Transiting Exoplanets

How can we detect planets orbiting distant stars?

Since planets do not produce their own light, it is extremely difficult to directly detect them. Therefore, astrophysicists use indirection detection methods to search for them. One such method, focuses on attempting to detect the transits of exoplanets in front of stars.

As a planet passes in front of its host star, there is a small decrease in the amount of light that we detect. Therefore, if we notice a periodic dimming of a star, this could imply that there is a planet orbiting it.

From these 'lightcurves', we can infer information about the planet. For example, if we know the size of the star and the depth of the lightcurve, we can estimate the size of the planet.

Whilst still in high school, I worked in the Astrophysics department with Professor Coel Hellier. As part of SuperWASP, the UK's leading exoplanet detection program, I was tasked with analysing possible signals in order to detect new planets.

During my work, I was lucky to discover WASP-142b, a previously unknown exoplanet orbiting a star nearly 3000 light years from Earth, and become the youngest person to have discovered an exoplanet. I worked with Professor Hellier to analyse the planetary system and several other recent discoveries and these findings were published in Monthly Notices of the Royal Astronomical Society.

In 2018, TESS, the transiting exoplanet survey satellite, launched and began its hunt for new exoplanets with improved apparatus, I'm excited to see the new worlds that it discovers!

Population Genetics

How does migration and changing environments affect genetic evolution?

There exist many predictions for the evolution of populations if we assume a constant environment and no spatial separation. It is relatively simply to produce simulations involving selection, mutation and genetic drift under these considerations. We can first assume that within the population, every individual has either an a or an A Allele. Then we can define mutation rates between these alleles and their relative fitnesses. Finally, using these parameters, you can evolve the population, with one random individual dying and one randomly weighted individual being born at each time step.

I created quick simulation here to help understand the concept. You can adjust the parameters to see how this affects the evolution and be sure to click the help button for an explanation of the exact mathematics of how the simulation proceeds if you're interested!

My research with Professor Michael Desai focussed on creating a model for populations that are exposed to environmental changes and are divided into spatially separated subpopulations. I spent time iterating on creating an effective model that reproduced experimental results. Eventually I settled on the following; split individuals into M demes and then at each time step, randomly change environment between two environments with fixed probabilities. The alleles can then be assigned different fitnesses based on their environment.

Based on this model, I wrote a simulation in Python that could evolve a population for tens of thousands of generations in a couple of hours. These simulations produced intriguing results. I found that the allele which was beneficial for more of the time but with a lower fitness benefit was favoured as we increased the number of demes.