Research On The Method Of Gas Parameters Retrieval Based On Rayleigh-brillouin Scattering Spectrum

Lidar technology plays an important role in the fields of atmospheric research because of its advantage in high spatial resolution for profile detection.Compared with traditional lidar technology that recieved the energy of backscattering signal,the retrieval method of Rayleigh-Brillouin lidar based on spectral detection is more simple and efficient.However,spectral processing method of Rayleigh-Brillouin lidar will be influenced by the electrical noise originated from photoelectric conversion.On the other hand,the retrieval method of Rayleigh-Brillouin lidar will be influenced by optical noise originated from extra spectrum in deconvolution operation.The retrieval accuracy will be decreased result from both above.The traditional method utilized relationship among spectral linewidth,temperature and pressure can not retrieve multi-parameters simultaneously.The retrieval accuracy of Tenti-S6 model will decrease result from the number of retrieval parameters increasing.In order to resolve above questions,the main work of this article are as follows:(1)For the problem that traditional method can not perfectly process dark count to obtain the accuracy spectrum due to the influence of temperature drift,a novel method was proposed in this article.For the new method,the baseline of darkcount has been taken as a parameter of Tenti-S6 model and take part in the fitting procedure for calculating the baseline of darkcount accurately by the means of the area normalization relationship between the darkcount in original spectrum and Tenti-S6 model.Finally,the accurate baseline of darkcount should be subtract from original Rayleigh-Brillouin spectrum in order to correct the spectral baseline.The residual between the spectrum processed by new method and Tenti-S6 model is 1.91%,the temperature retrieval error is1.9K,and the pressure retrieval error is 0.044 bar.Therefore,the retrieval error from new method is smaller than that of traditional method.(2)In order to resolve the problem that low frequency noise can not be removed by Butterworth filter and the noise in both ends of the spectrum can not be removed by Three-Point Zero-Order Savitzky-Golay Filter(ITZSGF),the rotating S-G filter has been proposed in this article.The rotating S-G filter connects the ends of the spectrum and smooths the noise in circulated way on the cycling spectrum until the effect of denoising approach perfect condition.Rotating S-G filter can effectively resolve the problem of low frequency noise and the noise in the end of the spectrum.The temperature retrieval error derived from rotating S-G filter is 1.49 K,pressure retrieval error is 0.01 bar,which are all smaller than those of Butterworth filter and ITZSGF.(3)In order to resolve the problem that the traditional method can not retrieve the gas temperature and pressure simultaneously,a new method that can retrieve gas temperature and pressure by using many characteristics of Rayleigh-Brillouin spectrum has been proposed in this article.New method obtained the relationship among gas temperature and many characteristics,and the relationship among gas pressure and many characteristics for establishing the characteristic retrieval model,then the best characteristic retrieval model has been established.The temperature retrieval error derived from the new method is 0.603 K,and the pressure retrieval error is 0.036 bar,which are all smaller than those of Tenti-S6 method.(4)In order to resolve the problem that the promoting of retrieval accuracy has been limited by traditional modeling method,the inversion modeling method that establish the retrieval model only use two groups of whole linewidth that arranged oppositely each other has been proposed in this article.Theoretical research show that,the fitting error and theoretical error of the whole linewidth model based on inversion modeling method are smaller than those from dual-parameter retrieval model.Besides,the new method due to the advantage of accurate retrieval can satisfy the requirement of remote sensing with short wavelength in Rayleigh-Brillouin lidar.The temperature retrieval error derived from the new method propose in this article is 0.211 K,and the pressure retrieval error is0.007 bar.In a word,the new method can be used for promoting the retrieval accuracy of traditional method.For the problem that current spectral processing method influenced by electrical noise and optical noise,the spectral baseline correction formula and rotating S-G filter have been proposed.Besides,for the problem that the gas parameter can not be retrieved by traditional retrieval model based on optical spectral detection,the improved and high-pricise retrieval model has been proposed.These theoretical researches can provide theoretical guidance and technical support to researchers in this field.