Modelling Biodiversity in Highway Stormwater Ponds

Examensarbete för masterexamen

Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.12380/236727
Download file(s):
File Description SizeFormat 
236727.pdfFulltext3.39 MBAdobe PDFView/Open
Type: Examensarbete för masterexamen
Master Thesis
Title: Modelling Biodiversity in Highway Stormwater Ponds
Authors: Calveiro, Ricardo Francisco Hermida
Abstract: The development of road infrastructures causes great disruptions in the biodiversity of the natural areas. The Norwegian Public Roads Administration is investigating the possibility of employing stormwater ponds for compensating the loss of biodiversity due to the construction of the E39 highway. To define the guidelines for the design of biodiversity-promoting stormwater ponds, a model predicting biodiversity in stormwater ponds based on abiotic and biotic factors is needed. The literature review performed in this thesis showed that specific examples regarding biodiversity prediction models are scarce. However, several modelling approaches were described and one of them was identified as the most suitable: the Machine Learning methods. Using this approach, a model for predicting biodiversity in stormwater ponds was constructed. The model was based on the monitoring data collected during a sampling campaign performed within the NORWAT project at the Norwegian Public Roads Administration. During the sampling campaign several stormwater ponds along several major roads near Oslo in Norway were studied. Due to the different number of samples for water and sediment quality, two different models were built. In order to measure biodiversity three indices were defined: Species richness, Shannon diversity index and inverse Simpson’s index. The models were feedforward Artificial Neural Networks trained with the backpropagation algorithm. The results showed that the prediction capabilities were rather poor in all the cases but one, which performed well. The two models that were built showed very similar performances. The performances were in accordance with other results found in literature. Out of the three biodiversity indices, the species richness presented the best performance. This model confirmed that the Machine Learning models can be useful for biodiversity prediction.
Keywords: Samhällsbyggnadsteknik;Vattenteknik;Building Futures;Civil Engineering;Water Engineering;Building Futures
Issue Date: 2014
Publisher: Chalmers tekniska högskola / Institutionen för bygg- och miljöteknik
Chalmers University of Technology / Department of Civil and Environmental Engineering
Series/Report no.: Examensarbete - Institutionen för bygg- och miljöteknik, Chalmers tekniska högskola
URI: https://hdl.handle.net/20.500.12380/236727
Collection:Examensarbeten för masterexamen // Master Theses



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.