The GPS contribution to the error budget of surface elevations derived from airborne LIDAR

When using airborne LIDAR to produce digital elevation models, the Global Positioning System (GPS) positioning of the LIDAR instrument is often the limiting factor, with accuracies typically quoted as being 1030 cm. However, a comprehensive analysis of the accuracy and precision of GPS positioning of aircraft over large temporal and spatial scales is lacking from the literature. Here, an assessment is made of the likely GPS contribution to the airborne LIDAR measurement error budget by analyzing more than 500 days of continuous GPS data over a range of baseline lengths (3960 km) and elevation differences (4002000 m). Height errors corresponding to the 95th percentile are $$ 0.15 m when using algorithms commonly applied in commercial software over 3-km baselines. These errors increase to 0.25 m at 45 km and $\,$0.5 m at 250 km. At aircraft altitudes, relative heights are shown to be potentially biased by additional errors approaching 0.2 m, partly due to unmodeled tropospheric zenith total delay (ZTD). The application of advanced algorithms, including parameterization of the residual ZTD, gives error budgets that are largely constant despite baseline length and elevation differences. In this case, height errors corresponding to the 95th percentile are $$ 0.22 m out to 960 km, and similar levels are shown for one randomly chosen day over a 2300-km baseline. © 2006 IEEE.