Improved Troposphere Blind Models Based on Numerical Weather Data

The troposphere blind model RTCA MOPS is the minimum operational performance standard for global positioning systems. With a standard deviation of 2.3% of the ZTD, it enables us to mitigate the main part of the tropospheric effect on GNSS signals. Nevertheless, the comparison of RTCA MOPS with modern troposphere models like the ESA model or GPT2 shows the limitation of RTCA MOPS and points out the potential of modern troposphere blind models based on climatological series derived from numerical weather data. The ESA model profits from a more advanced wet delay model and a higher spatial resolution. GPT2 shows the smallest mean bias on surface level in comparison to ray-tracing and IGS data and profits from additional mapping function coefficients – especially if the user is interested in tropospheric delay at low elevation angles. A revision of GPT2 called GPT2w combines the benefits of both aforementioned models. Copyright # 2014 Institute of Navigation. TROPOSPHERE CORRECTION MODELS When passing the neutral atmosphere (in particular the troposphere, the lowest layer of the atmosphere) GNSS signals experience a path delay dependent on the variation of the refractive index due to temperature, pressure, and water vapor content. The tropospheric delay can reach up to a few tens of meters for very low elevation angles. Hence, it is a limiting factor for most space geodetic applications. Several troposphere correction models have been developed to mitigate the tropospheric effect on GNSS signals. The troposphere blind model RTCA MOPS developed by Collins in 1999 [1] is the recommended troposphere model for Satellitebased Augmentation Systems (SBAS). RTCA MOPS [2] is based on a set of tabulated climatological data (pressure of air p, temperature T, water vapor pressure e, temperature lapse rate α, and vapor pressure decrease factor λ) and therewith is easy to implement and operable without any further information about the actual state of the atmosphere. The meteorological parameters are derived as mean values and annual amplitude from the U.S. Standard Atmosphere Supplements (1966). Its values are given in tabular form for five latitude belts (15°, 30°, 45°, 60°, and 75°). Variations of p, T, e, α, and λ are modeled as annual signals. The zenith hydrostatic delay at mean sea level (ZHD0) is calculated by the formula of Saastamoinen [3]. ZHD0 m 1⁄2 1⁄4 10 6 k1 Rd p gm (1) where k1 is the refraction coefficient for dry air (77.604 KhPa ), Rd is the constant of dry air (287.054 Jkg K ), gm is the mean gravity (9.784ms ) and p is the pressure of air at mean sea level in hPa. The zenith wet delay (ZHD0) at mean sea level can be obtained by the modified approach of [4]. ZWD0 m 1⁄2 1⁄4 10 6 k3 Rd gm λþ 1 ð Þ α Rd e