Determination of the ecological connectivity between landscape patches obtained using the knowledge engineer (expert) classification technique


SELİM S., SÖNMEZ N. K., Onur I., Coslu M.

Remote Sensing for Agriculture, Ecosystems, and Hydrology XIX 2017, Warszawa, Polonya, 12 - 15 Eylül 2017, cilt.10421 identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası: 10421
  • Doi Numarası: 10.1117/12.2278350
  • Basıldığı Şehir: Warszawa
  • Basıldığı Ülke: Polonya
  • Anahtar Kelimeler: Ecological connectivity, Green infrastructure, Knowledge engineer classification, Landscape patches, Landscape planning, Nature conservation, Probability connection index, Tree cover detection
  • Akdeniz Üniversitesi Adresli: Evet

Özet

© 2017 SPIE.Connection of similar landscape patches with ecological corridors supports habitat quality of these patches, increases urban ecological quality, and constitutes an important living and expansion area for wild life. Furthermore, habitat connectivity provided by urban green areas is supporting biodiversity in urban areas. In this study, possible ecological connections between landscape patches, which were achieved by using Expert classification technique and modeled with probabilistic connection index. Firstly, the reflection responses of plants to various bands are used as data in hypotheses. One of the important features of this method is being able to use more than one image at the same time in the formation of the hypothesis. For this reason, before starting the application of the Expert classification, the base images are prepared. In addition to the main image, the hypothesis conditions were also created for each class with the NDVI image which is commonly used in the vegetation researches. Besides, the results of the previously conducted supervised classification were taken into account. We applied this classification method by using the raster imagery with user-defined variables. Hereupon, to provide ecological connections of the tree cover which was achieved from the classification, we used Probabilistic Connection (PC) index. The probabilistic connection model which is used for landscape planning and conservation studies via detecting and prioritization critical areas for ecological connection characterizes the possibility of direct connection between habitats. As a result we obtained over % 90 total accuracy in accuracy assessment analysis. We provided ecological connections with PC index and we created inter-connected green spaces system. Thus, we offered and implicated green infrastructure system model takes place in the agenda of recent years.