Very High-Resolution Imagery and Machine Learning for Detailed Mapping of Riparian Vegetation and Substrate Types

Edvinas Rommel,Ina Quick,Paul Deffert,Maike Heuner,Katharina Fricke,Jens Bongartz,Gilles Rock,Björn Baschek,Filip Dzunic,Andreas Burkart,Paul Näthe,Maryam Asgari,Tina Mölter,Laura Giese,Frederik Kathöfer,Uwe Schröder

doi:10.3390/rs14040954

Abstract

Riparian zones fulfill diverse ecological and economic functions. Sustainable management requires detailed spatial information about vegetation and hydromorphological properties. In this study, we propose a machine learning classification workflow to map classes of the thematic levels Basic surface types (BA), Vegetation units (VE), Dominant stands (DO) and Substrate types (SU) based on multispectral imagery from an unmanned aerial system (UAS). A case study was carried out in Emmericher Ward on the river Rhine, Germany. The results showed that: (I) In terms of overall accuracy, classification results decreased with increasing detail of classes from BA (88.9%) and VE (88.4%) to DO (74.8%) or SU (62%), respectively. (II) The use of Support Vector Machines and Extreme Gradient Boost algorithms did not increase classification performance in comparison to Random Forest. (III) Based on probability maps, classification performance was lower in areas of shaded vegetation and in the transition zones. (IV) In order to cover larger areas, a gyrocopter can be used applying the same workflow and achieving comparable results as by UAS for thematic levels BA, VE and homogeneous classes covering larger areas. The generated classification maps are a valuable tool for ecologically integrated water management.

Highlights

Images acquired by gyrocopter were classified with the same workflow (RF algorithm) as for unmanned aerial system (UAS) data resulting in overall accuracy (OA) of 88.4% (BA), 86.4% (VE), 65.6% (DO) and 52% (SU)
We proposed an object-based image analysis workflow to classify vegetation and substrate types with different level of detail, using multispectral very highresolution UAS and gyrocopter imagery in a riparian ecosystem along the river Rhine in Germany
Classification results for UAS data with Random Forest (RF) decreased with increasing class detail from Basic surface types (BA) (OA = 88.9%) and Vegetation units (VE) (OA = 88.4%) to Dominant stands (DO) (OA = 74.8%) and Substrate types (SU)

Summary

Introduction

Surface waters, their riparian zones and floodplains are hotspots for highly specialized vegetation communities and fulfill valuable ecosystem services. Flooding and water availability drive the small-scale distribution of vegetation communities [1]. The interplay of vegetation and sedimentation and erosion processes leads to diverse terrain structures which provide habitats for endangered species [2,3]. Riparian zones perform several functions: they act as natural corridors for terrestrial wildlife, contribute to water purification, reduce flood vulnerability and serve as areas of recreation [4,5]. They are often degraded or eliminated by human activities, e.g., due to chemical contamination

Objectives

Methods

Results

Discussion

Conclusion