Methodology:

GNSS/INS buoy system was assembled by the University of Huddersfield. Each buoy system consists of a miniature inertial system (Spatial), a data storage (ILU), a GNSS antenna and a power supply. These sensor components were supplied from Advanced Navigation (advancednavigation.com) and Innovelec (innovelec.co.uk) (see Instrumentation for detailed information on the sensors), and were then assembled in a weather-proof plastic case for deployment. A total of six sets of the GNSS/INS buoys were shipped to the expedition, but the number of the buoys deployed on the ice varied throughout the expedition due to the logistical and environmental circumstances. The deployment of the buoys was made in the vicinity (within 2 km) of RV Polarstern at the MOSAiC drifting site. Kinematic correction requires the base station measurement should overlap the period of the rover measurement, so the availability of the kinematically corrected data is affected by environmental and operational limitations (see Temporal coverage for the data availability). This data set provides high-precision (centimetre accuracy) and high-frequency (10 Hz) ice dynamic data to study ice drift, lead/ crack opening/ closing and ridging of sea ice with the precision of a few centimetre scales.

Data collection:

Each GNSS sensor unit consists of Spatial, Data Interface Logging Unit (ILU), and antenna.

Spatial is a rugged miniature GPS aided inertial navigation system and AHRS that provides accurate position, velocity, acceleration and orientation. It combines temperature calibrated accelerometers, gyroscopes, magnetometers and a pressure sensor with advanced GNSS receiver to deliver accurate and reliable navigation and orientation (see advancednavigation.com/solutions/spatial/). The position accuracy can be a few centimetres as long as the antenna has a good sky view without any obstructions.

Data Interface Logging Unit (ILU) is a device server that interfaces to Spatial and provides data logging, a web interface, a time synchronisation server and different data input/output option.

Two types of GNSS antennas were used: Antcom (G5ANT-53A4T1) and Tallysman (TW3972).

Data quality:

The kinematic correction of the raw GNSS/INS data was conducted using Kinematica from Advanced Navigation (advancednavigation.com). The raw GNSS/INS data (in ANPP format) was converted to RINEX format for a kinematic correction with a moving base station. The processed data contain Moving Base Fix Type (i.e., the type of GNSS fix between the rover and the base station). For the data quality, the data with Fix Type* = 7 (i.e., RTK fixed - Carrier phase ambiguities have been resolved) were selected in this dataset.

Note that Leg5 data are not included in this dataset. No changes were observed in Leg5 and some errors were found during the data processing.

*Fix Type: Fix type 0 - No Fix; 1 - 2D; 2 - 3D fix (no base station data); 3 - SBAS; 4 - Differential (Base station data does not include carrier phase corrections); 5 - Omnistar; 6 - RTK Float (Carrier phase ambiguities cannot be resolved); 7 - RTK Fixed (Carrier phase ambiguities have been resolved).