fbpx What is a simulation? | The Morton Arboretum

What is a simulation?

The idea of simulations always intrigued me, even though I never knew much about them. I was interested in their predictive power, how modeling complex systems could be simplified into code, and what potential they have in many fields.

What are simulations and why do we use them?

Simulations are computer programs that model some phenomenon. Simulations represent simplified versions of these real-life systems. Simulations are extremely useful tools that have long been used in a wide variety of fields, such as physics and many areas of engineering. With increasing accessibility, computational power, and more user-friendly interfaces, researchers in other fields like medicine, ecology, evolutionary biology, have also begun using these tools. To run a simulation today, you most likely will not need any coding experience at all.

Simulations can be used to examine possible results or outcomes or to predict future responses. Simulations are extremely useful in projects where the study would be infeasible due to time constraints, or quite literally impossible to do. For example, if you wanted to study the evolutionary history of a particular species, you could examine real life taxa. However, this only represents one ‘dataset’ or one possible result. With simulations, one could possibly examine hundreds or thousands of possible results. 

In addition, researchers can manipulate the variables going into a simulation in order to examine outcomes in an idealized, controlled environment. This is particularly useful in many fields because we don’t have to worry about so many confounding variables.

Simulations have great implications in the biological sciences, especially in conservation biology. Simulations can predict species responses to environmental factors, assess management options, and compare outcomes of different breeding programs, for example.

What does a simulation look like?

Before I started work this summer, when I heard the word simulation I imagined something like the picture below. And it’s partially true, because a simulation can be a lot of things, even this video game.

Screenshot from an evolution sandbox videogame

Screenshot from 'Intelligent Design,' an evolutionary sandbox videogame. 

This game is called Intelligent design: an evolutionary sandbox. The simulation models evolution, so it represents a simplified version of this phenomenon. You have control over how many plants you grow and how many herbivores and carnivores you include in your ecosystem. Traits that you can curate are passed on from generation to generation, along with some random mutations. Eventually, you can ‘unlock’ new species. You can modify your species with different combinations of traits, and examine the ecosystem response. You watch in ‘real time’ as the system changes. This game provides a fun way to visualize an ecosystem you’ve created, and it is a type of simulation! 

However, a simulation can also look like this:

Simcoal program simulation software window
Screenshot of a simulation running in the software Simcoal 2.
This is Simcoal2, the simulation software I’m using this summer. Simcoal models molecular genetic diversity in populations of a species. To run a simulation in simcoal, first you need to create a parameter file. This is just telling the simulation how many populations you want your species to have, how large the populations are, what migration is occurring between populations, and a simplified version of the life history of the species (basically just the species traits). In other words, the parameter file sets up scenario you want to simulate.
Then, you run the parameter file through Simcoal, and it gives you a LOT of output files! The output files represent simulated genotypes (genetic data) of individuals. Here’s an example:
Simcoal simulation output file example
Arlequin file example. This file is the output file from Simcoal after the simulation is run.
There’s a lot of numbers here. The different numbers represent different alleles (different versions of a gene). This is genetic data, and that’s cool! You can see the amount of variation that the simulation creates. There will be thousands of other files similar to this one, all with different genetic data, when I run my final simulations. The data isn’t very readable or accessible in this format, so we do some conversions to get it into the software R to do our analyses.
Simulations are cool!
Simulations have great predictive power, and can provide insight into complex natural systems that would be difficult to do with a traditional empirical study. Simulations are useful for many broad applications, and they are becoming more and more accessible. With this increasing accessibility, people who aren’t trained in computer science or computational fields (like me!) can utilize simulations in their work, research, or just for fun!