Dataset for Influence and Prediction of Planetary Orbital Changes on Earth's Atmospheric Water Vapor Variations

AbstractThis dataset supports the study of planetary orbital changes' impact on Earth's atmospheric water vapor variations and future trend predictions, spanning 1959–2022. Core data are sourced from ECMWF’s ERA5 reanalysis, providing hourly, monthly, and yearly total column water vapor (TCWV) at 0.25°×0.25° resolution. Supplementary data include IERS Length of Day (LOD), Scripps CO2 concentrations, Fourmilab lunar phases, NASA JPL Earth-Sun distance, solar and lunar declination angles, and NOAA/SIDC solar sunspot data. Provided in NetCDF and TIFF formats, with supplementary CSV data, the dataset enables dynamic quantification and LSTM-based predictive modeling to analyze and quantify orbit-driven water vapor trends across multiple timescales, advancing orbital-climate interaction research.MethodsData CollectionTotal Column Water Vapor (TCWV): Global TCWV data were extracted from ECMWF’s ERA5 reanalysis dataset, covering January 1959 to December 2022, with 1-hour temporal and 0.25°×0.25° spatial resolution, spanning the globe.Length of Day (LOD): Daily LOD data for 1959–2022 were obtained from the IERS Prediction Center, U.S. Naval Observatory, reflecting Earth rotation variations.CO2 Concentration: Daily atmospheric CO2 data for 1959–2022 were sourced from the Scripps Institution of Oceanography archives, serving as a climate context variable.Lunar Phase: Daily lunar phase information for 1959–2022 was collected from Switzerland’s Fourmilab website to assess lunar influences on water vapor.Orbital Parameters: Daily Earth-Sun distance, solar declination angle, and lunar declination angle for 1959–2022 were acquired from NASA JPL ephemerides, representing planetary orbital changes.Solar Sunspots: Daily sunspot data for 1959–2022 were obtained from NOAA/SIDC, evaluating solar activity impacts on water vapor.Data ProcessingPreprocessing: Hourly ERA5 TCWV data were aggregated into global monthly and yearly averages using mean calculations and stored in NetCDF format. Supplementary data (LOD, CO2, etc.) were aligned to a daily timestep. Missing values (