Vegetation recovery in Saint Barthélemy after Irma

Last year, in this post, I showed the comparison of two Sentinel-2 images of Saint Barthélemy in the Caribbean before and after the powerful Hurricane Irma.

A new feature in the EO Browser enables to plot the evolution of the mean NDVI within a polygon. I drew a rough polygon of Saint Barthélemy to check the evolution of the vegetation after Irma from Sentinel-2 data.

Time series of the average Normalized Difference Vegetation Index in Saint-Barthélemy extracted from Sentinel-2 observations

Here I used L1C data but it is also possible to use the L2A products from ESA, although these data are not always available. I manually adjusted the cloud fraction to remove the most obvious artifacts in the mean NDVI due to cloud contamination (clouds cause abrupt drops in the NDVI) [1]. This nice tool is sufficient to see that the vegetation quickly recovered after the hurricane, in about 1 month [2]. Catastrophic disturbances like hurricanes are actually known to contribute to maintain tree species diversity in tropical regions [3].
In the cities, according to Le Point, most of the damages have been repaired and the island is almost back to normal. This is good news for the people of St Barth!
Notes and references
[1] Under the hood, it's a "local area cloud detection algorithm based on the Braaten-Cohen-Yang method" Milcinski, G. Multi-year time series of multi-spectral data viewed and analyzed in Sentinel Hub. Medium, Apr 5, 2018.
[2] This is very similar to what has been observed in other tropical areas, e.g. "a sudden drop in NDVI values after Hurricane Maria’s landfall (decreased about 0.2) which returns to near normal vegetation after 1.5 months", Hu, T., & Smith, R. B. (2018). The Impact of Hurricane Maria on the Vegetation of Dominica and Puerto Rico Using Multispectral Remote Sensing. Remote Sensing, 10(6), 827.
[3] Vandermeer, J., de la Cerda, I. G., Boucher, D., Perfecto, I., & Ruiz, J. (2000). Hurricane disturbance and tropical tree species diversity. Science, 290(5492), 788-791.

Erratum to "Landslides in Kyrgyzstan captured by Sentinel-2"

In this post "Landslides in Kyrgyzstan captured by Sentinel-2" I showed two landslides that I identified as the Kurbu and Ayu landslides, both featured in the AGU landslide blog. However, as pointed out by Isakbek Torgoev, the Kurbu-Tash landslide is located about 5 km farther south than the one I found on the Sentinel-2 image of Apr 29. The actual Kurbu-Tash was not visible in this image due to cloud cover, but today I found a more recent image (May 19), where the Kurbu-Tash landslide is evident.. and it's even more massive! From this image I estimate that the runout distance of the Kurbu-Tash is about 7 km and not 3.5 km.
Continue reading