WEEK 6: Predicting Future Climate Change
• Overall increase of 1.2°C since pre-industrial period, associated with variety of indicators
showing evidence of warming
• Particular focus on carbon dioxide emissions, contributing to greenhouse warming
• Attribution and detection methods lead us to conclude that 100% of temperature changes
are due to human activity
• How can we predict what to expect in the future?
o Empirical estimates and climate models
• Climate sensitivity: expressed as global temperature response to change in forcing
o Presented most regularly as response to a doubling of GHG forcing (+3.5 W/m2)
o Doesn’t necessarily have a huge practical purpose but it was used at the time, so
we continue to use it now
o This area is where there is still some debate in the literature
▪ Not that it will lead to warming, but how much warming
o Answers how sensitive the climate is to a change in radiative forcing
▪ Transient climate response and equilibrium
Transient Climate Response
• Transient climate response: measure of shorter-term response to
change in radiative forcing
• Only includes faster climate feedbacks i.e. shallow oceans, sea ice
• Originally calculated as change in global temperature for a 20-year
period at time of doubling of CO2
o Originally defined in terms of climate modelling so not very helpful
• Recent studies calculate it as the change in global temperature divided by the change in
radiative forcing over a period
o IPCC AR5 gives likely TCR range of ~0.27°C to ~0.69°C per W/m2
• Assumes a liner response between changes in temperature and radiative forcing
o Remember this method excludes slower feedbacks
o TCR = ∆T / ∆RF
▪ TCR = transient climate response
▪ ∆T = change in global temperature
▪ ∆RF = change in radiative forcing
• Whole host of climate forcings, can’t just consider GHG forcings for temperature (even
though they’re significant)
o Current estimate +2.58 W/m2
o TCR = 1.2 / 2.58 = 0.47°C W/m2
▪ Estimate right in the middle of the IPCC’s range, so it gives us some
confidence it’s pretty straightforward and matches with the best estimates
• Important to remember there is always a lag in the system due to energy imbalance
(approx. 0.6 W/m2 at earth’s surface)
o Maybe temperatures we’re currently seeing are reflective of changes minus the
imbalance amount
o Account for difference… TCR = 1.2 / 2.0 W/m2 = 0.6°C W/m2
▪ Subtracted 0.6 W/m2 imbalance estimate from values in equation
▪ This estimate keeps us on the higher range of the IPCC threshold
▪ We can pretty much rule out the lower end of that spectrum
• Applying this method to a doubling of GHG preindustrial forcing i.e. +3.6 W/m2
o 0.47°C W/m2 x 3.6 W/m2 = ~1.7°C
o 0.60°C W/m2 x 3.6 W/m2 = ~2.2°C
o ^Changes we expect to see for doubling of GHG forcing
o When will Earth reach this doubling point?
▪ Total radiative forcing to date is +2.6 W/m2
▪ Human radiative forcing currenting increasing at 0.44 W/m2 per decade
▪ By about 2043 Earth will reach +3.6 W/m2
• Summary: status quo emissions scenario
o Warming since preindustrial between around +1.7°C to +2.2°C by 2043
▪ Warming of 0.5°C to 1.0°C from now
Equilibrium Climate Sensitivity
• Equilibrium climate sensitivity: change in global surface temperature that will occur
once short- and long-term feedback processes reach
equilibrium; measure of long-term climate response to
change in radiative forcing
o Even though we might see transient climate response, expect climate to keep
warming slowly along longer scales
• IPCC AR5 gives likely ECS range of 1.5°C to 4.5°C for a doubling of CO2
o Much higher than TCR, showing we have better control over transient and we can
get substantial changes of really long time-periods
o Can further constrain ECS by looking at variety of different methods
▪ Usually best estimate over time is approx. +3°C
o Review paper by Sherwood et al. further narrowed range from IPCC numbers
(2.3°C to 4.5°C) -> More likely to be used in next IPCC report
o Scratch out 1.5°C lower bound… At least 2°C
o Helps reconcile with transient climate response, since short-term changes
sufficiently large enough that we know a lot of feedbacks are occurring
▪ If TCR is in upper 1 threshold, would expect ECS to be higher
• Linear association between ECS and TCR (x1.5 range)
Climate Modelling
• Discussed via attribution and detection, also considering past climate changes which can
be used as simulations, now discuss how they can be used in future
• Whole host of different groups around the world that do climate modelling
o Used to have just a few main models (mostly based in US and UK ~30 years ago)
o More global representation now, but difficult to keep people producing
comparable results
• Climate Model Intercomparison Project (CMIP)
o Goal to advance understanding of earth system
o Global coordination of climate modelling
o Define common experiment protocols, forcings and outputs
o Ensure inter-model comparability i.e. standardized scenarios