Cross-Comparison and Methodological Improvement in GPS Tomography

Abstract

GPS tomography has been investigated since 2000 as an attractive tool for retrieving the 3D field of water vapour and wet refractivity. However, this observational technique still remains a challenging task that requires improvement of its methodology. This was the purpose of this study, and for this, GPS data from the Australian Continuously Operating Research Station (CORS) network during a severe weather event were used. Sensitivity tests and statistical cross-comparisons of tomography retrievals with independent observations from radiosonde and radio-occultation profiles showed improved results using the presented methodology. The initial conditions, which were associated with different time-convergence of tomography inversion, play a critical role in GPS tomography. The best strategy can reduce the normalised root mean square (RMS) of the tomography solution by more than 3 with respect to radiosonde estimates. Data stacking and pseudo-slant observations can also significantly improve tomography retrievals with respect to non-stacked solutions. A normalised RMS improvement up to 17% in the 0-8 km layer was found by using 30 min data stacking, and RMS values were divided by 5 for all the layers by using pseudo-observations. This result was due to a better geometrical distribution of mid-and low-tropospheric parts (a 30% coverage improvement). Our study of the impact of the uncertainty of GPS observations shows that there is an interest in evaluating tomography retrievals in comparison to independent external measurements and in estimating simultaneously the quality of weather forecasts. Finally, a comparison of multi-model tomography with numerical weather prediction shows the relevant use of tomography retrievals to improving the understanding of such severe weather conditions. Global Positioning System (GPS) tomography considers the use of slant-integrated estimates, wet delays, or corresponding water vapour content estimated from the data records of ground-based GPS stations to respectively retrieve the 3D field of wet refractivity or water vapour density, as introduced by [1,2]. Comparisons of tomography retrievals with other techniques (e.g., those using a water vapour radiometer, radiosonde, raman lidar, or atmospheric emitted radiance interferometer) and with numerical weather models have shown relevant results and an encouraging understanding of meteorological situations ([3-17]) The resolution and configuration/geometry of the network of GPS stations are critical parameters with which to obtain the best scenario for applying GPS tomography to retrieve water vapour density or wet refractivity. These fields can be ideally retrieved for meteorological applications with a horizontal resolution of a few kilometres, a vertical resolution of~500 m in the lower troposphere, and a vertical resolution of~2 km in the upper troposphere, with a time resolution of 15 to 5 min. However, to obtain this remarkable geometrical resolution, data from a dense homogeneously distributed network of GPS stations (e.g., a netw...

Analyzing Different Parameterization Methods in GNSS Tomography Using the COST Benchmark Dataset

“…In addition, the effect of the topography of the area in tomographic modelling was studied. As a fact, tomography quality is a complex function of geometry (observations distribution, model resolution, time interval), inversion, regularization algorithms, and initial conditions used in the model [16]. Hence, all the research efforts that try to establish an understanding of these complex problems are of interest to the GNSS tomography community.…”

“…Instead, determination of water vapour using the Global Navigation Satellite Systems (GNSS) is a low cost and effective technique with reasonable precision and more spatio-temporal resolution than was provided by the previous techniques. Over past two decades, the potential of using GNSS to determine the 4D wet refractivity and water vapour fields using tomography has been evaluated in various studies [14][15][16][17][18][19][20][21][22][23][24][25][26]. GNSS tomography is an all-weather condition remote sensing technique that can be used to reconstruct the water vapour or the wet refractivity in the Earth's atmosphere.…”

“…In this method, the atmosphere is modelled through a finite number of three-dimensional elements (voxels). As recent studies show, the tomography solution [16] is very sensitive to the a priori data and much less dependent on the number of observations used in a single epoch. One of the clear limitations of a few models (e.g.…”

“…One of the clear limitations of a few models (e.g. TOMO2) tested by Brenot et al, [16] was a simplification of the parameterisation: the signal path was parameterised as a straight line and no topography model has been considered. Therefore, the signal path in the model is, on average, shorter than in reality, which introduces biases [27].…”

50 anos de sinergia entre Geodésia Espacial e Meteorologia: do erro no posicionamento GNSS a aplicações de previsão de precipitação de curtíssimo prazo

“…Um estudo recente propõe o uso de receptores GNSS para prever (com antecedência de 30 minutos) eventos extremos de precipitação (>35mm) (SAPUCCI et al, 2019) e há projetos visando explorar dados do COSMIC-2 (constelação de satélites equatoriais dedicados para RO-GNSS) para prever eventos extremos de precipitação (BANOS et al, 2018). Um desenvolvimento em paralelo vem sendo a tomografia do vapor d'água (BRENOT et al, 2020), com potencial de melhor detalhar a distribuição temporal e espacial dessa variável (MOELLER et al, 2020). É consenso que previsão de tempo em condições atmosféricas extremas (severas) é desafiadora e encontra-se na fronteira do conhecimento humano, no qual a sinergia em diferentes áreas do conhecimento é um dos ingredientes mais importantes para se obter o sucesso.…”