5.8 3-D Point data

Since the first excavations undertaken by the LRC at Cook's Quarry, West Heslerton, from 1977-86, all finds have been treated and recorded as individual items with their 3-D location. The laborious task of gridding areas and levelling individual find-spots applied in the first few years was replaced in 1984 when an EDM mounted on a theodolite was employed for the first time to calculate the 3-D point location for every find-spot. Since then increasingly sophisticated Total Stations have been used for all survey purposes. Using our own logging software, a wide variety of survey data can be simply collected using a numeric type code to identify the type of data collected, such as an individual context, drawing point, find-spot, contour survey point or the crude outline of a feature.

A single 3-D coordinate is recorded for the centre of the surface of every context identified; in the case of a cut-feature 'the centre of the measured surface' is taken as the lowest point of the cut. The incorporation of a 3-D point in every record facilitates the production of simple plots and distributions such as plots of post-holes using a symbol, sized according to the feature depth, a characteristic that may not be immediately apparent when looking at the digitised plans covering a number of buildings. The inclusion of the coordinates in the context record serves as a check against other parts of the record and, given the very large areas excavated, a simple method of finding the location of a single context among many thousands.

The individual recording of all finds, including their 3-D location, is fundamental to the LRC recording system and is in contrast to more common practice, where only those finds deemed to have some special importance (so-called small finds) are considered worthy of this sort of treatment. The detailed approach to recording at Heslerton affords us the opportunity for comprehensive spatial analysis of the excavations and the assemblages that they contain, allowing far more detailed analysis of the formation and deformation processes that create the archaeological deposits than would otherwise be possible. Such an approach may not be possible in the case of deeply stratified urban sites with huge finds assemblages. In the rural environment within which the LRC operates, the 3-D recording of sparse distributions allows approaches that effectively pad out the context record to be avoided. For example, during the excavations at Cook's Quarry in the late 1970s, long and very often characterless ditches were excavated and recorded in short segments to provide a degree of spatial control with regard to finds recovery, despite the lack of any distinguishing feature between one segment and the next: this generated multiple cut and fill records that were effectively copies of the first excavated segment. However, by recording the individual find-spots an absolute density is recorded and single context records can be generated covering many metres of a single ditch, defined not on the basis of regular segments but on the basis of unified sections.

In the countryside, where we are most often concerned with horizontal stratigraphy and vertical stratigraphy which is often barely visible within the highly uniform soils that fill most of the features, the approach does generate a far more powerful dataset. The 3-D recording of find-spots and subsequent individual recording of all finds can be a daunting process in areas where finds densities are high. However, the ability to examine distributions in three dimensions offers powerful analytical potential, witness the rotating plot (see Fig. 17a and Fig. 17b). The investigation and analysis of the blown sands in Trench 28AC, discussed above, would not have been possible without this approach to finds retrieval and recording. Many of the arguments regarding the nature of the fills of Grubenhäuser discussed by Jess Tipper and arising out of the West Heslerton Anglian Settlement excavation could not have been developed or tested without the analysis of three-dimensional distributions of the finds within these features (Tipper 2004). This approach to finds retrieval and recording requires a disciplined and well-organised approach to fieldwork in addition to the availability of appropriate data management and software; in the case of the LRC we are fortunate to have our own software, G-Sys, which is dedicated to our data and allows us to undertake interactive analysis and data visualisations that are not available within other off-the-shelf software.

The method adopted for 3-D recording the finds is simple and relies on the use of uniquely numbered tags used to mark each find location. As each find is recovered an individual nailed tag, printed initially with ten find numbers, is used to mark the find-spot. The find is individually bagged and the lowest number on the tag (that furthest from the nail hole through the tag) is then written on the finds bag with the Site Area and Context number. Once work over a surface is complete or a deposit has become so crowded with tags that work is made difficult, the tags are surveyed and removed, recording the Site, Area, Context, tag number and coordinates in the total station data logger. The tags are lifted as surveyed and then the used numbers cut off the tag, which can then be reused. The unique tag number is recorded in the field in the Object Record, providing the link required to import the coordinate data into the record.

One result of the approach adopted for tagging finds is that the density of find-spots across the site is often clearly visible on the ground during excavation. In some cases this has revealed the relationships between intercutting features filled with visually indistinguishable fills, where the intersection is matched by a variation in density, reflecting where one feature, incorporating a high density of finds, is cut through by one with only a few, or vice-versa.

In addition to gathering point locations for Objects and Contexts or contour surveys, the total station is used to gather boundary points that define simple polygons used for basic pre-excavation planning of large areas. Each polygon is automatically generated, including the KEY_ID as the Total Station data is downloaded.


© Internet Archaeology/Author(s) URL:
Last updated: Wed Nov 11 2009