Abstract
This study deals with the accuracy of GPS surveys in forest. Several factors affect the accuracy of GPS static positioning using carrier phase analysis. The degree of spatial openness without obstruction caused by tree canopies or slopes surrounding a survey station affects the station's ability to receive satellite signals; the position dilution of precision and changes in the epochs of satellites receivable may affect the survey results. This study consists of two series of observations, one of 1-hour surveys at 43 stations and another of 7-hour surveys at 15 stations. The logged data for the 7-hour surveys were separated into fourteen 30-minute data sets, nine 45-minute data sets, and seven 60-minute data sets as different trials. Baseline processing was used to produce a covariance matrix of accuracy. The second set of observations was used to calculate the distance between the baseline coordinates and the average coordinates calculated using the seven 60-minute data sets. It was found that the degree of openness without obstruction affected the accuracy of the survey. We estimated the 90% confidence limits using an exponential formula. Consequently, when surveying in forest, the degree of spatial openness without obstruction should exceed 80%. This guarantees that at least the amount of epochs logged is the same as 4 satellites receivable for all of the survey period and PDOP is less than 7 for at least 2/3 of the survey period. Difference in the length of the survey period did not have a clear effect on the average coordinates, but shorter periods showed increased variance.
