Reconstructing NDVI Time Series from Landsat and MODIS and Developing a Species-Specific Validation Dataset for Vegetation Classification in Frontenac Provincial Park
Loading...
Authors
Zajaczkiwsky, Sophia Grace
Date
2025-09-30
Type
thesis
Language
eng
Keyword
Phenology , Remote Sensing , MODIS , Landsat , Frontenac Provincial Park , Growing Degree Days (GDD)
Alternative Title
Abstract
Understanding the timing of vegetation green-up and species composition in forests is critical for monitoring ecosystem dynamics and assessing how forests are shifting under climate change. This thesis investigates spring start of season (SOS) using remote sensing and develops a field-based species survey dataset for Frontenac Provincial Park (FPP). SOS was analyzed from MODIS-derived NDVI (2000-2023) and a gap-filled Landsat (1988-2018) time series, each with distinct workflows but using a 0.7 NDVI threshold. R package phenofit was applied to gap-fill the MODIS timeseries, while Landsat was processed independently to identify SOS DOY trends. Growing degree days (GDD) increased significantly (slope = 2.78; p = 0.009) from 1972-2023, reflecting regional warming. However, SOS DOY showed no statistically significant shifts in either dataset. For MODIS, other seasonal metrics—including end of season (EOS, p = 0.028) and length of season (LOS, p = 0.021)—showed significant changes, suggesting asymmetric phenological responses, with a lengthening fall rather than an earlier spring onset.
In parallel, a survey of 80 sites in FPP was conducted to capture tree species composition and generate a spatially explicit dataset. This dataset was integrated into an interactive StoryMap to visualize species distributions and support knowledge sharing. While a full species-specific classification was not undertaken, the dataset provides a foundation for future classification efforts as well as a baseline for understanding potential changes to forest composition in the future.
Collectively, this work provides insights into phenological trends and establishes a field-based dataset for FPP, offering a starting point for future research in species-specific classification or ecological monitoring. It emphasizes the value of combining remote sensing, field surveys, and interactive data products to improve understanding of forest vegetation dynamics and their responses to climate change.
Description
Citation
Publisher
License
Queen's University's Thesis/Dissertation Non-Exclusive License for Deposit to QSpace and Library and Archives Canada
Intellectual Property Guidelines at Queen's University
Copying and Preserving Your Thesis
This publication is made available by the authority of the copyright owner solely for the purpose of private study and research and may not be copied or reproduced except as permitted by the copyright laws without written authority from the copyright owner.
Attribution-NoDerivatives 4.0 International
Intellectual Property Guidelines at Queen's University
Copying and Preserving Your Thesis
This publication is made available by the authority of the copyright owner solely for the purpose of private study and research and may not be copied or reproduced except as permitted by the copyright laws without written authority from the copyright owner.
Attribution-NoDerivatives 4.0 International