What you need to know
- 5.19(i)Data can be extrapolated to make predictions used in models of future climate change.
- 5.19(ii)Models for climate change have limitations.
You do not need named IPCC scenarios (RCP/SSP) as recall. If a graph labels them, read the axes. 5.25 belongs with who commissioned the run, not with the physics of a grid cell.
By the end, you should be able to…
- Describe extrapolationExtending a trend beyond the range of data actually collected, to make a prediction. It is less reliable than the measured range. as extending a pattern beyond the measured range.
- State that such projections are inputs to climate models.
- Give several limitations of those models.
Build your understanding
Beyond the last data point
To extrapolate is to extend a relationship beyond the range of the measurements. If emissions have risen, a simple extrapolationExtending a trend beyond the range of data actually collected, to make a prediction. It is less reliable than the measured range. might assume they keep rising at the same rate. That is a prediction, not a measurement. Different assumed future emissions give different future concentrations.
Climate models use those assumed futures, plus physics of radiation, water, ice and carbon, to project temperature, rainfall and sea ice. NASA describes models as mathematical representations on a 3-D grid, because we do not have a spare Earth to experiment on. The output is a range, not a single promised weather forecast for 12 June 2084.
Why the range is wide
Future human behaviour is unknown: that is not a failure of physics. Cloud feedbacks, aerosols, ice-sheet dynamics and regional rainfall are harder than the global greenhouse calculation. Incomplete data — sparse oceans in older records, uncertain paleoclimate calibrations — feed into the starting state. Grid cells smear local mountains and storms.
A model that cannot resolve a thunderstorm can still be useful for global mean temperature. Matching the last century’s warming when greenhouse gasA gas, such as CO₂ or methane, that absorbs infrared radiation in the atmosphere. are included, and failing when they are left out, is a test. It is not a proof that every regional drought forecast is exact.
Using models without worshipping them
Limitations named in a 5.19 answer might include: unknown future emissions; incomplete data; processes that are simplified or missing; coarse spatial scale; difficulty with tipping points; and the fact that a projection is not an observation. “Models are only guesses, so ignore them” is the opposite error: they are constrained tools with stated uncertainty.
If two groups publish different warming ranges, ask which emissions pathway, which climate sensitivity, and who funded the work (5.25). Disagreement about the third decimal place is not disagreement that extra greenhouse gases warm the surface.
| Idea | What it is | Limitation |
|---|---|---|
| Measurement | CO₂ or temperature already recorded | Does not, by itself, give 2100 |
| Extrapolation | Extending a trend or relationship | The future drivers may change |
| Climate model | Physics plus assumed futures on a grid | Simplified processes, coarse scale, unknown emissions |
Make your reasoning visible
Name two limitations, not one shrug
“Unknown emissions” and “incomplete cloud data” score as biology; “models are fake” does not.
Keep weather out of it
A five-day forecast failing is not 5.19.
Avoid this misconception
The claim: If one winter is colder than the model’s mean, the model is worthless.
The correction: Climate projections are about long-term statistics. Year-to-year weather still varies around a shifting mean.
How warm will it be in 2100?
Nobody can measure the future. Climate scientists extend today’s trends and use physics to project what might happen. How much can we trust the answer?
What the model shows
Measured CO₂ and warming since 1850, an extrapolated future, and a simple model linking the two, with a shaded range.
What to look for
Whether the model matches the past, and which choices change the 2100 answer the most.
What it tells us
Models turn assumed futures into projections. The answer is a range, because emissions, sensitivity and some processes are uncertain.
Be careful what you claim
Only CO₂; a two-layer model; illustrative futures.
Try it in the model: set Extra CO₂ since 1850 to Left out.
Projecting warming to 2100
Test a simple climate model against the measured warming, then extrapolate CO₂ to 2100 under different assumed futures and see how sensitivity and a missing feedback change the range. Measured values are approximate; futures are illustrative.
What you see
- A graph of warming above 1850–1900 from 1850 to 2100: the measured warming (dots), the model (line) and the likely range for the chosen future (shaded).
- A graph of CO₂ concentration: measured to 2025, then extrapolated (dashed). The chain at the bottom links CO₂ in 2100 to forcing and to warming.
How to use it
- Choose an assumed future for CO₂, set the climate sensitivity, leave the extra CO₂ out to test the model, and add a carbon-cycle feedback.
What it shows
- Measured data can be extrapolated to give assumed futures, which are fed into climate models to project temperature (5.19(i)).
- Projections depend on unknown future emissions, uncertain sensitivity and simplified or missing processes, so models give a range (5.19(ii)).
