Litter Survey – Methodology
This document summarizes a current methodology used for conducting litter audits in various Canadian and USA jurisdictions.
MGM Management has conducted, or has planned 15 major litter audits including:
* Ontario – conducted under supervision of Dan Syrek, 1990
· Ontario – Toronto area 1994, done by McKenney with Syrek assistance
· Newfoundland Litter Audit - 1994
· City of Toronto - Street Litter Audit 2002
· Regional Municipality of Peel - Street Litter Audit 2003
· Regional Municipality of York - Street Litter Audit 2003
· Regional Municipality of Durham - Street Litter Audit 2003
· City of Toronto – Street Litter Audit 2004
· City of Toronto – Parks Litter Audit 2004
· City of Toronto - Street Litter Audit 2005
· City of Toronto - Street Litter Audit 2006
· City of San Francisco Street Litter Audit – April 2007
· Alberta Rural Highways Litter Characterization – May 2007
· City of Edmonton - Street Litter Audit 2007 (June 2007)
The Toronto 2002; 2004, 2005 and 2006 Litter Audits were conducted using our methodology. The Regions of Peel, York and Durham are major urban municipalities around Toronto – these surveys were also conducted in 2003 using our methodology. We conducted audits in 2007 for the City of San Francisco, CA; and for the City of Edmonton. Also in 2007, we conducted a litter characterization study on behalf of the Recycling Council of Alberta (RCA), funded in part by Alberta Environment, examining litter along rural roads and highways in that province. Our survey methodology uses proven and recognized techniques to randomly locate litter survey sites and for counting 84 ty pes of large litter, and 16 categories of small litter (small litter was not be included in the Alberta RCA project).
In the USA - over 30 litter count surveys have been done by Syrek, (and reviewed by MGM Management). More recently the University of Florida at Gainesville has completed five excellent surveys across all of the 29 counties of Florida. MGM Management has been trained in the methods of both the Syrek and in the Florida techniques, directly by those researchers.
In some instances, local environmental groups have done litter counting. These methodologies may not, in some cases, be scientific in their development and tend not to be reproducible. Measurement techniques need to be unbiased, scientifically rigorous, and reproducible to be defensible. Comparison to other jurisdictions is not usually possible with local methods. This methodology's approach can be reproduced and compared. This survey uses statistically proven and recognized methods of identifying litter survey sites and for counting litter.
In 1993 the Florida Legislature directed the Florida Centre for Solid and Hazardous Waste Management to conduct a statewide litter count. The Centre had developed a method for surveying litter that was understandable, simple and statistically valid. MGM Management has conducted litter audits using the principles of this methodology.
The method counts "accumulated litter". This is in contrast to "fresh litter" counts, where a sight is cleaned, then researchers return after a set time to count the number of pieces of litter that have been deposited. Accumulated litter allows for an examination of the occurrence of litter as it is has developed over time. Fresh litter count surveys are much more labour intensive than accumulated litter counts.
Site Selection Process
Random Site Selection
In selecting sites to survey it is important to have an unbiased method of selection. The survey teams are not allowed judgment in the field in selecting sites, but rather this is done using a pre-selection technique. In this way, neither the "dirtiest" nor the "cleanest" locations are picked. The survey teams count litter at sites that are selected well in advance of traveling to the location.
To select sites for a given jurisdiction a geographical information system (GIS) database is acquired (software used was ArcView GIS 3.2 by Environmental Systems Research Institute Inc.). Using the program, centre-line coordinates for all potential public street locations within the jurisdiction is selected.
Then using a random number generator feature of a spreadsheet program (Microsoft Excel) samples of this data are assembled. The data locations outputted centre-line locations for a statistically significant sample of potential sites.