Observed vegetation dynamics through social media


There are currently vast amounts of freely available data online that have yet to be analyzed for potential applications to the earth sciences. These data include millions of publicly available images taken across the globe. Though the original intent of these photos was likely not primarily to drive earth science related analyses, they nonetheless contain information regarding the earth system. In particular, many photos are taken of various landscapes, capturing the state of the local vegetation. An integration of the time series of these photos can be used to observe vegetation dynamics at regions around the world. This research presents a proof of concept analysis of photographs taken of US National Parks posted to Twitter. These photographs were processed per park, and a relative greenness index was calculated as a metric for vegetative dynamics. We compared these Twitter data to satellite observations of greenness index, and describe the potential value and shortcomings in using social media as a source of data in the earth sciences.