Domension Bogart Contessa Sofa

Time series of NDVI in the oak forests of the Sierra Nevada

plants to perform photosynthesis, actively absorb only a portion of the visible light spectrum (called PAR: Photosynthetically Active Radiation ) , covering wavelengths from 0.4 to 0.7 microns (micrometers). Wavelengths between 0.7 and the 1.1 microns (corresponding to the near infrared) are reflected by the leaves, because they negatively affect the photosynthetic apparatus from excess heat.

This difference in absorption / reflection of light is useful for determining photosynthetic activity of plants from remote sensing. A remote sensor captures the light reflected from the surface of the earth, and stores the intensity values \u200b\u200bof different wavelengths in different bands. For example, the sensor VEGETATION on SPOT 4 and 5 work with four bands (global expansion at a spatial resolution of about 1100 meters per cell, with daily temporal frequency): B0

• 0.43 to 0.47 microns (blue)
• B2 0.61 to 0.68 microns (red)
  
• B3 0.79 to 0.89 microns (near infrared)
  
• MIR 1.58 - 1.75 microns (medium infrared)
  

With these data, the best way to get an estimate of photosynthetic activity of vegetation is to use NDVI (Normalized Difference Vegetation Index) is calculated by the equation: NDVI = (B3-B2) / (B3 + B2). Possible outcomes have a range from -1 (no activity) and +1 (maximum activity).



To avoid the masking effect produced by clouds, images of VEGETATION sensor are processed to retain the maximum value of NDVI in periods of 10 days, with the S10 products (ten-day Synthes), available free on free.vgt.vito.be .

I downloaded the series on the European continent in the 1999-2010 period to focus on NDVI analysis of the oak forests of Quercus pyrenaica in Sierra Nevada, consisting of 432 images in HDF . GRASS GIS is a specific module to import these images called i.in.spotvgt , which processes them ready for use.


From this set of images, I have analyzed the time series of NDVI of the different populations Quercus pyrenaica in Sierra Nevada (grid of black border on the map above). A byproduct of this analysis is the following video, which shows the series of NDVI images of Sierra Nevada, and graph the maximum, average and minimum for all populations.




To create this video I used:

• GRASS to generate maps.
• R (pack Zoo) to generate graphs of time series.
• Imagemagick to unite couples of images of the same dates.
• avi mencoder to generate all images.
• Audacity to edit audio (although there was much room for improvement ...)
  
• pitivi to generate the final video and audio transitions.
  

A more interactive animation (and with better quality image) you can see here (expected to load all the pictures, and press the play button and not see anything, patience).

Later I will discuss the test results, and if there is popular demand, I will prepare the scripts for the video.

Greetings!