This is a book about the use of simulations to understand the cosmos. In most other areas of science, scientists are able to make predictions, then perform experiments to test those predictions. One of the challenges with astronomy is that things develop so slowly over time that scientists don’t have time to wait and see if their predictions come true. Simulations provide a way around this problem. By capturing their knowledge and predictions in computer code, and running simulations, they can predict how astronomical objects such as galaxies would change over time. Then they can compare these predictions with what they see through telescopes. If galaxies in the real universe look similar to those that result from the simulations, it suggests that the rules used within the simulation are correct.
The book starts by explaining the principles of simulations in a more familiar area: the weather. It’s also an area in which things develop quickly enough that predictions can be tested by seeing how the weather changes over the next few days. The book charts the history of weather forecasting using simulations, and discusses the difficulties and successes that have been achieved. One of the interesting concepts introduced at this point is that of “sub-grid rules”. In any simulation, the part of the universe being simulated is divided up into a grid. But it’s not possible (with current computer technology) to simulate every single particle and all its interactions. Sub-grid rules are general rules that are applied to make the simulation more realistic. In some ways, these are a fudge to get the simulation to come out right. But they can also be useful in letting scientists know if their understanding of the universe is correct: adding a sub-grid rule that encapsulates some understanding about how clouds form, for example, allows scientists to understand if they’ve got that rule right. It also allows them to see what happens if that rule is removed.
The book then goes on to discuss the author’s area of expertise, which is simulations of galaxies and other astronomical phenomena. Simulations have been crucial in discovering the mysterious dark matter and dark energy that make up the majority of our universe. If dark matter is not included in galactic simulations, the galaxies fly apart. Clearly this is not what happens in the real universe, so scientists can infer than dark matter exists. In galactic simulations, the sub-grid rules also play a role, such as in correctly capturing the behaviour of black holes on the formation and development of galaxies. Some science is covered during the book in relation to all the simulations, such as general relativity, quantum mechanics, the ongoing search for a theory of quantum gravity, and what may have happened at the very start of the universe to produce the structures that we see today.
The book takes an interesting turn towards the end, becoming more philosophical. Pontzen discusses the possibility that we might actually be living inside a simulation ourselves, and compares several points of view on this. He also talks about the possibilities of quantum computers for running simulations.
Another topic discussed is whether simulations can be considered to be experiments in the normal scientific sense, and the author makes the case that they can. But he’s also careful to point out the pitfalls that can arise when simulations are taken too literally, or are tweaked in order to get the results you’re looking for.
This book doesn’t go too deeply into the details of how simulations work, or the science behind them. But I enjoyed it as an introduction to a world that I’d never thought much about before, but which has been crucial for our understanding of the cosmos.