Modelling residential solid waste production in Amsterdam-Oost using demographic and socio-economic data

Organised by Laboratory of Geo-information Science and Remote Sensing

Thu 29 September 2016 09:30 to 10:00

By Steven Schrauwen (the Netherlands)


The objective of this research was to generate a better understanding of the drivers of solid waste production in Amsterdam by analysing the waste collection measurements of 2015 for temporal patterns and applying a regression analysis with socio-economic and demographic data. The waste is divided in four waste streams: paper, glass, plastic and mixed. Each of these waste streams was analysed separately. First time series were generated by recalculating the waste collection measurements to daily waste production per bin. Then the time series were summed per neighbourhood. For the regression analysis a stepwise multiple linear regression method was applied. The result of the time series analysis showed that considerable variation exists in the day to day waste production, 20 to 43%, but no systematic variation throughout the year. The regression analysis generated four models with varying performance. The mixed waste model had the highest R2adj (0.903) and the plastic waste performed the poorest (0.357). Overall it is concluded that population size is the main driver on a neighbourhood scale and other variables had little influence, this is likely due to the neighbourhood scale that averages out the differences per household.