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 (<1%) were filled via linear interpolation.Global Analysis: ERA5 spatial data were processed as global averages or gridded datasets in NetCDF files, producing hourly, monthly, and yearly time series to capture global water vapor trends. TIFF files provide spatial visualizations of TCWV.Standardization: TCWV data are stored in NetCDF format with spatial-temporal grids; supplementary data are organized in CSV format, with each row corresponding to a date and columns including LOD, CO2, lunar phase, orbital parameters, and sunspots. Units are explicit (e.g., mm, ppm, degrees). TIFF files are high-resolution images documenting analysis results.Quality Control: Outliers (e.g., physically implausible TCWV values) were removed through cross-validation with source data. All processing steps are documented in the metadata.Data AnalysisDynamic Quantification Model: A dynamic quantification model, based on proximity difference methods, was developed to eliminate Earth’s internal variability and quantify the impact of orbital parameters (Earth rotation, lunar revolution, Earth revolution, etc.) on TCWV variations.LSTM Prediction: A Long Short-Term Memory (LSTM) deep learning model was trained using orbital parameters and supplementary data as inputs to predict global TCWV trends over the next decade. The model was trained on 1959–2022 time series, with cross-validation for robustness.File StructureMain File:tcwv_era5.nc: NetCDF file containing global TCWV data for 1959–2022 (hourly, monthly, and yearly), with 0.25°×0.25° spatial resolution, including time, latitude, and longitude dimensions.Visualization File:tcwv_spatial.tiff: TIFF file providing global spatial visualizations of TCWV, reflecting monthly or yearly trends.Supplementary File:orbital_auxiliary_data.csv: Daily time-series data including LOD, CO2 concentrations, lunar phase, Earth-Sun distance, solar and lunar declination angles, and sunspots.Metadata: Variable definitions, units, sources, and processing steps are detailed in the Dryad submission form and this README.Research AreasThis dataset pertains to the following research areas:Climate Science: Analyzing global water vapor variations and climate context to study orbit-driven climate trends.Planetary Science: Exploring the impact of planetary orbital changes on Earth’s systems using orbital parameters and lunar phase data.Geophysics: Investigating Earth rotation and water vapor interactions through LOD data.Data Science and Machine Learning: Applying LSTM deep learning and dynamic quantification models for water vapor trend prediction.