A Global Troposphere Model for VLBI Simulations

VLBI simulation studies play an important role in various applications, including optimizingstation network designs, exploring innovative concepts, advancing technologicaldevelopments, and testing observation strategies. Realistic simulations are crucial forinformed decision-making and must account for various error sources such asmeasurement noise, clock drift, and atmospheric effects. Accurate simulation oftropospheric effects, particularly incorporating spatio-temporal correlations, is essential inthis context.Traditionally, tropospheric simulations rely on Kolmogorov turbulence theory combinedwith the frozen flow assumption. These simulations are parameterized using the refractiveindex structure constant (Cn), alongside auxiliary parameters like wind velocity andtroposphere height. Although Cn values can be derived from GNSS observations, mostexisting studies assume a generalized average troposphere, independent of specificlocations or seasonal variations. Only a few incorporate simplistic, location-basedconditions, often through a basic latitude-based interpolation. Furthermore, reliance onGNSS data limits the ability to test potential network extensions where such observationsare unavailable.This work advances tropospheric simulations by developing a global, three-dimensional(latitude, longitude, time) Cn model. The model is based on Zenith Wet Delay (ZWD)estimates from 21,000 GNSS stations (2000-2023) and meteorological data from theERA5 reanalysis model, including specific humidity across 11 pressure levels (300 to 1000hPa) and wind velocity. It is also adaptable for short-term prediction scenarios using HRESweather forecasts, making it suitable for operational applications.The proposed model manages to capture both spatial and temporal patterns adequately,resulting in more realistic simulations. Additionally, it can be used to derive global monthlyaverage estimates on a 0.25° x 0.25° grid, offering a practical and sufficiently accuratesolution for most simulation studies. The model's performance is further validated throughcomparisons with real VLBI observations, demonstrating significant improvements insimulation accuracy.