WALMET - Hyperspectral characterisation of lead dispersal from abandoned metalliferous mining in Wales, UK

Context and objectives

The aim of this study is to map the mine waste within the Rheidol valley, faster and more
safely than by using geochemical analysis alone. Remote sensing data has been used in past
projects to map mine waste in different forms. The MINEO study was the first of these studies
within BGS to use hyperspectral data to map mine waste in the UK. The area studied in this
case was the tin mining region of Cornwall.

The objectives of the study are:

• To map mine waste using hyperspectral image processing techniques already devised
from previous studies
• To asses different methods used for mine waste mapping and attempt to replicate these in
this area
• To produce a map showing the different mine waste present at each of the mine sites

Project outcome

Expected scientific results

The result of the final version of the SAM classification is shown below. The end member
spectra used for classification are also shown in figure below. The colours used in both
figures match to illustrate where each end member spectrum is located in the imagery and on
the ground. Colours were chosen to highlight the regions of mine waste against the
background colours of vegetation and natural materials.