Document Type

Journal Article

Department/Unit

Department of Biology; Croucher Institute for Environmental Sciences

Abstract

Waste generation projections inform waste policy formulation and are an indispensable process in waste management planning. However, not only is valid waste generation projection difficult, their reliability is also difficult to prove. Between the two major methodological approaches in forecasting municipal solid waste generation, the time-series approach only uses past data and their distribution to determine future waste trends. The factor model, on the other hand, explains and predicts waste trends with explanatory variables such as socioeconomic factors of the waste generators. This latter approach not only aims at making predictions on waste quantities, it also aims at unveiling hypothetical causal relationships between factors for the prediction of waste quantities. In this article, results of waste generation projection studies conducted between 1989 and 1999 by Hong Kong's environmental authority on domestic waste growth were verified against actual waste data for determining the accuracy of these predictions. It was then followed by the use of a factor-model based technique, autoregression, to forecast domestic waste growth for Hong Kong based on historical (1979–2007) waste data and other socioeconomic factors. Although the use of multiple factor autoregression model appeared to rectify the overestimation tendency of classical linear regression model used by Hong Kong's environmental authority, a number of empirical constraints that are also typical of other factor-model-based techniques were encountered. It is essential that waste policy makers are aware of, these constraints when they are making decisions based on the results from such models.

Publication Year

2010

Journal Title

Environmental Engineering Science

Volume number

27

Issue number

1

Publisher

Mary Ann Liebert

First Page (page number)

13

Last Page (page number)

20

Referreed

1

DOI

10.1089/ees.2009.0106

ISSN (print)

1092-8758

Keywords

municipal solid waste, waste projection, classical linear regression, autoregression, factor models, time series models

Included in

Biology Commons

Share

COinS