Modellergic – The reaction experienced by climate denialists whenever they encounter science containing the word ‘model’.
Nova rejects almost any science if it involves a “climate model”; a computer application performing billions of calculations in order to simulate how the planet responds to various scenarios such as increased greenhouse gases. Her “strawman argument” is that the models don’t get everything 100% correct, therefore they can’t ever be trusted. That’s an expectation that no climate scientist shares and even the great Deep Thought never stood a chance.
So are computer models good for anything?
Why are they always wrong?
There are many things in our environment that we cannot predict; systems so complex that we just cannot replicate.
Models can’t predict Volcanoes.
Models can’t predict the Solar cycle.
Models can’t predict the El Nino cycle.
Models can’t predict how much aerosols man will release in the future.
Models can’t predict how much CO2 man will release in the future.
Models can’t predict how much methane man will release in the future.
All of these are factors influence the surface temperature but are too complicated or not well enough understood to predict. The ENSO cycle is a great example, switching between El Nino and La Nina, it affects global temperatures dramatically on a short term basis (as seen below), but it can’t explain the long term trend.
Our ability to forecast the future state of the El Nino cycle is extremely limited. Various models trying to determine what the ocean might do, arrive at a range of outcomes as depicted in the following image.
The El Nino cycle adds a natural variation to the climate trend, with El Nino’s increasing the global surface temperature, whilst La Nina’s bring cooler global surfaces temperatures. These are not changes in the total energy of the planet, just cyclic movement of the energy through the ocean/air boundary.
Because of unpredictable changes, such as ENSO, Models will never get a projection 100% correct. Expecting such accuracy is to demonstrate a misunderstanding of the use of a climate model. Climate scientists are more concerned about the long term trend, rather than replicating short term variability.
How else to calculate?
What’s the alternative to climate models? How else should we calculate the impact of increasing a particular combination of gases and examine the impact on the planet?
Whilst Nova continues to criticise Models for their inaccuracies, she (and her climate denialist friends) have no approach for calculating attributions of recent warming. Jokingly, a few suggestions have been:
- Get a larger abacus.
- Examine the entrails of Schrödinger’s cat.
- Read tea leaves.
- Consult a Ouija board.
GCMs – How they work
Rather than simply shrug their shoulders, different climate science teams have worked to create numerous computer models such as the one described here. Over the years climate models have become ever more complex with greater amounts of detail.
Rather than wait to see how a model performs against future observations, we can run a model starting further back in time and compare it to the known observations. The results show models do better job of replicating the observations when GHG are included. The warming cannot be accounted for by natural causes alone.
Inaccuracy and uncertainty in the models is cause for concern, after all, in the absence of denialist tools, these are the only means by which we might forecast future changes to this planet. The theory of global warming doesn’t rely on models. Scientists know that greenhouse gases are rising, they know they trap heat and that from previous climate changes the planet has a climate sensitivity values around 3° for a doubling of CO2.
But models do get a number of items incorrect, but not always for the better. Sea level rise was recently announced to be occurring at 60% faster than the IPCC’s previous estimates.
Despite the Models not being perfect, they do actually perform reasonably well when some of the natural variability is taken into account. Rahmstorf, Foster, and Cazenave 2012 adjust for the effects of solar variability, volcanoes and ENSO and find much more agreement than Nova likes to admit.
Joanne Nova unreasonably expects perfection, criticises the Models on minor points, ignores the main features that they do replicate well and fails to deliver any calculations of her own.
Climate models aren’t perfect, but they sure beat Jo and her Ouija board.