Raw ERA5-Land vs NOAA ISD Lite Global Temperature Comparison Records (62-Station Roster, January 2024-December 2025)

Raw ERA5-Land 2 m temperature vs NOAA ISD Lite surface observation comparison records, backing Section 5 and Table 3 of the corresponding Geodesy and Geodynamics manuscript (companion Supplementary Tables S1-S2). This file contains the complete, unfiltered set of 2,826 station-date query results from a 62-station global roster (45 of which returned contributing valid data), January 2024-December 2025, three dates per month (5th/15th/25th), 12:00 UTC target time. Two exclusion filters are specified in the manuscript methodology (Section 5.1) and must both be applied to reproduce the published statistics (N=2,644; global median 1.38 degC, IQR 0.67-2.55 degC): 1. Failed ERA5 retrievals (T_era5 = NaN): 178 records, from 3 stations that returned no valid data at all (CYVR, KBOS, WSSS). 2. Non-physical temperature differences (|delta_T| > 20 K, indicating an ingestion or coordinate error): 4 records, all from one station (FMEE, Reunion Island), traced to an isolated data-ingestion fault producing an implausible near-freezing surface reading (1-2 degC) at a tropical station whose ERA5 value and all other 55 dates correctly show 25-31 degC. This was identified during final data-archiving verification; see the manuscript's Response to Reviewers (Comment 7) for full documentation. Applying both filters (182 records total excluded) yields the 2,644 valid comparisons reported in the manuscript. Field schema: icao (station ICAO code), date (YYYY-MM-DD), season (meteorological, hemisphere-adjusted), climate (Koppen-style zone label), region, lat, lon, elev (m), T_metar (surface temperature in degC; despite the field name, this is sourced from NOAA ISD Lite, not live METAR reports - the field name is a legacy artifact of the generating script), T_era5 (ERA5-Land 2 m temperature, degC), delta_T (T_metar - T_era5, signed), abs_delta_T (absolute value of delta_T). Published as-is (unfiltered) for full transparency and independent reproducibility, including of the exclusion process itself.