While California has just started to emerge from its worst drought in 1,200 years (2011-2016), it is now being hit by severe rainfalls, causing the evacuation of 188’000 northern California residents under the threat of spillway failure at Oroville Dam.

 

The Sentinel-2 mission began to monitor the western USA in 2016, so we can make an image comparison of the Lake Oroville in January 2016 (after five years of drought) and 2017 (after the heavy rainfalls). Here I used the NIR color composition (B8/B4/B3) in the Sentinel-Playground the highlight the contrast between the water and the vegetation surfaces.

Sentinel-2A images of Lake Oroville (NIR/red/green color composites)

As explained in this previous post, the normalized difference water index (NDWI) enables to better detect the lake extent.

Sentinel-2A images of Lake Oroville (colormap based on the NDWI)
Sentinel-2A images of Lake Oroville (colormap based on the NDWI)

Here the NDWI is computed from level-1 images (top of the atmopshere reflectances). We can see that the water mask derived from the NDWI is not perfect in this mountainous region, as shaded (north-facing) slopes also have high NDWI values and can be confused with water surfaces. A better result could probably be achieved with level-2 images (top of canopy, flat surface reflectances).

5 thoughts on “Lake Oroville in January 2016 vs. 2017

  1. Could be interesting then to process those images with sen2cor and obtain the level-2 images with the SNAP toolbox? I would like to do it with Python

    1. Absolutely! You can download the latest release of sen2cor for free and process the images yourself. Don’t hesitate to share your results with us here!

  2. I will certainly do it then!! by the way I have just added you in Twitter, you always post interesting Sentinel-2 stuff 🙂

Répondre à antonio vecoli Annuler la réponse

Votre adresse e-mail ne sera pas publiée. Les champs obligatoires sont indiqués avec *

Ce site utilise Akismet pour réduire les indésirables. En savoir plus sur comment les données de vos commentaires sont utilisées.