Date Approved
8-20-2025
Graduate Degree Type
Thesis
Degree Name
Biology (M.S.)
Degree Program
Biology
First Advisor
Robert Hollister
Second Advisor
Jeremy May
Third Advisor
Rob Larson
Academic Year
2024/2025
Abstract
Numerous studies have shown changes, both increases and decreases, in Normalized Difference Vegetation Index (NDVI) across Arctic ecosystems. Higher NDVI values have been associated with larger deciduous shrubs and increases in graminoid cover. We used the differences in NDVI collected throughout the growing season to estimate the above-ground abundance, cover and biomass, of different plant functional types at Atqasuk, Alaska. We examined NDVI at three critical times during the growing season: an initial measurement a few days after snowmelt, the second occurred shortly after leaf-out for deciduous shrubs, and again at peak season. We also examined the relationship between biomass and plant cover, measured with a point frame. To model cover and biomass, plant functional types (PFT’s) with similar phenological green up patterns were combined together. The combinations of PFT’s include “evergreen’s” (evergreen shrubs, lichens and bryophytes), “non-evergreen’s” (deciduous shrubs, graminoids and forbs), “all plants” (combination of all types). We also attempted to split the non-evergreens into graminoids and deciduous shrubs. Each combination of plant functional types (e.g. all plants, evergreens, non-evergreens, deciduous shrubs, and graminoids) cover and biomass were modeled using the hypothesized difference in NDVI relevant to them. Each model was cross validated using leave one out cross validation, and the error between predicted and observed values was calculated to assess each model. This process yielded significant models for both cover and biomass respectively: All Plant (R2: 0.71, R2: 0.50), Evergreen (R2: 0.65, R2: 0.67), Non-Evergreen (R2: 0.69, R2: 0.20), deciduous shrubs (R2: 0.62, R2: 0.42), and graminoid (R2: 0.27, R2: 0.19). The hypothesized differences in NDVI provided accurate models, except for the graminoid models, where a different relationship better modeled cover and biomass (R2: 0.70, R2: 0.80). Future work will see if this approach can be applied across the North Slope of Alaska and elsewhere in the Arctic. This method has the potential to characterize vegetation quickly and allows for large spatial coverage. Monitoring vegetation change is fundamental to understanding the changes taking place across the Arctic ecosystem. This method provides a novel, intermediate approach between ground sampling and higher spatial scale remote sensing techniques.
ScholarWorks Citation
Doorn, Taylor A., "Using Seasonal NDVI to Predict the Aboveground Abundance of Tundra Plants" (2025). Masters Theses. 1158.
https://scholarworks.gvsu.edu/theses/1158
