Positioning precision and sampling number of DGPS under forest canopies

Experimental results were studied to determine the relationships between positioning precision of a differential global positioning system (DGPS) and forest type, antenna height, and season, and to clarify the relationship between sampling number and the convergence of positioning precision. Observation was carried out for 24 h. Mean circular area probability (CEP95) was 2.80 m for deciduous broadleaved trees, and 4.99 m for conifers. The mean CEP95 taken at 7 m height (3.14 m) was higher than that at 1 m height (3.92 m) at all sites. The mean CEP95 taken during the defoliation season (2.65 m) was slightly better than during the foliation season (2.96 m). There were significant differences between forest types (P < 0.001) and antenna heights (P < 0.05). Positioning precision was not noticeably improved if the sampling number was around ten or less. A sampling number of 100–1000 or more is required before substantial improvements can be expected. As long as high positioning precision is not required, it is acceptable to use 2D & 3D modes and relatively few samples to take measurements.