Digital Classification of Anthropogenic Features for Natural Terrain Hazard Assessment in the Quasi-natural Heritage Landscape of the Lin Ma Hang Lead Mine

Much of the Hong Kong landscape consists of densely vegetated steep hillside and may give the impression of natural terrain untouched by man-made activities. However, much evidence of old human activities occurs in our vegetated landscape. The old lead mine workings in the Lin Ma Hang district of the northeast New Territories form a significant industrial heritage site now hidden by dense vegetation. Extensive old anthropogenic activities are seen in site reconnaissance. Most of the man-made features were formed during the mining period (1860-1960) and the WWII (1941-45) occupation of the mine site. Some features have more obscure origins associated with cycles of agricultural activity and settlement of more than 1000 years. The unique and diverse nature of the Lin Ma Hang hillsides provides an ideal case study to demonstrate the benefits of systematic assessment of anthropogenic features in Natural Terrain Hazard Assessment. Some of these man-made features may create impacts as potential adverse Hillside Pocket scenarios and require inventory and classification during natural terrain hazard and other geotechnical studies (Ho & Roberts, 2016). Over the past decade, the application of airborne LiDAR data for site characterization has grown significantly, in part due to advances in handling of very large data sets. Through 3D topographic models using LiDAR in combination with visual data, landforms are revealed and terrain classification is enhanced allowing identification of anthropogenic features of varying scale and origin within their geomorphological setting. The authors discuss the application of a digitally aided approach for terrain mapping with emphasis on the identification and classification of anthropogenic features based on size, type, origin, material, extent and location. These are classified within a Hong Kong-based framework of an 80 class classification following from Styles & Law (2012).