Forest canopy, terrain, and distance effects on global positioning system point accuracy

Tests were conducted to determine the realizable accuracies of the Global Positioning System under eastern North American forest conditions. The effects of terrain, forest canopy, number of consecutive position fixes, and PDOP on accuracy were evaluated. Position accuracies were determined for a total of 27 sites: three replicate sites selected for all combinations of three canopy (deciduous, conifer, open) and three terrain (ridge, slope, valley) types. Each site was visited over a span of nine months to collect position data, for ten replications per each of 27 sites to collect 60, 100, 200, 300, and 500 position fixes. The mean differentially corrected positional accuracy for all sites was 4.35 metres, with 95 percent of the mean positions estimated within 10.2 metres of the true value. The least accurate differential position data were observed at conifer sites. Positional accuracy was higher for deciduous sites and highest at open sites. Mean positional accuracy increased from valley to ridge locations. Mean accuracy increased with increasing number of position fixes collected per point. When the number of position fixes increased from 60 to 500, mean accuracy increased from 5.9 to 3.1 m under deciduous canopies, from 6.6 to 4.4 m under conifer canopies, and from 3.9 to 2.2 m under open skies. The average time required by the GPS receiver to lock onto four satellites and begin collecting positions varied from one to two minutes, and collection times increased from open, through deciduous, to conifer sites. There was an observed, but statistically non-significant, trend between accuracy and the field receiver's distance from the base station. Nine replicates of 300 position fixes were averaged for six sites, which ranged from 43 kilometres to 247 kilometres from a base station. Mean accuracy ranged from 1.48 metres to 2.43 metres.