Time Series Analysis of Municipal Solid Waste Generation Rates

Abstract

A methodology for data analysis and stochastic modeling of daily municipal solid waste production rates is presented. The data sets examined are the daily quantities of municipal solid wastes for consecutive days and for each day separately. Each sequence of observations was modeled by Box-Jenkins stochastic models as a function of autoregressive, moving average, and seasonal terms. For the overall time series, a seasonal ARMA model (1, 0) × (1, 1)5 was found to be adequate. The observed seasonality of length 5 was due to the municipal solid waste (MSW) collection pattern. For the separate daytime series, simple autoregressive (AR) models were adequate without inclusion of any seasonal terms. In general, these models demonstrated statistical fit, and modeling of the trends was satisfactory. The forecasting ability of the Box-Jenkins models was compared to simpler statistics, such as the mean value and the moving average values. Depending on the specific day, different models gave optimum forecasting results.