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.

Visualizing the vegetation dynamics in Africa

The Global Imagery Browse Services (GIBS) is a great new API provided by the NASA to access its vast collection of Earth Observation products. For example the following URL allows you to download a jpeg picture of the Terra MODIS NDVI on 2018 Jan 01 over Africa:,-37.6,53.1,39.0&epsg=4326&layers=MODIS_Terra_NDVI_8Day&worldfile=false&format=image/jpeg&width=820&height=872

By looping on the "TIME" parameter (in a bash script using wget) I could get all the available products from 2016-07-30 until yesterday, and make this animation (using ffmpeg).

Continue reading

Sentinel-2 adds colors to the 2017 super bloom in California

Texas had the super bowl in 2017 but California got the super bloom!


After its worst drought in 1,200 years, California is now experiencing one its wettest year. The Sierra Nevada snowpack is currently one of the biggest ever recorded. In February, severe winter rainfalls caused the evacuation of many California residents under the threat of the Oroville Dam failure.


Now, this unusual level of rainfall triggered the blossoming of flowers that were in a dormant state. The Washington Post collected some photographs and noted that it was so so massive that you can see it from space (images were provided by Planet, a California-based remote sensing company).


How does it look like in Sentinel-2 imagery? Below is the comparison between two Sentinel-2 images acquired in spring 2016 vs. spring 2017 near the Carrizo Plain National Monument (see by yourself in the sentinel-playground).


Continue reading