Posted by: norstarnewengland | July 31, 2016

Picture Post Project – Tracking Change in Foliage Color

One of the previous blogs had featured the Picture Post Project for Science Friday’s Science Club’s Take a Sample to Answer a Question about the World.  In that blog, were two graphs showing the change in the percentage of red, green, and blue, from the end of March to the end of May, one from 2016 and the other from 2015.  They are included here, again.

GPH_SciFriSampEv_BHT5SS_RGBIndex_Spring2016

GPH_SciFriSampEv_BHT5SS_RGBIndex_Spring2015

The color index is based on the fact that the color on every digital image is a combination of the brightness of each of three color sensors, red, green, and blue, that make up a pixel.  The idea is that though every color is a combination of three colors, more vegetation would result in higher values in the green, and less of the other two.

These graphs were created from the south facing images uploaded to the Picture Post web site for Blue Hill Tower.  First, images from each week were opened in Irfanview (a photo editor application), and the same area was clipped and exported.  Then, in the Picture Post’s application, Analyzing Digital Images, the clipped areas were analyzed to determine the percent of red, green, and blue in each pixel to the sum of red, green and blue values. The result is the value in the y axis, plotted for each image, arranged chronologically in the x axis.

The end of Science Friday’s challenge was May 13, so the graph created was up to the weekend before.  Thus, the results showed no change in the relative positions.  This seemed to be because leaf out progression was not at the point where photosynthesis had really become established.  The following is a graph that continues the analysis to the end of May:

Graphs_IndexesS

This graph is similar to the graph for 2015.  The main difference is when the green value crosses the red.  In this year’s graph, it occurred on 5/21, while the previous year was between 5/12 and 5/19.  Thus, the change is about a week later than the previous year.  While pictures are taken about a week apart, and, thus, due to this ‘temporal resolution,’ the difference is basically ‘within the error of the data,’ the images do show that the method used is viable, and, with more frequent visits, differences may be detected.

The next blogs will introduce the Picture Post project and each of the sites on Great Blue Hill.

 

 

Advertisements

Categories

%d bloggers like this: