COS 222-5 - Predicting tree mortality using spectral indices derived from multispectral UAV imagery

Background/Question/Methods

Past research has shown that remotely sensed spectral information can be used to predict tree health and vitality. Recent developments in unmanned aerial vehicles (UAVs) have now made it possible to derive such information at the tree and stand scale from high resolution imagery. Using this data could allow us to predict tree mortality across large areas and improve our understanding of tree mortality.
We used visible and multispectral bands from UAV imagery to calculate a set of spectral indices for 52 845 individual tree crowns within 38 forest stands in western Canada. We then used those indices to predict mortality of these canopy trees over the following year using random forest and logistic regression algorithms. We evaluated whether including multispectral indices leads to more accurate predictions than indices derived from visible wavelengths alone, and how performance varies among three different tree species (Picea glauca, Pinus contorta, Populus tremuloides).

Results/Conclusions

Our results show that spectral information can be effectively used to predict tree mortality, with a random forest model producing a mean AUC of 89.8% and a balanced accuracy of 83.3%. The exclusion of multispectral indices worsened model performance, but only slightly (AUC = 87.9%, balanced accuracy = 81.8%). We found variation in model performance among species, with higher accuracy for the broadleaf species (balanced accuracy = 85.2%) than the two conifer species (balanced accuracy = 73.3% and 77.8%) . Despite the fact that all models overestimate tree mortality, our results demonstrate that imagery from UAVs has strong potential for predicting annual mortality for individual canopy trees and can be used to assess tree mortality risk on forest stand level. Further improvements could include the use of hyperspectral data, training the model with a larger and more balanced data set, using cost-sensitive learning algorithms, and testing its performance in new regions.