Terrestrial Laser Scans of the on-ice Remote Sensing instrument site on the Multidisciplinary Drifting Observatory for the Study of Arctic Climate (MOSAiC) in April 2020 (Leg 3)

• Main Author: David Clemens-Sewall
• Format: Dataset
• Language: unknown
• Published: Arctic Data Center 2022
• Subjects: Snow; Sea Ice; Terrestrial Laser Scanning; LiDAR; Arctic; Arctic Ocean; Sewall; Sea ice
• Online Access: https://doi.org/10.18739/A25717P9Q

These data are superseded by the following dataset, which has superior alignment quality: https://doi.org/10.18739/A26688K9D. Data are available for download at: https://arcticdata.io/data/10.18739/A2QJ78024/ These Terrestrial Laser Scanning (TLS) data were collected on April 17, 22, and 30, 2020 at the Multidisciplinary Drifting Observatory for Arctic Climate (MOSAiC) Expedition. The MOSAiC Expedition continuously made in-situ measurements on a collection of drifting sea ice floes In the Arctic Ocean from October 2019 to May 2020. TLS, also known as Terrestrial Light Detection and Ranging (LiDAR), is an active remote sensing modality in which a tripod-mounted, rotating laser emitter and detector (the TLS sensor) scans the surroundings emitting pulses of light and tracking the time it takes for a pulse to return. From the time of flight and the orientation of each pulse, the sensor creates a point cloud of the surroundings. The TLS sensor used here was a Riegl VZ1000, and the data were acquired using Riegl proprietary RiSCAN software. The data have been converted into entirely open-source formats (see below). This dataset contains point clouds of the same unit of sea ice acquired on different days. In addition to the raw data, we provide the rigid transformations needed to align all data into a common, ice-fixed reference frame (see below). Within this data, we can quantify snow deposition and erosion within the different footprints of the remote sensing instruments and elsewhere. These data are provided in a directory structure that contains meaningful information about each file. To download these data in this directory structure, please press 'Download All'. Beneath the top level directory, is a directory for the TLS scan collected on each day (e.g. 'mosaic_rs_170420.RiSCAN' is the data collected on April 17, 2020). Within each scan directory are the files 'tiepoints.csv' and 'ScanPosXXX.DAT' where XXX is the scan number (e.g. '008'). Additionally, there are subdirectories: 'lasfiles', 'npyfiles', and 'transforms'. 'lasfiles' contains the raw data collected by the scanner and written out in the LAS 1.4 format, a standard open file format for LiDAR data: https://www.asprs.org/committee-general/laser-las-file-format-exchange-activities.html. The raw data are collected in the orientation of the TLS sensor on the tripod, which is not exactly level due to human error (most operators can level the tripod to within 2 degrees). The scanner contains built-in level sensors, and from these the RiSCAN software computes the 4x4 rigid transformation matrix (https://en.wikipedia.org/wiki/Transformation_matrix) that levels the raw data. These 4x4 matrices are written out in ascii format in the 'ScanPosXXX.DAT' files. The raw data do not distinguish between TLS points from wind-blown snow particles (a major source of noise in TLS scans from snow in windy environments) and otherwise. These data have been processed and analyzed by FlakeOut (Clemens-Sewall 2021: https://github.com/davidclemenssewall/flake_out/tree/v1.0.0 or https://zenodo.org/record/5657286#.YZRRcLtOlH4) to label wind-blown snowflakes with the Classification flag '65'. The processed data are provided in the 'npyfiles' subdirectory. The processed data have been saved in the open NumPy format (.npy: https://numpy.org/devdocs/reference/generated/numpy.lib.format.html). This format has several advantages over LAS 1.4 in terms of the speed of reading and writing data and is more widely used in the earth science community than LAS 1.4. Finally, although the scans collected on different days were examining the same area of the ice, the tripod cannot be located in precisely the same location each time, and so transformations are need to align the scans into a common, ice-fixed reference frame. This reference frame is anchored by a set of tiepoints that were manually identified to be the same object in each scan (and whose locations are available in the 'tiepoints.csv' files). The rigid transformations are provided in .npy format in the 'transforms' subdirectory.