Limits of the model
- Measured CO₂ and warming are approximate rounded values. Forcing uses 5.35 × ln(C ÷ 285) W m⁻²; warming uses two layers (upper ocean and land, and deep ocean). Only CO₂ is modelled: other greenhouse gases and aerosols are left out, and they roughly offset each other in the past.
- The futures and the feedback strength are illustrative, not named IPCC scenarios. There is no weather, no year-to-year variation and no regional detail: a real model works on a 3-D grid.
Guided tour · step 1 of 8
- Compare the model line with the measured warming to 2020.
- In Model, just below these steps, set Extra CO₂ since 1850 to Left out.
With the measured CO₂ rise included, the model follows the measured warming to within about 0.1 °C. Leave the extra CO₂ out and it stays flat and misses the warming completely. Matching the past is a test of the model, not a proof of every detail of its future.
Model
Questions
Check your understanding
Try each question before opening the suggested answer.
Practice question
Retrieval and explanation · suggested answer
What does it mean to extrapolate climate data?
Show suggested answer
To extend a measured relationship or trend beyond the range of the observations to make a prediction.
Practice question
Retrieval and explanation · suggested answer
How are such predictions used in climate models?
Show suggested answer
Assumed future emissions or concentrations are fed into models that then project temperature, rainfall and other variables.
Practice question
Retrieval and explanation · suggested answer
Give two limitations of climate models.
Show suggested answer
Any two of: unknown future emissions; incomplete data; simplified or missing processes (for example clouds or ice sheets); coarse grid; uncertainty about tipping points.
Take the idea further
Apply the same specification knowledge to an unfamiliar problem. Any extra assumptions are supplied in the question.
Stretch question
A newspaper prints one line: “up to 4 °C by 2100”. List three questions a 5.19 student should ask before using the figure.
Show a hint
Pathway, model, uncertainty.
Explore the reasoning
Which emissions pathway was assumed? Is 4 °C a mean, an upper bound, or a regional maximum? Which models and what uncertainty range? Also: compared with which baseline year?
Recall this lesson
Lesson recall questions
Question 1 of 5: List three reasons climate-change models have limitations.
Climate models and their limits
Practice question
Retrieval and explanation · suggested answer
List three reasons climate-change models have limitations.
Lesson: 5C 7: Models of Climate Change and Their Limitations
Try this from memory first.
Put your biology to work
Explore where these ideas appear in the world. Enrichment, not extra required recall.
A laboratory we cannot build in the field
5C · Models of Climate Change and Their LimitationsNASA’s public climate-model explainers treat models as simulations on a grid, used because we cannot turn greenhouse gases up on a spare planet. That is the 5.19 idea…Specification 5.19(ii)Read & use the scienceClose article
Why this matters
NASA’s public climate-model explainers treat models as simulations on a grid, used because we cannot turn greenhouse gases up on a spare planet. That is the 5.19 idea in institutional language. Supercomputer detail is enrichment; the specification still wants extrapolation and limitations. A model can reproduce some past patterns and still be uncertain about a particular region or future emissions. Scenarios are assumptions about human choices; they are not predictions that the choices have already happened.
Back to the specification
Pearson Edexcel IAL Biology · Issue 2, February 2021 · Unit 4, pp. 29–32. The application extends beyond required recall.
- 5.19(ii) understand that models for climate change have limitations
Use the science
Why is “we ran it on a supercomputer” not a reason to hide the limitations?
Explore a possible response
Computing power does not supply missing observations or known future politics. A finer grid can still be wrong if clouds or ice are simplified.
Explore the source
A climate model is tested against yesterday
5C · Models of Climate Change and Their LimitationsNOAA explains that climate models can be tested by seeing whether they reproduce historical patterns before using them with possible future emissions pathways. This process is called hindcasting.…Specification 5.19(i) · 5.19(ii)Read & use the scienceClose article
Why this matters
NOAA explains that climate models can be tested by seeing whether they reproduce historical patterns before using them with possible future emissions pathways. This process is called hindcasting. A model that matches a past global average can still have regional biases, especially for rainfall or clouds. Future scenarios also depend on human decisions that physics alone cannot predict. That is why a range of models and scenarios is more informative than one precise-looking number. Extrapolation is useful, but its limitations should travel with the result. A model is neither an untested guess nor an infallible forecast.
Back to the specification
Pearson Edexcel IAL Biology · Issue 2, February 2021 · Unit 4, pp. 29–32. The application extends beyond required recall.
- 5.19(i) understand that data can be extrapolated to make predictions and that these are used in models of future climate change
- 5.19(ii) understand that models for climate change have limitations
Use the science
Why might two valid models project different rainfall for one region even if both simulate past global warming well?
Explore a possible response
They may represent regional processes differently, and future rainfall also depends on scenario and natural variability.
Explore the source
- Climate modelsNOAA Climate.gov