Ground Penetrating Radar Data, Ice thickness fields and bedrock topography maps of Perito Moreno, Viedma and Upsala Glacier

Data acquisition The dataset was collected during two helicopter-borne GPR campaigns conducted in March to April 2022 and October 2024. In total, six survey flights were carried out: three over Perito Moreno Glacier, two over Viedma Glacier, and one over Upsala Glacier, covering a combined distance of 393 km (171 km Perito Moreno, 149 km Viedma, 73 km Upsala). We used a shielded bistatic radar system with a center frequency of 25 MHz, suspended beneath a helicopter as a sling load. Surveys were flown at 15–20 m above the glacier surface at ~70 km h⁻¹. The ADC converter digitized signals at 400 MHz (4096 samples per trace), stacking 256 pulses per trace and recording at 10 Hz, resulting in an along-track spacing of ~1.94 m (≈500 traces km⁻¹). A GNSS rover mounted at the antenna center provided precise positioning, later synchronized with the GPR internal clock during post-processing. Ground Penetrating Data Processing All helicopter-borne GPR data were georeferenced using two Leica GS16 multi-frequency GNSS receivers operated in a base–rover setup. Base-station coordinates were determined via precise point positioning (PPP) with ±0.05 m accuracy, and the rover trajectory was post-processed kinematically to ±0.10 m. This ensured that all radar measurements were accurately tied to a global coordinate system. Prior to processing, GPR and GNSS data were time-synchronized by correcting the 18 s GPST–UTC offset and aligning GNSS timestamps with the radar’s internal clock. GNSS data logged at 1 Hz were trimmed to full-second intervals for matching. The radar data were processed in REFLEXW v8.1 (Sandmeier Geophysical Research) following a uniform workflow applied to all flights. The processing included trace repositioning, subsetting into transects, zero-time correction, background noise removal, bandpass filtering (10–40 MHz), amplitude compensation, manual layer picking, 2D migration using a velocity model (0.3 m ns⁻¹ in air, 0.168 m ns⁻¹ in ice), air-layer correction, and interpretation of bedrock reflections. Final two-way travel times were converted to ice thickness and merged into a single, fully processed dataset. Ice thickness data and bedrock topography maps Corresponding .tif files include all three glaciers. The ice thickness reconstruction was performed in two main steps. First, the processed GPR-derived thickness measurements were spatially interpolated using a physically based reconstruction algorithm that accounts for glacier flow and mass continuity. In a second step, the interpolated thickness fields were used to infer the underlying bedrock topography by subtracting the modeled ice thickness from the glacier surface elevation. This approach ensures a consistent and physically plausible distribution of ice thickness and bedrock geometry across each glacier. For more details on the reconstruction see: Fürst et al. 2017 and Fürst et al. 2024. The bedrock elevation files have a resolution of 20 x 20 m (UTM 18S). The ice thickness fields have a resolution of 100 x 100 m (UTM 18S). Further details can be found in the corresponding publication in Earth System Science Data: PLACEHOLDER TEXT