Abstract
Data describing aircraft position and attitude are essential to computing return positions from ranging data collected during airborne laser scanning (ALS) campaigns. However, these data are often excluded from the products delivered to the client and their recovery after the contract is complete can require negotiations with the data provider, may involve additional costs, or even be infeasible. This paper presents a rigorous, fully automated, novel method for recovering aircraft positions using only the point cloud. The study used ALS data from five acquisitions in the US Pacific Northwest region states of Oregon and Washington and validated derived aircraft positions using the smoothed best estimate of trajectory (SBET) provided for the acquisitions. The computational requirements of the method are reduced and precision is improved by relying on subsets of multiple-return pulses, common in forested areas, with widely separated first and last returns positioned at opposite sides of the aircraft to calculate their intersection, or closest point of approach. To provide a continuous trajectory, a cubic spline is fit to the intersection points. While it varies by acquisition and parameter settings, the error in the computed aircraft position seldom exceeded a few meters. This level of error is acceptable for most applications. To facilitate use and encourage modifications to the algorithm, the authors provide a code that can be applied to data from most ALS acquisitions.
Highlights
Airborne laser scanning (ALS), or light detection and ranging (LiDAR) as this technology is alternatively known, has been used, and often relied upon [1], to generate detailed and precise digital descriptions of ground surfaces [2], model hydrologic systems [3], assess forest above-ground biomass [4], map snow depth [5], monitor powerlines [6], and measure terrain displacement induced by seismic events [7]
We investigated systematically the effects time block duration (∆t) and closest point of approach (CPA) intensity, defined as the time interval between points considered in the cubic spline optimization process detailed above, have on airborne platform trajectory estimates obtained for flight lines in each of the five study areas
The median residual is almost double for trimmed swaths and short ∆t, but slowly converges to that observed for complete swaths for higher ∆t values
Summary
Airborne laser scanning (ALS), or light detection and ranging (LiDAR) as this technology is alternatively known, has been used, and often relied upon [1], to generate detailed and precise digital descriptions of ground surfaces [2], model hydrologic systems [3], assess forest above-ground biomass [4], map snow depth [5], monitor powerlines [6], and measure terrain displacement induced by seismic events [7]. Over forested landscapes, the pulse angle is known to affect the computation of tree height, leaf area index, gap fraction, and canopy cover [21,22,23,24], all important parameters for forest resource assessment and planning. This is primarily because as the angle increases, the length of the pulse trajectory intersected by a tree crown increases, and with it, the probability for a crown return. This phenomenon, manifested as a horizontal displacement of ground returns with respect to returns from the canopy, is sometimes referred to as ‘LiDAR shadow’ and it is more pronounced with trees featuring a high leaf area index [13]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
