The current generation of near-infrared imaging of the Milky Way with photometric catalogues enables us to see further into the structure of the inner galaxy. With recent deep photometric catalogues extracted out of data released from the Vista Variables in Via Lactea (VVV) survey, we reconstruct the stellar density of the Galactic bulge via several methods. We fitted parametric models which are used as priors in our non-parametric maximum entropy and algorithmic smoothing regularised likelihood optimisation using a semi-analytic luminosity function of red giant stars constructed from stellar isochrone models. Our reconstruction naturally inpaints overcrowded and high extinction regions with curvature based inpainting via the algorithmic smoothing. Assuming two-fold symmetry we obtain our non-parametric bulge model over the inner 40◦ × 40◦ region centred on the Galactic centre. Our resulting bulge properties from our best fitting model and systematics were found to be consistent with other investigations utilising the VVV data.

In further investigation of the VVV bulge data, we observed a stellar density in front of the bulge which provided a near side counterpart to the spiral arm structure behind the bulge previously observed in the literature with VVV catalogues. We determined that these features are sensitive to the modelling of the Red Giant Branch Bump and Asymptotic Giant Branch Bump components in our luminosity function construction.

Anomalous emission in the GeV energy range has been detected towards the centre of the Milky Way by the Fermi-Large Area Telescope (LAT). Theories proposed for this Galactic Centre Excess (GCE) include: self annihilation of dark matter particles, an unresolved population of millisecond pulsars, or a series of burst events. The diffuse emission of gamma-rays via cosmic-rays interacting with the interstellar gas and radiation fields dominates the sky in the energy range of the excess. To apply models of the GCE to Fermi-LAT data, this background must be accounted for with care. We utilised a method of modelling the interstellar gas position that exploits hydrodynamic modelling in conjunction with previous gas modelling techniques. We found the hydrodynamic model improves the quality of fit to the Fermi data over previous methods applied. We also found a boxy Galactic bulge and nuclear stellar bulge produced a statistically better fit to the GCE than a spherical excess that would be more indicative of dark matter self-annihilation. Based on this result we argue that the GCE is likely associated with the stellar population of the Galactic bulge rather than dark matter. Applying new bulge models from our morphological analysis of the Red Clump in VVV infrared data further improved the statistical preference for a non-spherical excess rather than a dark matter interpretation.