansion (poleward shift under warming):
  Observed: ~1–2° latitude shift per °C of global warming
  Consequence: subtropical dry zones expand poleward (Mediterranean, SW USA, SW Australia drying)

ENSO (El Niño–Southern Oscillation):
  Walker circulation in tropics: East-to-west along equatorial Pacific
  El Niño: weakened trades → warm pool moves east → global teleconnections
  ΔT (Niño-3.4 region) > +0.5 °C for 5 consecutive months = El Niño event
  PDO and AMO (decadal oscillations) modulate ENSO amplitude and global temperature trend.

6. Climate Models — From Zero-D to Global Coupled GCMs

Climate models span a hierarchy from the simple zero-dimensional energy-balance model (one equation, one variable) to fully coupled general circulation models (GCMs) with hundreds of millions of grid cells in atmosphere, ocean, land and sea ice. Each rung of the hierarchy adds physical processes at the cost of computational expense and structural uncertainty.

Model hierarchy:
  0-D EBM:       C dT/dt = F_in − F_out   (teach concepts; fast)
  1-D EBM:       adds latitude; meridional heat transport Q = −D d(T)/dφ
  2-D RCM:       radiative-convective model; vertical structure; atmosphere columns
  AOGCM:         coupled atmosphere-ocean GCM (CMIP5/6 models)
  ESM:           Earth System Model = AOGCM + interactive carbon cycle, chemistry, etc.

Horizontal resolution (typical CMIP6):
  Atmosphere: 100–200 km (hindcast), 25–50 km (HighResMIP)
  Ocean:      25–100 km (eddy-parameterised), 10 km (eddy-resolving: very expensive)
  Sub-grid processes (cloud microphysics, convection, turbulence) are parameterised.

CMIP6 model ensemble (Zelinka et al. 2020):
  ~ 40 models from 20+ institutions worldwide
  Historical warming 1850–2014: all models reproduce observed record within ~0.2 °C
  ECS range in CMIP6: 1.8–5.6 °C (wider than CMIP5: 2.1–4.7 °C)
  High-ECS models (above 5 °C) now assessed as unlikely (IPCC AR6 constraints).

IPCC AR6 (2021) projections for 2100 vs 1850–1900 baseline:
  SSP1-1.9 (aggressive mitigation): +1.0 – +1.8 °C
  SSP2-4.5 (intermediate):          +2.1 – +3.5 °C
  SSP5-8.5 (fossil-fuel intensive):  +3.3 – +5.7 °C
  Sea level rise (likely range):     0.32–1.01 m (SSP1–SSP5)
  Note: 2024 observed warming ~ +1.45 °C above pre-industrial average.