05.11.2016 13:38

Twenty years after


Twenty years ago, NASA approached the biodiversity community to explore the possibility to build a research program in NASA that would support biodiversity research. It was a time, where the state of the art have been single Landsat scenes purchased at around USD600 and thus a veritable bottleneck both in conservation and research. One of the main interests of the biodiversity community has thus been free access to the Landsat imagery, and implicitly other remote sensing products. The argument made has been that through this, the distribution of species and ecosystems could be plotted on this unique reference system (georeferenced images), and at the same time this will help planning of trips, conservation management and much more. 

What happened on the NASA side is well known and extremely encouraging and productive. Not only did our community  get open access to the Landsat images, but a wealth of new sources, composite images. This not only from NASA, but all the members in GEOSS

What happened on the biodiversity side is still in development. Millions of specimens are being digitized and label data converted and made accessible at institutional level and in many cases made available through GBIF. The main bulk of new data is now being produced through citizen scientist and covers mainly birds. Whilst the latter is comparable to NASA and their partners stream of new Earth observations, both in term of being real time and with a very high geospatial resolution, the former is like NASA going back into their archives and converting early remote sensing data into a format that can be read today. Whilst today’s spatial resolution is down to few meters, old Landsat images had 30m resolution which for many uses is still good enough. A lot of the collection data being georeferenced has a resolution of many kilometers at best. There is still a way to go to complement the bird community with additional, regularly sampled taxa to get an adequate sample for global biodiversity monitoring proposed by Globis-B, EU-BON or IPBES. The development of taxonomic publication seems to be a good way to see the development to samples taken with GPS reads.

After twenty years we are finally able to fully automatically find taxonomic papers on the Web, grab them, process them to extract hundreds of facts including geographic information to taxonomic treatments and visualize the content. Furthermore, we can disseminate the observation records to GBIF, the illustrations to the Biodiversity Literature Repository, so that programs like GEOSS have access to this bit of information: Every point on the map can now be traced back to the source data, publication and the actual specimen cited therein, if available. 

These observations are not yet “big data” per se, but often the only evidence that there is a little-known species out there in the world. However, the data provided in the taxonomic treatments add up to tens of millions of facts about the species. But to bring the remote sensing data together with somebody’s observation on the ground (accompanied with a GPS read) is just spectacular. And it works now routinely and available at our daily list of new treatments or the Twitter alert.  The results  displayed not least depend on what the authors and editors provide in their articles, and how the scientists collected in the field.

These developments over the last twenty years will hopefully be stimulating enough to launch a needed global monitoring programs on the ground  to understand and with that conserve this unique assett of planet Earth.