Modelling axial and circular data for vegetation stripes at Marion Island

dc.contributor.advisorNagar, Priyanka
dc.contributor.advisorBekker, Andriette, 1958-
dc.contributor.emailu18023691@tuks.co.zaen_US
dc.contributor.postgraduateVon Holtzhausen, Francois Victor
dc.date.accessioned2025-02-10T10:48:04Z
dc.date.available2025-02-10T10:48:04Z
dc.date.created2025-04
dc.date.issued2025-02
dc.descriptionMini Dissertation (MSc (Advanced Data Analytics))--University of Pretoria, 2025.en_US
dc.description.abstractIn this study, a bivariate model is proposed to analyse the joint distribution of Marion Island’s vegetation stripe and wind direction data. The objective is to investigate whether wind contributes to the formation of these irregular vegetation stripe patterns. Using a copula-based approach, the joint density function is modelled with a bivariate wrapped Cauchy circular component combined with various circular and axial distributions. Due to multimodality in the data, a finite mixture model is proposed to accurately model the overall density. This finite mixture model incorporates the slope angle and cone aspect as concomitant variables. The results indicate that a three latent component finite mixture model with von Mises and axial normal distributions as marginals provides the best fit. Using the proposed model it was determined that wind influences vegetation stripe orientation on the southern sides of cones, while no clear relationship is observed on the northern sides, likely due to harsher wind and sunlight exposure. These findings highlight the role of wind and other environmental factors, such as cone aspect and slope, in shaping vegetation patterns.en_US
dc.description.availabilityUnrestricteden_US
dc.description.degreeMSc (Advanced Data Analytics)en_US
dc.description.departmentStatisticsen_US
dc.description.facultyFaculty of Natural and Agricultural Sciencesen_US
dc.description.sdgSDG-13: Climate actionen_US
dc.identifier.citation*en_US
dc.identifier.doihttps://doi.org/10.25403/UPresearchdata.28369058en_US
dc.identifier.otherA2025en_US
dc.identifier.urihttp://hdl.handle.net/2263/100650
dc.language.isoenen_US
dc.publisherUniversity of Pretoria
dc.rights© 2023 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria.
dc.subjectUCTDen_US
dc.subjectSustainable Development Goals (SDGs)en_US
dc.subjectCopulaen_US
dc.subjectExpectation-maximisation algorithmen_US
dc.subjectMixture modelen_US
dc.subjectVegetation stripesen_US
dc.subjectAxial dataen_US
dc.subjectCircular dataen_US
dc.titleModelling axial and circular data for vegetation stripes at Marion Islanden_US
dc.typeMini Dissertationen_US

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
VonHoltzhausen_Modelling_2025.pdf
Size:
7.21 MB
Format:
Adobe Portable Document Format
Description:
Dissertation

License bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description: