0000000000925924
AUTHOR
Ennio Ottaviani
Random Forest model and TRIX used in combination to assess and diagnose the trophic status of Bizerte Lagoon, southern Mediterranean
International audience; A combined multimetric trophic index (TRIX) and the Random Forest (RF) model were used to characterize the trophic status of Bizerte Lagoon. The RF model was used to build a predictive model of chlorophyll a using physicochemical variables (nitrite, nitrate, ammonium, phosphate, oxygen, temperature and salinity) as predictors. The approach is based on physicochemical and biological parameters measured in samples collected twice weekly from January to December 2012 at one representative sampling station located at the lagoon center.The observed TRIX values vary from 5.18 to 6.12, reflecting waters ranging from moderate to poor quality with a high trophic level. The re…
Machine learning predictions of trophic status indicators and plankton dynamic in coastal lagoons
Abstract Multivariate trophic indices provide an efficient way to assess and classify the eutrophication level and ecological status of a given water body, but their computation requires the availability of experimental information on many parameters, including biological data, that might not always be available. Here we show that machine learning techniques – once trained against a full data set – can be used to infer plankton biomass information from chemical and physical parameter only, so that trophic index can then be computed without using additional biological data. More specifically, we reconstruct plankton information from chemical and physical data, and this information together w…