What is in the PAS Database Part II: Spatial Data and Fuzziness

Records in the PAS database contain general geographic data in the form of the finds’ European region, county (or Local Authority Area), district and parish name, and geographic coordinate data given as the UK Ordnance Survey (OS) grid reference, easting/northing (derived from the former) and longitude/latitude. Access to the coordinate information is restricted to OS four-figure grid reference level unless special access is granted, such as a for research purposes.

The spatial data of PAS finds is typically based on information supplied by the finders. Its precision varies depending on how the findspot has been recorded. Taking into account this geospatial uncertainty – or ‘fuzziness’ – is particularly important for study of the PAS finds data in its local landscape context. The graph below gives a breakdown of the degree of precision of the locational information in the database with reference to the records’ OS grid reference information.

blog 2. spatial precision breakdown.png

A 4-figure OS grid reference identifies a square 1 km to a side within which the find is located. Each successively higher figure reduces that distance by an order of magnitude, so that a 10-figure reference identifies 1m2 square. A bit of technical detail is good to keep in mind: the OS grid reference marks the coordinate location where the south-westernmost corner of the specific OS square lies on the ground; the actual findspot will be somewhere to the north and east of that location. Theoretically a finder may also have reported the coordinate reference of the nearest OS square corner, with the find actually lying outside the square itself. But in practice finders tend to report the coordinate data of the actual square the find falls into, as taken either from a handheld GPS device or an OS map.

There are local and regional variations in how the data is recorded, as is shown on the below map displaying PAS findspots in eastern England (finds recorded as centred on parish, village or field are included at the 4-figure precision level).

blog 2. PAS spat accuracy mapAs can be see, a majority of finds west of the Wash are recorded at a 6-figure spatial precision, while in East Anglia recording at 8-figure precision is more common. Moreover there are local cluster of finds records at 10-figure precision; often these are the result of metal-detecting rallies, or the work of industrious individual detectorists or local societies. The map below presents an overview of the average accuracy of PAS geospatial findspot data across England and Wales per 100 km2 (cold to hot with quantile breaks, squares containing less than 100 finds excluded).

blog 2. PAS Spatial Accuracy.pngOverall about 85% of the PAS entries can therefore be identified within a hectare. For statistical and pattern analysis on a regional or national level 6-figure spatial precision normally suffices. But for more fine-grained research variations in the spatial precision of the records may have a significant impact on understanding the relationship of the PAS data to the historic landscape, with 8- and 10-figure records potentially allowing specific activity sites to be identified

There are factors, however, that complicate the study of local historical activity patterns, even when a high degree of coordinate accuracy is available. The vast majority of finds are made by metal-detecting on arable land (see below for finds figures in 2015); the vast majority of these are recovered from the plough zone – within the depth turned over by the mechanised ploughing of the soil. PAS finds are therefore typically divorced from their archaeological context.

blog 2. PAS Land Use.png

A valuable recent examination of the post-deposition dispersal of PAS finds has been conducted by Adam Daubney in his 2015 PhD dissertation Portable Antiquities, Palimsests and Persistant Places. The action of the plough, for example, moves the objects within the boundaries of a field – possibly up to 5-10 m per episode of ploughing. The importation of topsoil or dredging may disperse objects yet further or bury them beyond the reach of ordinary means of discovery. And not all post-deposition movement of objects is modern. Artefact scatters may be the result not of concentrated one-site historical activity, but of historical manure and refuse disposal from settlements onto the surrounding fields. Examining the PAS data for such misleading patterns is essential for the close, local analysis of finds data. The key to understanding the PAS data in its local context may involve talking to the finders and local metal detectorists, who often have a deep understanding of the local landscape and the conditions affecting finds recovery.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s