![]() Randomness/stochasticity is also necessary for modelling the presence of artifacts within an intersected test-pit. Where p is the probability of finding one or more artifacts, a is the area of the test-pit and d is the artifact density. The pseudo-code for this process is as follows: If the total pit density is greater than 1, the site is recorded as detected, and the site detection tally is increased by 1. All the pit-squares then calculate the total pit density by adding together the artifact-densities of each pit-square in the pit-set and then dividing that figure by the total pit area. If this site overlaps a test-pit, the artifact density of the site at the point of overlap is transferred to the pit-square. The detection calculation occurs in a similar manner, however a detection threshold of 1 is set here. This allows sites which only partially overlap a test-pit to be tallied as intersected by the test-pit as a single unit. During this calculation, all the pit-squares in a pit-set ask their fellow pit-squares if they intersected any sites in that run. If a site overlaps a single test-pit, the intersection tally for that test-pit increases by 1. The number of sites added is determined by the user through the variable ‘trial-runs’, which is set by default to 1000. During the simulation the sites are added to random locations within the survey area. The survey area and the test-pits are now loaded and drawn, and the characteristics of the sties have been established on the user-interface. This density distribution is calculated with the following equation: Decay then proceeds linearly to zero at the margin of the site. The centre of the site will have the highest density, which is set at 3.0 times the average density. If the density distribution is set to a linear, then a regular density decay occurs. In the uniform density distribution all points within the site will have the same density as the average density. The density distributions are described below. The density distributions which can be selected are: uniform, linear, cosinusoidal, and the Lake George Regression (LGR) curve. The density of each point within the site is calculated according to the average-density and the density-distribution. The size of the site is determined by the variable ‘site-diameter’. These properties can be adjusted between runs. The site-properties are established during the initialisation on the interface with the variables ‘average-density’, ‘site-diameter’ and ‘density-distribution’. The sites consist of concenrations of artefacts. The following pseudo-code describes this process: This allows sites which partially overlap a pit-set to be recorded as intersected/detected by the test-pit as a single unit. These pit-squares have the same ID as the seed-pit, and the seed-pit and pit-squares keep a tally of how many sites the test-pit intersects and detects. This allows test-pits of different sizes to be imported and constructed. The initial seed-pits create a pit-set of multiple pit-squares, which together define the extent of each test-pit. The ID, x coordinate, y coordinate, length and width are imported with the shapefile, and the area is then calculated from the length and width variables.Įach pit consists of a pit-set, which includes a single seed-pit, which is imported as a point during the importation of the shapefile, and additional pit-squares. ![]() Pits have the variables: ID, x coordinate, y coordinate, length, width, area, artifact-density, number of sites intersected and number of sites detected. The model then calculates the site density at that point and compares it with the test-pit size to determine whether the density is great enough for detection to also occur. When overlap does occur, the site is recorded as intersected. ![]() A simulation is run which randomly places sites, the characteristics of which can be varied between runs, in a sampling area with the specified test-pit layout to determine the probability of overlap between the test-pits and sites with the selected characteristics. what is the probability that sites of a certain size, density and density-distribution will be, or were, missed by the sampling program. Specifically, the aim of the model is to determine the inherent biases of the specified sampling program, i.e. The purpose of the DICI model is to assess the effectiveness of a specified subsurface sampling program in the detection of archaeological sites (Way and Tabrett, in press). This data article describes how the Dig It Check It (DICI) model operates.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |