PROMICE: Greenland ice velocity maps v2026

Abstract. We present an updated version of the Programme for Monitoring of the Greenland Ice Sheet (PROMICE) ice velocity product (https://doi.org/10.22008/FK2/K70OPK; Solgaard and Kusk (2026)), providing a continuous time series of Greenland Ice Sheet velocity mosaics from January 2016 to the present. The product is derived from Sentinel-1 synthetic aperture radar (SAR) data, gridded at 200 m spatial resolution and updated every 12 days using data spanning two consecutive Sentinel-1 repeat cycles (24 days). Data are typically released within 10 days of the final acquisition and include all valid 6- and 12-day image pairs within the 24-day window.

This update includes several important improvements to the processing chain. The spatial resolution has been refined from 500 m to 200 m, justified by the implementation of an adaptive correlation template size approach for offset tracking, improving velocity retrievals and enhancing delineation of narrow outlet glaciers. We further implement a new mosaicking strategy, which reduces noise associated with ionospheric disturbances. Additional improvements include enhanced error handling and outlier rejection. The full processing workflow is described, including data selection, mosaicking, uncertainty estimation, and filtering procedures.

Validation against in-situ GNSS measurements over the full time series shows that the standard deviation of the differences between satellite- and GNSS-derived velocities (with corresponding bias) is 22 m/yr (-0.3 m/yr) and 38 m/yr (-0.4 m/yr) for the easting and northing components, respectively. These values fall within expected ranges, although a substantial fraction of the discrepancy likely reflects uncertainty in the GNSS measurements. This interpretation is supported by validation over stable terrain, where substantially lower values are obtained: 9 m/yr (0.1 m/yr) and 15 m/yr (-0.1 m/yr) for the easting and northing components, respectively. Compared to the previous product version, uncertainties are higher due to a prolonged period when only one Sentinel-1 satellite was operational, resulting in increased noise and reduced temporal sampling. We quantify the impact of these conditions on spatial coverage.

Overall, coverage is highest during winter, when radar coherence is strong and acquisitions are most comprehensive, whereas summer coverage is reduced due to surface melt. Despite these seasonal and mission-related constraints, the PROMICE ice velocity product provides consistent temporal sampling and broad spatial coverage, supporting investigations of ice-sheet-wide and glacier-specific dynamics and ice discharge on seasonal to multi-year timescales.