|
Forests are dynamic, complex and multidimensional ecosystems and play an irreplaceable role in social, economic, environmental, ecological and cultural context. Eucalyptus is the most common exotic species in Portugal forests. This species is fundamental in the industries related to the pulp paper production and the concern about their effects in ecosystems is growing. Geographical Information Systems (GIS) combined with Remote Sensing (RS) data can help to understand this complex ecosystem. Moreover, GIS and RS are commonly used in forest management. GIS allows the manipulation, analysis, and generation of considerable amounts of environmental information. This information can be used in the evaluation of ecosystems’ conditions and for decision making. The study case of this project was the municipal lands included in “Serras do Porto” and Valongo’s Nature 2000 network (Porto district, Portugal). The study zone considered in this work is a landscape of extreme relevance to Porto Metropolitan Area. For decades this area was extensively explored with eucalyptus plantations in order to produce cellulose for paper industry. Due to the characteristics of the area and its extension (40 hectares) the use of GIS became the most accurate and reliable alternative to characterize it. The combination of GIS tools and RS data allows the characterization of terrain relief, namely the analysis of altimetry, hypsometry, hydrography, the creation of environmental indexes such as Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI) and Normalized Difference Water Index (NDWI), and the Digital Elevation Model (DEM). RS technology offer the potential to explore the effects of land-use changes and disturbances on forest dynamics at large spatial scales. A Sentinel-2A image was used to produce NDVI, EVI, and NDWI environmental indexes and to generate the Land Use Land Cover (LULC) map, through Semi-Automatic Classification Plugin from QGIS software using Minimum Distance algorithm. The LULC was classified with two classes because the study area only presents two types of species: eucalyptus and bare soil. The LULC map obtained was validated through field points collected in the study area with a GPS receptor. An overall accuracy of 92.98% and a kappa statistic of 0.842 was obtained. Also, some of the geographic information obtained in the field was then integrated in QGIS software. Moreover, a phenological study was performed using NDVI values obtained from Sentinel-2A images, to understand the eucalyptus behavior in a certain period of time.. Because of that RS data provided useful information about the landscape dynamics allowing the assess to forest cover change and land use helping to create decision making plans and forest conservation measures.
|
