OJ Research Contact

Research

I am primarily focused on researching atmospheric radative transfer (see SOCRATES). However, I am also investigating making use of machine learning to improve modelling efficiency. Prior to joining the Met Office, I was a researcher in jet physics. More specifically, jets emanating from black hole x-ray binaries and AGN/Blazars.

X-ray binaries

Radio to Infra-red flat spectra observed in X-ray binaries have presented a long standing conundrum of how the emitting lepton population can be re-accelerated. This re-acceleration is necessary if adiabatic energy losses are to be taken into account. In order to investigate this I, alongwith Prof. Rob Fender and Dr. Christian Kaiser, developed a new simulation (iShocks) where multiple internal shocks are used as a re-acceleration mechanism. We show that, for the first time, it is possible to reproduce a radio to infra-red flat spectrum with the adiabatic energy losses present.



The multiple internal shocks set up also provides a natural set-up to also investigate the variability from these jets. We know that optical/infra-red emission from X-ray binary jets varies on short (sub-second) time-scales. Such variability in the spectrum in observed when using iShocks. This is demonstrated in the animated gif below, where as the time increases can see the spectrum extending to lower frequencies as well as showing rapid variability. We are currently further investigating this variability and applying iShocks to reproduce various observations.

Blazars

During my post-doc at Ohio University I developed a new numerical model to simulate blazar jets (the model itself is scale agnostic and can be used to simulate XRB jets as well). The main focus of this work was to implement synchrotron radiation and Compton scattering with full angle dependence within distributions framework. Using electron/photon distribtuions instead of a ray tracing Monte-Carlo simulation allows us to get higher photon statistics, particularly at high energies, without extremely high computational costs.