Modelling and forecasting of beach litter assessment values

Walvoort, Dennis; Loon, Willem van; Schulz, Marcus; André, Silvère


A model has been developed for forecasting total counts of beach macrolitter on European beaches. A negative binomial generalized linear model (NB-GLM) was used, because this model is suitable to model asymptotic trends and is a natural choice for count data with overdispersion. NB-GLMs were fitted for the Dutch-North Sea and German-North Sea beach litter. The model predicts that the threshold value (TV) could be reached in Germany in 2026 (90% confidence interval: 2023-2031) and in the Netherlands in 2047 (90% confidence interval: 2037 and later than 2050). The forecasts can be used to evaluate whether the current reduction rate is su cient, or if increased national and regional reduction measures may be needed.