Remote sensing as a tool for monitoring plant invasions: Testing the effects of data resolution and image classification approach on the detection of a model plant species Heracleum mantegazzianum (giant hogweed)
Abstrakt
Plant invasions represent a threat not only to biodiversity and ecosystem functioning but also to the character of traditional landscapes. Despite the worldwide efforts to control and eradicate invasive species, their menace grows.
New techniques enabling fast and precise monitoring and providing information on spatial structure of invasions are needed for efficient management strategies to be implemented. We present remote sensing assessment of a noxious invasive species Heracleum mantegazzianum (giant hogweed) that integrates different data sources, spatial and spectral resolutions, and image processing techniques.
Panchromatic, multispectral and color very high spatial resolution (VHR) aerial photography (1947-2006, resolution 0.5 m), and medium spatial resolution satellite data (Rapid Eye 2010, resolution 5 m) were analyzed to assess their potential for hogweed monitoring by using pixel- (both supervised and unsupervised) and object-based image analysis (OBIA, automated hierarchical, iterative, and rule-based). Both point and grid based accuracy assessment was carried out.
Described methods of object-based image analysis of VHR data enabled monitoring of hogweed at high classification accuracies measured by various means, regardless of the spectral resolution of the data provided that the data came from the species flowering period. Although the proposed automated processing of VHR data is relatively time-effective and standardized, application over large areas would be rather demanding due to the size of datasets, and multispectral satellite data of medium spatial resolution (lower than the size of individuals) was therefore tested.
High accuracies achieved for VHR data indicate the possible application of described methodology for monitoring invasions and their long-term dynamics elsewhere, making management measures comparable precise, fast and efficient.