Variable Resolution Simulations over the Tibetan Plateau: insights from MPAS-A transitioning across Convective-Permitting, Gray Zone, and Quasi-Hydrostatic Scales

This dataset provides gridded atmospheric variables used to analyze the sensitivity of precipitation and thermodynamic processes to horizontal resolution in variable-resolution MPAS-A simulations over the Tibetan Plateau. The data support analyses presented in the accompanying manuscript and are intended to ensure full reproducibility of the results. Target grid and horizontal resolution All datasets are provided on regular latitude–longitude (lat–lon) grids to enable grid-to-grid comparison among MPAS-A configurations and reference products. The primary analysis grid is 0.05° × 0.05°, used for most hourly variables and diagnostics (e.g., precipitation, CAPE, moisture flux and moisture flux divergence, and event-based composites). Daily near-surface temperature fields (2-m air temperature: mean / maximum / minimum) are provided on a 0.10° × 0.10° lat–lon grid. Source data and regridding strategy MPAS-A native output is produced on an unstructured SCVT mesh (cell-centered variables). All MPAS-A fields distributed here have been interpolated/remapped from the native MPAS mesh to the target lat–lon grids listed above. Reference datasets are also provided on the same target grids: ERA-Interim (native coarse lat–lon grid) is regridded to the target grid for direct comparison. CMORPH and IMERG precipitation products are regridded to the target grid. Temporal resolution Hourly precipitation. Hourly CAPE. 6hr Surface moisture flux and moisture flux divergence (and related moisture transport diagnostics where provided). 6hr Event-based/composite diagnostics (see below), aligned to hourly sampling. Daily (Δt = 1 day) products: 2-m air temperature (daily mean, daily maximum, daily minimum) on the 0.10° grid. Reinitialization experiments and time coverage consistency All MPAS-A outputs come from reinitialization (RI) experiments with effective horizontal resolutions of approximately 60 km, 60–10 km, and 60–3 km. The released time series are organized to match the analyses in the accompanying manuscript (e.g., precipitation amount, diurnal cycle, event evolution, thermodynamic instability, and land–atmosphere moisture exchange), with consistent timestamps and sampling intervals across model configurations and reference datasets. Event-based composite diagnostics (0–200 h relative time) Composite diagnostics are constructed relative to precipitation onset (defined in the manuscript) and are provided on a relative time axis spanning 0–200 hours. These composites support analyses of event frequency and intensity and their co-evolution with thermodynamic and surface–atmosphere coupling diagnostics, sampled at hourly resolution on the 0.05° grid. Directory and file organization Each subdirectory corresponds to one physical variable or one diagnostic family (e.g., hourly precipitation, daily T2, CAPE, moisture flux/divergence, composites). Within each subdirectory (or .tar file), separate NetCDF files are provided for each MPAS-A configuration (60 km, 60–10 km, 60–3 km) and for each reference dataset (ERA-Interim, CMORPH, IMERG), ensuring one-to-one comparability on the same grid and time sampling.