International Snow Science Workshop Proceedings - Montana State University Library

Item: SfM-based 3D point clouds in determination of snow depth from high-resolution UAS data as alternative methods: Is it possible to use?

• • Title: SfM-based 3D point clouds in determination of snow depth from high-resolution UAS data as alternative methods: Is it possible to use?

    Proceedings: International Snow Science Workshop Proceedings 2018, Innsbruck, Austria

    Authors:

    • Remzi Eker  [ Düzce University, Faculty of Forestry, Konuralp Campus, 81620 Düzce, Turkey ]
    • Abdurrahim Aydın  [ Düzce University, Faculty of Forestry, Konuralp Campus, 81620 Düzce, Turkey ]
    • Yves Bühler  [ WSL Institute for Snow and Avalanche Research SLF, Flüelastrasse 11, 7260 Davos Dorf, Switzerland ]
    • Andreas Stoffel  [ WSL Institute for Snow and Avalanche Research SLF, Flüelastrasse 11, 7260 Davos Dorf, Switzerland ]

    Date: 2018-10-07

    Abstract: The DEM of Difference (DoD) method is most common technique used to obtain snow depth (HS) maps obtained from remote sensing data such as unmanned aerial system (UAS) data. But in the present study, M3C2 method, an advanced point cloud comparison method, was tested. For this aim, data of two UAS flights carried out in the upper Dischma valley (Davos, Switzerland) were used: one for snow free data in 6th June 2016 and one for snow covered data in 17th May 2017. HS values also were able to be measured by using snow probe for only 7 points at same day of UAS flight. DoDbased and M3C2-based HS values were compared to reference data. While M3C2-based HS values had 12.3 cm of RMSE, DoD-based HS values had 5 cm of RMSE. Even though the quite less number of field data for the comparison of the results, HS maps were compared at 30 points located over boulders without snow cover, where the vertical difference between two DSMs is expected to be zero. According to results, M3C2-based HS values were obtained less than DoD method (19 cm in average). However, M3C2 distance at 30 snow free points had values that close to zero than DoD, as expected. While 90% of M3C2 distance values of those snow free points varied between -4 cm and 4 cm, only 17% of DoD values varied between -4 cm and 4 cm. All remaining points were obtained higher than 4 cm.

    Object ID: ISSW2018_P04.14.pdf

    Language of Article: English

    Presenter(s):

    Keywords: DoD, M3C2, Point cloud, Snow depth, UAS

    Page Number(s): 398-402

  • Get Full Article

Back to Home page