AI and climate prediction

With the threat of climate change looming large, we will need to use every weapon in our arsenal to combat it. One of our newest is AI. While AI cannot be used to directly combat climate change, it can provide far more sophisticated predictions for future climate change and its effects, as well as measuring carbon emission, and calculating reduction strategies. AI is ideal for climate prediction, which is based on large meteorological data sets and then analysis of previous patterns. AI is excellent for predicting natural hazards, especially hurricanes, at which it has a 80-95% success rate. Currently AI seems to be accurate in the short term, but long term models generated by AI have differences to ones created by scientists. As the AI does not explain how it reached these conclusions, they are often deemed unreliable by climate scientists. More innovation in the field, by both people and AI is needed to create more accurate computer generated predictions. 

AI predicts the intensity of natural hazards like hurricanes will increase dramatically due to climate change.

Credit: https://connect.agu.org/naturalhazards/home

AI is still useful in the field of long term climate predictions. Bodies such as the IPCC (intergovernmental panel on climate change) use an average of many different climate models to come tho their central predictions. AI is used to determine which models are more reliable, by tracking their accuracy, and then giving them more weight within the average. This increases accuracy of the average predictions, which are often the ones presented to the media and governments.

The introduction of AI has also been used as an argument for the existence of climate change, as based on past records its gives similar predictions to scientists. This could make it a powerful tool in the fight against climate denial as there can be no human error or opinion based argument. 

Currently AI is well suited to short term predictions but is unable to agree with scientists on long term predictions, partly due to the sheer number of possibilities and permutations, which can significantly change predictions. It is also used in the statistical element of the calculating predictions to ensure weighting is optimised. Future innovation will perhaps see all predictions based on AI , with checks from scientists rather than visa versa.

Thumbnail : https://www.forbes.com/sites/bernardmarr/2018/02/21/the-amazing-ways-we-can-use-ai-to-tackle-climate-change/#16a00d804a35

Alfie Birkett