The application of remote sensing to environmental archaeology

An area in the Fenlands of Eastern England was used to assess the spectral, spatial, and seasonal requirements of airborne multispectral data for identifying wetland archaeological features by detecting crop and soil marks. Ordination of data from a scanner with 11 spectral channels was achieved using the Sheffield method, which calculates the wavebands that produce a 3 band composite image with optimum contrast using band variances and interband correlations. the spatial detail requirements for multispectral data were investigated by applying an edge enhancing filter to single waveband images. the loss of feature clarity with increased levels of spatial smoothing is visually apparent. Seasonal variation in crop mark visibility was quantified by cross tabulating feature visibility in fields known to contain features with time period and sensor type. If remote sensing is to be used for repeated operational archaeological survey then the quality and quantity of information which can be gained from the data must be evaluated. This study attempts to define some of the necessary operational requirements for the use of multispectral data for archaeology.