The evidence for the Severan campaigns (Birley 1999b, 170-187; Breeze 1982 , 128-136) comes primarily from literary sources, propaganda coin issues and the distribution of temporary camps, although the dating of these latter is less secure than often claimed. However, they broadly support a large army heading up Dere St to the Forth, reducing in size slightly when heading into north-east Scotland; the northern limit of campaigning is unclear, but the east-coast focus is notable, indicating where the problem areas were seen to be. The navy was clearly a vital element in the supply chain, as indicated by the rebuilding of (coastal) South Shields as a massive grain store, the reoccupation of (coastal) Cramond and the use (new or continuing) of the coastal fortress at Carpow. A string of denarius hoards of this general date up the north-east has traditionally been linked to the progress of the Severan army, but is better seen as a broader diplomatic phenomenon in the decades before and (to a lesser extent) after the invasion (Hunter 2007c).
A fresh study of this period with a broad perspective would be of benefit, looking at the invasions in a wider chronological context (from the late Antonine period), considering evidence from elsewhere on the northern frontier (such as the South Shields rebuilding, the enigmatic Vindolanda circular huts, and the evidence from York; Bidwell 1999, 73-8; Birley 2009), reviewing the evocative propaganda evidence (not just the well-known coinage, but sculpture (eg Piggott 1968) and gemstones (eg Elliot and Henig 1999; Marsden 2011), and considering the effects on the local population; there are arguments for seeing this time as pivotal in the history of indigenous societies due to Roman political as much as military interference (Hunter 2007a).
Some workers (e.g. Whittington and Edwards 1993; Martin 1995) have suggested a dramatic impact on the local population tantamount to genocide; the evidence for this is not strong, but the question merits further work, especially to strengthen the palaeoenvironmental arguments of Whittington and Edwards substantially.