For my capstone project in machine learning at EPFL, I wrote a classifier capable of sorting 3D scans of archaeological objects by culture.
Digitization of museum collections is currently a major challenge faced by cultural heritage and natural history museums. Museums are expected to digitize the collections to improve not only the documentation of artifacts, but also their availability for research, reconstruction and outreach activities, and to make these digital representations available online.
Much of biodiversity is discovered in museum collections, sometimes years after the specimen has been collected. Ploughing through expedition notes and logs is then required and therefore having a way to summarize the contents of a large text corpus can be very interesting. In this example, I will graphically summarize “On the Origin of Species” by Charles Darwin (it seemed a suitable choice) to demonstrate this technique.
In this post, I will use a divergent color scale to plot two distributions on the same map. As an example, I chose to plot the European distribution of two species of corvids: the carrion crow (Corvus corone) and the hooded crow (Corvus cornix). There has been some adjustments to the taxonomical status of the hooded crow (see Parkin et al., 2003 for details), hoewever, currently, they are regarded as different species.
In this map, I will use a divergent color scale to show areas in Europe where each species is dominant, and also show areas where both species are present.
In a previous post, I discussed how to plot GBIF occurrence data using OpenStreetMaps. Here, I will plot a distribution map. Distribution maps differ from occurrence maps in that occurrences are aggregated and plotted as a heat map. Additionally, the map has to be projected using an equal area projection. I will illustrate these two features by plotting the distribution of the tawny owl (Strix aluco) in Europe.
In a previous post, I discussed how to plot occurrence data from GBIF on a map. In this post, I will discuss how to plot a bird migration by producing an occurrence map for each month of the year. I will use the migration of the stork (Ciconia ciconia) as an example.
The Global Biodiversity Information Facility (GBIF) is a data aggregator for biodiversity data. The big advantage of using an aggregator like GBIF over getting data directly from the original data source is that an aggregator provides a single point of entry to many data sets, so analysing one data set is technically interoperable with any other data set.
Spectrograms are a common visualization of sound data. Visualizing sound data can be useful when doing a presentation or for publication. Additionally, machine learning algorithms for classifying sound data generally use spectrograms as their starting point, instead of the sound data itself, as many advanced algorithnms for classifying images are readily available. The example uses the R packages warbleR (Araya-Salas & Smith-Vidaurre, 2017), seewave (Sueur, Aubin, Simonis, 2008) and tuneR (Ligges et al., 2018).
This example draws the spectrogram of the call of a tawny owl (Strix aluco).
For a project on noise pollution in the oceans at the Natural History Museum in Berlin, I recently made this plot of the hearing and vocalization ranges of selected marine animals. Range plots are generally not-so-common plots. In this example, I plotted the hearing and vocalization range (frequency) for several species of whales, dolphins, seals, turtles and fishes.
If you are using R, then you are probably familiar with the mtcars data set, that is used in many R tutorials. The “car analogy” is so common in text books, that this so-called technique has its own Wikipedia page. For the rest of us, who don’t understand anything about motor cars, R-core comes with a wide selection of example data sets, some of which relate to ecology or biodiversity.