OBJECT BASED IMAGE ANALYSIS OF DYING SPRUCE STANDS IN GORCE AND TATRA NATIONAL PARKS USING THE PLANE

Interaction between natural processes and human activity results in changes in land use and land cover. Using GIS analysis, it is possible to estimate the dynamic of these changes and therefore detect former trends or extrapolate changes in the future.

The aim of the project was to develop the map of changes in the structure and health of forest areas adjacent to hiking trails in the Tatra and Gorce National Parks (TPN and GPN) using a Geographic Object-Based Image Analysis (GEOBIA) in order to highlight the advantages of this innovative, semi-automatic method of image analysis. Special emphasis will be laid on dead standing trees, which do not represent value to the natural landscape. Providing security on hiking trails should include removing dead and diseased trees that could threaten safety of individuals, who stay on trails and paths. Project will use remote sensing scenes from RapidEye, Planet and Airborne Laser Scanning data (ALS, project ISOK in Poland).

The development of remote sensing and high-resolution satellite images with a daily return determined the direction of modern research oriented at automating the process classification of the data. One of the very rapidly developing methods of automatic classification is Object Based Image Analysis (OBIA). An important advantage of the OBIA method is the ability to effectively combine data from various sources, such as image data, LiDAR, digital surface model, as well as GIS vector data.

The aim of the present study was to use object classification for automatic detection of dead or dying trees in spruce stands of Gorce and Tatra National Parks. In this paper the PlanetScope satellite images were used (GNP – 30 July 2017; TNP – 24 August 2017), they were characterised by high spatial resolution (GSD 3.125m), spectral (4 spectral bands: R, G, B and NIR), radiometric (16 bit), as well as LiDAR data, nDSM. Data was analyzed using eCognition Developer (TRIMBLE) software.

As a result of the analysis, a map of the distribution of dying trees in the area of two mountain National Parks was obtained. The 644 hectare of the Tatra National Park covers dead trees, which constitute 3.03% of the TNP. In Gorce NP it was 85 hectare (1.21% of the GNP).

Analysis of the accuracy of the obtained results showed that the OBIA method gives quite good results (Kappa coefficient equal to 0.8) in the automated process of generating current maps of the occurrence of dying trees in mountainous areas of southern Poland based on PlanetScope satellite images and LiDAR data.

Additionally, smartphone application (Android, iOS) called „Bezpieczny Szlak” was developed to allow input of data by hikers who would notice unmarked trees matching criteria. To encourage usage of this application it will show a map of the region and therefore will function as a hiking guide. Information gathered by users will be processed to create overview of situation in selected area.

The analyzes will be repeated in subsequent years on similar data.