Scholarlink

<p>The global atmospheric model IGGatm is established from the gridded Chinese reanalysis data CRA-40 with 1° horizontal resolution and optimized with a flexible vertical grid and formulas. It can provide empirical three-dimensional fields of pressure, temperature, and water vapor pressure from the troposphere to the lower stratosphere for satellite positioning and related applications. Accurate slant atmospheric delays and mapping functions are calculated using the revised piecewise linear ray tracing method from the modelled atmospheric field. To reduce model parameter redundancy, the annual mean coefficients are represented on the non-uniform vertical grid determined from the tropopause height feature, and different polynomial fitting formulas are applied to characterize the seasonal amplitude coefficients across all latitude bands. The resulting optimized version, IGGatm Poly, maintains the same level of accuracy while reducing the parameter count to ~ 11% of the initial uniform grid version, which is about 1.3 times that of the GPT3 model. A revised extrapolation scheme combining state equation extrapolation with the U.S. Standard Atmosphere model (USSA) further ensures more reliable slant delay estimation. Validation of the IGGatm shows that the global mean root mean square (RMS) errors at the surface are approximately 6.5&#xa0;hPa for pressure, 4.9&#xa0;K for temperature, and 3.1&#xa0;hPa for water vapor pressure (vs. radiosonde), while the zenith total delay (ZTD) error is about 3.9&#xa0;cm (vs. GNSS tropospheric products). By the optimized vertical representation method, the IGGatm series captures vertical variations of the atmospheric parameter and ensures consistent accuracy across different altitudes. In contrast, the GPT3 model exhibits unrealistic biases at higher atmospheric levels. The performance of the mapping function estimation is verified through statistics on equivalent slant delays at 5° elevation angle. The global mean RMS errors of equivalent slant hydrostatic delay (SHD) and slant wet delay (SWD) estimations derived from the IGGatm Poly are 1.65&#xa0;cm and 0.83&#xa0;cm, respectively, for the surface levels of radiosonde stations. The IGGatm series achieves an average improvement of 22% in equivalent SHD accuracy compared with the GPT3 model, with the improvement reaching up to 50% in tropical regions. The IGGatm series also show slight improvement in equivalent SWD accuracy. Furthermore, the IGGatm series effectively captures the variation of the hydrostatic mapping function with height and also accounts for its latitude-dependent characteristics. Overall, the IGGatm series provide the convenience of complete atmospheric parameters and ensure consistent performance for higher levels. They facilitate both ground and airborne users and also show potential in low-orbit satellite-aided positioning, in which the GPT3 demonstrates systematic biases in low elevation-angle mapping function estimations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The global atmospheric model IGGatm based on Chinese reanalysis data and polynomial fitting formulas

  • 期刊论文

  • 发表日期：

    2026年1月7日

查看全文
  • Beibei Yu,
  • Wei Li,
  • Hongxing Zhang,
  • Yunbin Yuan,
  • Liu Hui

摘要

The global atmospheric model IGGatm is established from the gridded Chinese reanalysis data CRA-40 with 1° horizontal resolution and optimized with a flexible vertical grid and formulas. It can provide empirical three-dimensional fields of pressure, temperature, and water vapor pressure from the troposphere to the lower stratosphere for satellite positioning and related applications. Accurate slant atmospheric delays and mapping functions are calculated using the revised piecewise linear ray tracing method from the modelled atmospheric field. To reduce model parameter redundancy, the annual mean coefficients are represented on the non-uniform vertical grid determined from the tropopause height feature, and different polynomial fitting formulas are applied to characterize the seasonal amplitude coefficients across all latitude bands. The resulting optimized version, IGGatm Poly, maintains the same level of accuracy while reducing the parameter count to ~ 11% of the initial uniform grid version, which is about 1.3 times that of the GPT3 model. A revised extrapolation scheme combining state equation extrapolation with the U.S. Standard Atmosphere model (USSA) further ensures more reliable slant delay estimation. Validation of the IGGatm shows that the global mean root mean square (RMS) errors at the surface are approximately 6.5 hPa for pressure, 4.9 K for temperature, and 3.1 hPa for water vapor pressure (vs. radiosonde), while the zenith total delay (ZTD) error is about 3.9 cm (vs. GNSS tropospheric products). By the optimized vertical representation method, the IGGatm series captures vertical variations of the atmospheric parameter and ensures consistent accuracy across different altitudes. In contrast, the GPT3 model exhibits unrealistic biases at higher atmospheric levels. The performance of the mapping function estimation is verified through statistics on equivalent slant delays at 5° elevation angle. The global mean RMS errors of equivalent slant hydrostatic delay (SHD) and slant wet delay (SWD) estimations derived from the IGGatm Poly are 1.65 cm and 0.83 cm, respectively, for the surface levels of radiosonde stations. The IGGatm series achieves an average improvement of 22% in equivalent SHD accuracy compared with the GPT3 model, with the improvement reaching up to 50% in tropical regions. The IGGatm series also show slight improvement in equivalent SWD accuracy. Furthermore, the IGGatm series effectively captures the variation of the hydrostatic mapping function with height and also accounts for its latitude-dependent characteristics. Overall, the IGGatm series provide the convenience of complete atmospheric parameters and ensure consistent performance for higher levels. They facilitate both ground and airborne users and also show potential in low-orbit satellite-aided positioning, in which the GPT3 demonstrates systematic biases in low elevation-angle mapping function estimations.