Back

Mapping vegetation types in Antarctic Peninsula and South Shetlands islands using Sentinel-2 images and Google Earth Engine cloud computing

Fonseca, E. L. d.; Santos, E. C. d.; Figueiredo, A. R. d.; Simoes, J. C.

2021-09-15 ecology
10.1101/2021.09.14.460232 bioRxiv
Show abstract

The Antarctic vegetation maps are usually made using very high-resolution images collected by orbital sensors or unmanned aerial vehicles, generating isolated maps with information valid only for the time of image acquisition. In the context of global environmental change, mapping the current Antarctic vegetation distribution on a regular basis is necessary for a better understanding of the changes in this fragile environment. This work aimed to generate validated vegetation maps for the North Antarctic Peninsula and South Shetlands Islands based on Sentinel-2 images using cloud processing. Sentinel-2 imagery level 1C, acquired between 2016 and 2021 (January-April), were used. Land pixels were masked with the minimum value composite image for the "water vapor" band. The NDVI maximum value composite image was sliced, and its classes were associated with the occurrence of algae (0.15 - 0.20), lichens (0.20 - 0.50), and mosses (0.50 - 0.80). The vegetation map was validated by comparing it with those from the literature. The present study showed that Sentinel-2 images allow building a vegetation type distribution map for Antarctica Peninsula and South Shetlands Islands.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.