Mapping Alkaline Fens, Transition Mires and Quaking Bogs Using Airborne Hyperspectral and Laser Scanning Data

The aim of this study is to evaluate the effectiveness identification Natura 2000 wetland habitats (Alkaline fens—code 7230, and Transition mires quaking bogs—code 7140) depending on various remotely sensed (RS) data acquired from an airborne platform. Both remote sensing botanical reference were gathered for mentioned in Lower (LB) Upper Biebrza (UB) River Valley Janowskie Forest (JF) different seasonal stages. Several classification scenarios tested, ones that gave best results analyzed indicated each campaign. In final stage, a recommended term acquisition, as well list products, which allowed us achieve highest accuracy mapping these two types habitats, presented. Designed integrated hyperspectral products such Minimum Noise Fraction (MNF) bands, spectral indices derived Airborne Laser Scanning (ALS) representing topography (developed SAGA), or statistical OPALS—Orientation Processing Scanning). image classifications performed using Random (RF) algorithm multi-classification approach. As part research, correlation analysis developed was carried out, Recursive Feature Elimination with Cross-Validation (RFE-CV) select most important RS sub-products thus increase efficiency developing habitat distribution maps. showed alkaline fens are better identified summer (mean F1-SCORE equals 0.950 UB area, 0.935 LB area), transition bogs evolved on/or vicinity autumn 0.931 summer, 0.923 dystrophic lakes spring 0.953 spring, 0.948 JF area). also points out both highly improved when combining selected (MNF indices) ALS topographical products. This article demonstrates information provided by synergetic use sensors can be used monitoring its future functional assessment and/or protection activities planning high accuracy.