?? This app shows the snow cover duration in the Sierra Nevada (Spain) since January 01 in comparison with the average over the same period of the year between 2001-2019. The snow cover was obtained from Nasa’s Terra/MODIS satellite observations. The app is automatically updated when new observations are available.

?? Cette application montre la durée de l’enneigement dans la Sierra Nevada espagnole depuis le 01 janvier par rapport à la moyenne sur la même période de l’année entre 2001-2019. L’enneigement a été obtenu à partir des observations satellitaires Terra / MODIS de la Nasa. L’application est automatiquement mise à jour lorsque de nouvelles observations sont disponibles.

?? Esta aplicación muestra la duración de la capa de nieve en la Sierra Nevada desde el 01 de enero en comparación con el promedio durante el mismo período del año entre 2001 y 2019. La capa de nieve se obtuvo de las observaciones satelitales Terra / MODIS de la NASA. La aplicación se actualiza automáticamente cuando hay nuevas observaciones disponibles.

Click here to view in full screen.

Below is an example on Mar 18, the snow cover duration was much lower than the average due to the exceptionally warm winter.

Method

  • The daily NDSI is extracted from the MOD10A1.006 collection (~500 m spatial resolution). Missing values are filled using a linear interpolation in the time dimension [1].
  • The resulting series of daily gap-free NDSI is converted to a series of binary snow maps (snow / no-snow) using the threshold NDSI>0.2.
  • The snow duration is the sum of the snow maps along the time axis.
  • The snow maps over 2001-2019 were pre-computed and stored as an asset to reduce the app loading time (but it’s still slow).

Earth Engine users can get the source code in my Apps repository:

git clone https://earthengine.googlesource.com/users/sgascoin/apps

[1] Missing values account for about 50% of the pixel.days in the Pyrenees, mostly due to cloud cover. Although more sophisticated algorithms exist to fill gaps in MODIS snow cover maps, the linear interpolation in the time dimension is a good trade-off between efficiency and accuracy.

This work was done as part of the CLIM’PY / OPCC project (FEDER/Poctefa) https://opcc-ctp.org