by Laurent Lebreton, data scientist at dumpark
A recent study on marine plastic pollution published in the journal PlosOne estimates that some 5.25 trillion plastic particles weighing more than 268.000 tonnes are floating on the world’s oceans.
Over a six-year period, researchers from six different countries collected plastic density data from 24 expeditions into the five main oceanic basins and various marginal seas. By trawling nets for microplastics and using systematic visual sightings for the larger macroplastics, count data was collected for 1571 field locations and weight data for 1333 of those.
The dimensionless numerical model that we developed earlier and wrote about here on Sciblogs last year has now been calibrated to fit the reported measurement data via a linear system of equations of the form:
where Y is a series of observations, S a matrix of dimensionless model outputs for K scenarios and N observations, and β and ε respectively vectors of computed weighting coefficients and error terms.
This method can be used to fit an arbitrary number of model output cases to any number of measured data points producing a weighting coefficient and error term for each case. For this study, we used a set of three model results (K = 3), corresponding to three different input scenarios:
- urban development within watersheds,
- coastal population and
- shipping traffic.
To visualise the model-predicted concentration, we produced an interactive map that shows the weight density estimates as a dot density map, where each dot represents 20 kg of floating plastic, totalling more than 13 million dots globally. We calculated the number of dots for each model cell -an area of around 400 km²- and randomly distributed them within each cell using a circular Gaussian function.
The map also displays the different expeditions that participated in the data collection effort from 2007 to 2013. For each individual expedition stage, heatmaps show the measured plastic densities for every survey location and four debris size classes.
Based on conservative fragmentation rates, researchers originally expected to find more small particles than larger ones, but this study surprisingly shows that the smallest fragments are generally less abundant than the second smallest size, but are often found in remote regions outside of the garbage patches.