| 1. |
- Montaghi, Alessandro, et al.
(författare)
-
Analysis of effects of scanning angle on ALS-derived vegetation metrics in a nationwide airborne ALS acquisition
- 2012
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- In the summer of 2009, the Swedish government, under the co-ordination of Lantmäteriet (the Swedish mapping, cadastral and land registration authority), started a five-year project (2009- 2013) using airborne laser scanning for production of the New National Elevation Model (in Swedish: Ny Nationell Höjdmodell, NNH) for all of Sweden (450,000 km2). The primary aim of this project is to produce a 2 meter grid DEM (Digital Elevation Model) in which the standard error is better than 0.5 m. A total of about four hundred scanning areas, with a size of 25 km by 50 km, are being scanned with a nominal density of 0.5-1 point per square meter, and with a maximum scanning angle of ± 20 degrees (Petersen and Burman Rost, 2011). The acquisition is primarily done using Leica Geosystems ALS50-II and ALS60 sensors, with Optech ALTM Gemini as a complementary sensor. In each scanning area, 21 parallel flight lines were flown, with a nominal overlap of 20%, in addition also three perpendicular crossing lines where acquired. This flight line arrangement was designed to allow strip adjustment techniques, based on sensor parameter calibration, to be used for creation of a seamless final product (Toth, 2009). Finally, main and cross strips were merged together and delivered in blocks of 2.5 km by 2.5 km. The ALS data acquired for the NNH project is a resource of interest also for forest estimation. However, since the ALS survey is being carried out for a purpose other than measuring forest parameters, there are a number of issues that need to be considered. These include the effect of different time points for the scanning, the relatively large view angles used, and the positional accuracy for the NFI plots. In this paper, the effect of one of these issues, the view angle influence on LiDAR forest metrics, will be reported. At the conference also initial results from predictions based on LiDAR data and SPOT HRG satellite data, trained with national forest inventory sample plots, will be presented.
|
|
