8. SUMMARY AND FUTURE PROSPECTS

The vast amount of work on the M31 halo over the past 15 years has led to an exceptional situation where, in many ways, our knowledge of the peripheral regions of that system far exceeds our knowledge of those regions in our own Milky Way. Studies have been able to identify and characterise coherent debris streams in the M31 halo out to projected galactocentric distances of ∼ 120 kpc, and detect and map the properties of the ‵smooth' halo out to ∼ 150−200 kpc. Even in the inner halo (≲ 50 kpc), we have a much clearer understanding of the nature and origin of the tidal debris, thanks to our external perspective. This can be contrasted with the situation in the Milky Way where many of the inner halo/outer disk structures are very poorly understood (e.g. the Monoceros, Hercules-Aquila, Tri-And, and Virgo overdensities; see Chapters 3 & 4).

In comparison to the Milky Way, the M31 inner halo is littered with metal-rich debris. Some of this material has been stripped from the GSS progenitor, while the rest is likely to be material from the M31 disk heated by the recent impact of this satellite. M31 may also have more outer halo tidal streams than the Milky Way, although most Milky Way surveys to date have not had the sensitivity to detect substructures in these parts. Additionally, M31 (1) has a higher fraction of its total light in the halo component compared to the Milky Way (e.g. Ibata et al. 2014, Gilbert et al. 2012), (2) is characterised by a smooth density profile unlike the broken power-law of the Milky Way (Deason et al. 2013) and (3) has a substantially larger population of halo GCs, many of which lie along streams (Mackey et al. 2010b, Veljanoski et al. 2014). It is likely that these differences result from the unique accretion histories experienced by the two systems. M31 may have experienced more accretions than the Milky Way, or it may simply have experienced a more prolonged history of accretion. Indeed, there is a tantalising similarity between the properties of the Sagittarius dwarf and those inferred for the GSS progenitor – both are early-type galaxies with estimated initial masses in the range 0.5−1 × 10⁹ M_⊙. The major building blocks of these halos may well have been comparable, but their orbits and accretion times rather different. In this spirit, it is interesting to speculate how the Milky Way and M31 halos would have compared to each other ∼ 2 Gyr ago, before the GSS progenitor entered M31's inner halo.

A number of outstanding questions remain regarding the M31 system, and much exciting progress can be expected over the next decade. Some of the most pressing issues that remain to be addressed include:

• The development of a definitive understanding of M31's most significant recent accretion event: what was the true nature of the GSS progenitor and where is the remnant now? How did it come to be on such an extreme orbit and how did it survive in this state until ∼ 1 Gyr ago? What is the explanation for the second velocity component in the GSS and is there any connection between the GSS progenitor and the similarly metal-rich component of Stream C?
• Understanding the origin of the outer halo debris. What and where are the progenitors of the outer halo debris streams? Do these features result from one or many accretion events? The associated GCs are likely to play an important role in answering these questions, and also in using these streams to derive refined estimates of the mass and potential of M31. While line-of-sight distances to stream GCs are possible with HST observations, proper motion measurements may be possible for the most compact objects with Gaia.
• Understanding the origin of the coherent rotation in the outer GC population, and how it can be reconciled with the supposedly chaotic accretion of parent dwarf galaxies into the halo. Additionally, do the underlying debris streams and the smooth field halo also exhibit this rotation? Given the sparse nature of the stellar populations in these parts, very large-scale spectroscopy campaigns covering a significant fraction of the halo will be required to answer these questions.
• Developing a holistic picture of M31's evolution that links its accretion history to its global evolution and current structure. There are fascinating hints that the recent interaction and accretion history of M31 can explain a variety of puzzling observations. The close passage of M33 could excite the strong asymmetric warps and bursts of star formation observed in both systems, while the accretion of the GSS progenitor could further disrupt the M31 outer disk, displace some fraction of the disk stars into the halo and deposit a substantial amount of metal rich debris in the inner halo. Much work is required to verify and fine-tune these ideas and there is a particular need for further detailed N-body modelling.
• Searching for the edge of the M31 stellar halo. Tidal streams, field stars and GCs have been found to the very edge of the PAndAS survey, suggesting that they could extend yet further. New wide-field mapping facilities such as Hyper-Suprime Cam on the Subaru telescope will be required to efficiently explore beyond the limit of PAndAS and may benefit from the use of specialised filters to discriminate between foreground and M31 populations.

Acknowledgements We thank Edouard Bernard and Jovan Veljanoski for their help in creating Figs. 7 and 8. AMNF acknowledges support from an STFC Consolidated Grant (ST/J001422/1) and the hospitality of the Instituto de Astrofisica de Canarias while completing this chapter. ADM is grateful for support from the Australian Research Council through Discovery Projects DP1093431, DP120101237 and DP150103294.