| Radio occultation(RO)technique,as an important part of GNSS(Global Navigation Satellite Systems)meteorology,can provide high vertical resolution,global coverage,calibration-free,all-weather,and high precision atmospheric information for weather and climate studies.From the Global Position System/Meteorology(GPS/MET)mission launched in 1995 to the recent Constellation Observing System for Meteorology,Ionosphere,and Climate(COSMIC-2)program,a continuous RO dataset of more than 20 years has been collected,and a new opportunity for long-term climate analyses using multi-mission RO observations has subsequently arisen.Therefore,the primary purpose of this research is to investigate how best to use those multi-mission RO observations for climatological purposes as well as to evaluate the quality and consistency of long-term RO-based climatologies.The main contributions of this research are summarized as follows:1)Using ERA5 reanalysis and radiosonde data as a reference,the precision of atmospheric temperature and refractivity profiles from four RO missions(including COSMIC-1/2,Challenging Minisatellite Payload,Meteorological Opera-A/B,and Feng Yun-3C/D)between 2001 and 2020 were evaluated.The bias between RO and the reference data was calculated and the spatiotemporal variations of the bias were analyzed.In the lower troposphere(0 km～8 km),the mean bias of atmospheric temperature and refractivity is between –0.2 K～1.2 K and –1.5%～0.2%,respectively.The magnitude of the bias is most significant near the surface and decreases gradually with the increase of altitude;the bias in the low latitude region is larger than that in the middle and high latitude regions;the bias in summer months is larger than that in winter months.In the upper troposphere and lower stratosphere(8 km～22 km,UTLS),the mean bias of temperature and refractivity ranges from –0.3 K to 0.3 K and –0.3%to –0.2%,respectively,and the variations of the bias with altitude,latitude,and season are stable.In the upper stratospheric region(22 km～40 km),the mean bias falls into the range of –6 K～–1 K and –0.2% ～0.6%(temperature and refractivity).The amplitude of mean bias increases with the increase of altitude;the mean bias is more pronounced in high latitude than middle and low latitude regions,and more significant in winter months than summer months.According to the characteristics of the bias between RO and reference data,an empirical error model was developed using a piecewise function,and the parameters of the empirical model were calculated.2)The performance of different setup methods for RO-based climatologies including the "bin" method,the maximum search radius method,and the spherical harmonics method was assessed.The influences of spatial and temporal resolution,maximum search distance,and maximum spherical harmonic degree on the reliability of RO climatologies were investigated.The reliability of the “bin” method is mainly affected by the amount and latitudinal distribution characteristics of RO observations.Under the influences of the above two factors,the proportion of invalid bins is largest near the polar and equatorial regions and smallest in the middle-latitude region.The performance of the maximum search radius method is closely related to the density of the RO profile,and the proportion of invalid grid points is the largest near the equator and gradually decreases toward the North/South poles.The key problem of the spherical harmonics method is to find the most appropriate maximum degree.Based on the RO profiles in June 2008,the recommended value of the maximum degree is range from 9 to 12.Taking RO temperature climatologies of 2008 as an example,the precision of the three methods was analyzed.Results show that the global mean bias of the “bin” method and the spherical harmonics method ranges from 0.0 K to 0.3 K in the lower troposphere,0.0 K to 0.5 K in the UTLS region,and 0.6 K to 1.0 K in the upper troposphere;the mean bias in winter is slightly larger than that in other seasons.The global mean bias of the maximum search radius method is about 0.1 K to 0.5 K,–0.4 K to 0.6 K,and 0.1 K to 1.2 K respectively in the above three regions.Overall,the precision of the “bin” method and spherical harmonics method show good agreements with each other,while the precision of the maximum search radius method is relatively low.3)Taking atmospheric temperature as an example,the accuracy of multi-mission long-term RO climatologies was evaluated.Based on the mechanism of the climatological error,the total error of RO climatologies is divided into three components,including sampling error,statistical error,and systematic error.The characteristics and spatiotemporal variation of those components were analyzed.The sampling error of RO climatologies originates from the fact that RO atmospheric measurements cannot give a uniform and sufficient sampling of the atmosphere.The values of the sampling error range from-1.5 K to 1.5 K and the most significant sampling error was observed in the lower troposphere.The impact of sampling error on RO-based climatologies can be evaluated by sampling error estimation.By subtracting the estimated error from the RO climatologies,most of the sampling error can then be removed,leaving a residual sampling error,which is about 30% of the original values(i.e.–0.5K ～ 0.5K).The statistical errors result from the accidental error of RO atmospheric profiles and will thus be gradually diminished by averaging over many profiles.The values of the statistical error range from 0.0 K to 0.2 K and the maximum values were observed in the upper stratosphere region.The systematic error accounts for potential residual biases in RO observations as well as in the retrieval process and generally dominates the total climatological error.The values of systematic error fall into the ranges from-1.0 K to 0.5 K,with the most significant error found in the low troposphere and upper stratosphere region.Based on sampling error,statistical error and systematic error,the total climatological error for RO climatologies can then be calculated with a reasonable assumption that the three components are uncorrelated with each other.In the lower troposphere,the total climatological error is about 0.0 K to 1.5 K,and the largest error occurs in the southern hemisphere high latitude region.In the UTLS region,the total climatological error remains from 0.0 K to 0.5 K.In the upper stratospheric region,the magnitude of climatological error increases gradually with the increasing altitude,and the maximum value can reach about 2 K.4)Taking ERA5 reanalysis and radiosonde data as references,the homogeneity of RO climatologies from 2001 to 2020 was evaluated,and the temporal evolution characteristics of RO atmospheric temperature series were analyzed.The RO temperature series can be divided into four parts: long-term trend term,seasonal term,temporal mean shifts,and residual term.Based on the characteristics of the RO temperature series,a strategy of multiple changepoints detection for RO climatologies were introduced by the combination of STL(Seasonal-Trend Decomposition Procedure Based on Loess)and PMT(Penalized Maximal t-Test)methods.As a result,a total number of 751 mean shifts were detected,with 187,106,and 458 of them being in the lower troposphere,the UTLS region,and the upper stratosphere region,respectively.Then,with the help of the total climatological error of RO climatologies,a changepoints homogenized method was proposed.After the changepoints homogenized process,the accuracy of RO climatologies is significantly improved,especially in the upper stratosphere region.The bias between RO climatologies and ERA5 data is reduced from –3 K～3 K to –1.5 K～1.5 K.The homogenized RO climatologies were then used to study the long-term trend change of atmospheric temperature.Results show that from 2001 to 2020,atmosphere warming was mainly observed in the lower troposphere region,while in the upper stratosphere region atmospheric temperature decreases gradually.In this thesis,some key problems of RO technology were investigated and discussed in-depth,including the RO retrieval method,accuracy evaluation of RO atmospheric profiles,error analysis,and homogeneity assessment of RO climatologies.It is believed that this research can provide some valuable theoretical and application references for the RO technique in weather and climate study.There are 58 figures,5 tables,and 150 references in this thesis. |
PDF Full Text Request