Hyperspectral modeling of ecological indicators – A new approach for monitoring former military training areas

摘要: Abstract Military areas are valuable habitats and refuges for rare endangered plants animals. We developed a new approach applying innovative methods of hyperspectral remote sensing to bridge the existing gap between technology demands nature conservation community. Remote has already proven be monitoring instrument. However, approaches lack consideration applied which includes legal EU Habitat Directive. Following idea Vital Signs Monitoring in USA, we identified subset highest priority indicators our study area. analyzed continuous spectral response curves tested measurability N = 19 on basis complexity levels aggregated from extensive vegetation assemblages. The differentiability floristic as well faunistic revealed values up 100% accuracy. point out difficulties when it comes distinguishing habitat requirements several species adapted dry open landscapes, this case results O verall accuracy 67, 87–95, 35% error matrix. In summary, provide an applicable feasible method facilitating military by following.