Radioholographic Technique For GPS/LEO Radio Occultation

Atmospheric parameters of the Earth can be resulted from inverting GPS(Global Positioning System) signals by employing the GPS/LEO (Low Earth Orbit)radio occultation (RO) technique. This technique can provide atmospheric profileswith high accuracy, near-real-time, high vertical resolution, all-weather capability,global coverage, low expense, long-term stability, etc. It is independent of traditionalatmospheric detection techniques, and can be used as a tool of comparision andsupplentment to the traditional atmospheric detection techniques. The GPS/MET(METeorology) program demonstrated the potential of GPS/LEO RO technique inmonitoring the atmosphere of the Earth. Reduction of both random and systematicerrors was found by applying RO data to numerical weather forecast and climateanalysis assimilation system. GPS/LEO RO data can be used in climate and globalchange monitoring, space environment detection, and atmosphere modelimprovement. Moreover, it is of great importance in other research areas, such asmeteorology, geodesy, geodynamics, space weather, etc.Retrieving of atmospheric parameters from RO data often encountersdifficulties in the moist lower troposphere. Complicated structure of atmosphericrefracitivity, caused mainly by the water vapor, results in the degradation of theaccuracy of retrieved refractivity. The inversion of atmospheric parameters can alsobe affected by signal scattering, attenuation, noise, multipath propagation,superrefraction effect, etc. Various radioholographic methods for processing RO datain atmospheric multipath zones have been proposed to improve retrieval accuracy inthe moist lower troposphere: back propagation (BP) method, sliding spectral (SS)method, canonical transform (CT) method, full spectrum inversion (FSI) method. Inthis dissertation, the theoretical background, key points, accuracy and limitation ofdifferent radioholographic methods will be discussed in detail. The atmosphericpropagation of GPS signals under multipath conditions and their detection are simulated by using multiple phase screens technique. Under the assumption of idealsignal, bending angles computed by the different radioholographic method arecompared with corresponding solutions to the true value obtained with the Abelintegral. The above process can be used to assess the performance of differentmethods under multipath conditions. The accuracy of radioholographic methods isalso validated throught refractivity comparisons between COSMIC (ConstellationObserving System for Meteorology, Ionosphere and Climate) occultation andECMWF (European Centre for Medium-Range Weather Forecasts) analysis.The main works of the present dissertation can be summarized as follows:1、 Different low Earth orbit satellite missions, such as GPS/MET, CHAMP(CHAllenging Minisatellite Payload), GRACE (Gravity Recovery and ClimateExperiment), SAC-C (Satelite de Aplicaciones Cientificas-C), COSMIC, areintroduced. Standard inverse algorithm for neutral atmoshphere is alsodescribed in detail.2、 Under multipath conditions, the propagation of GPS signals in the neutralatmosphere is simulated by multiple phase screens technique. The signalsimulation is useful in studying phenomena such as multipath propagation,suprefraction effect, noise, etc.3、 The BP method is confirmed to be able to reduce multipath effect to someextent under multipath conditions and is not very sensitive to the noise.Optimal selection of auxiliary plane position is a key problem in the BPmethod. Simulation results show that the retrieved accuracy is related to theposition of auxiliary plane, which can not be determined in advance.4、 The SS method takes into account the whole spectral content of the signals inthe small aperture, and thus can provide feasible results in multpath area.Moreover, the SS method is sensitive to noise in the signal. Refractivityretrieved by the SS method from COSMIC data contains noticeable positiveand negative bias in the lower troposphere. The SS method can obtain moreoccultations than the BP method in the lower troposphere. 5、 The whole spectral content of RO signal is implemented with the classical SSmethod, without consideration of detecting and removing the noise. In themoist lower troposphere, due to the noise, the accuracy of the bending angleprofile reconstructed by the SS method is degraded. A modified slidingspectral (MSS) method is designed to reduce the noise influence by using thesignal amplitude and spectral power information. Simultaion tests andstatistical comparisons between COSMIC data and ECMWF analysis showthat the MSS method can decrease the influence of the noise to some extent,and therefore MSS method is superior to the SS method in the aspects of bothnoise reduction and vertical resolution.6、 The CT method can retrieve accurate bending angle profiles in the multipatharea. The position of auxiliary plane in the CT method can be fixed for alloccultaion, while in the BP method it must be found individually for eachoccultation. The disadvantage of the CT method is the necessity ofback-propagation, which is the most time-consuming part of the numericalalgorithm.7、 The FSI method has the capability of distentangling multiple rays accuratelyand efficiently in multipath areas. However, the result given with the FSImethod may be affected by the noise in the signal. Statistical comparisons ofthe derived refractivity by the FSI method between the COSMIC data andECMWF analysis indicate that the retrieval bias of the FSI method is smallerthan0.5%above5km height. Systematic negative bias in refractivity isexhibited below about2km due to superrefraction in the lower moisttroposphere.