University of Twente Student Theses

Login

Woody species diversity using remote sensing and environmental data with Neural Network

Redowan, Mohammad (2010) Woody species diversity using remote sensing and environmental data with Neural Network.

[img] PDF
1MB
Abstract:Most of the remote sensing studies of plant diversity focused only on one aspect of plant diversity i.e. species richness, paying limited attention to other aspect i.e. species evenness. Moreover, environmental data is comparatively less used in the studies of remote sensing plant diversity and evenness, despite having potential to increase map accuracy. This study compares the mapping accuracy of woody species diversity and evenness of an Italian forest site by using two medium resolution imageries of Landsat 5 Thematic Mapper (TM) and Advanced Land Observing Satellite (ALOS) with Neural Network classifier. The study jointly used satellite data with four other environmental variables namely elevation, slope, aspect and solar radiation in different combinations. The highest overall map accuracies obtained using TM and ALOS imageries are 60% and 53% respectively for woody species diversity and 59% and 46% respectively for woody species evenness. Use of environmental data increases the classification accuracy significantly. The highest overall map accuracies obtained by using TM and ALOS data jointly with environmental data were 91% and 81% respectively for woody species diversity and 83% and 80% respectively for woody species evenness. Landsat TM maps both woody species diversity and evenness more accurately than ALOS, with or without environmental data. Coarser pixel size and higher spectral resolution of TM proved to be contributing factors in getting higher map accuracy compared to ALOS. Key words: Woody species diversity, woody species evenness, Shannon Diversity Index, Shannon Evenness Index, Neural Network, TM, ALOS, multicollinearity, forest types.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:https://purl.utwente.nl/essays/92509
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page