Inversion Method Of Satellite Ultraviolet Remote Sensing To Retrieval Ozone Concentration In Troposphere

Tropospheric and ground-level ozone pollution aroused more and more concerns along with growing speed of global urbanization and industrial development.Thus quick and accurate measurement of atmospheric ozone concentration become vital important nowadays,which serves bases for real-time atmospheric ozone monitoring,air quality forecasting and contributes to further studies of photochemical pollution or even environmental and climate effects caused by ozone pollution.Currently,ground-level ozone monitoring based primarily on air quality monitoring stations,which characterized on scattering measuring points clustered around big cities.This measuring method has its advantages:like providing precise ozone concentration and nice vertical profiles of ozone concentration around the stations.However,this traditional method suffers from extensive cost of building and maintaining these air quality monitoring stations.Moreover,the uneven and low density of station distribution results in low credibility of ozone concentration products in places far away from the monitoring stations.Satellite remote sensing's synoptic view,global coverage as well as low cost all weather monitoring characteristics make it high advantages over the traditional station ozone observation,and the remote sensing ozone concentration measuring technique is widely accepted and applied worldwide.Most of the remote sensing ozone monitoring applications is focused on broad scale like global monitoring while rarely applied in local regions like in urban region ozone concentration monitoring.To promote satellite remote sensing in application of lower tropospheric ozone monitoring,remote sensor design parameters are simulated based on Bayesian theory to provide theory and technical support for future hypespectral remote sensors application in accurate measurement of tropospheric ozone concentration.Lower tropospheric ozone concentration inversion algorithms are designed according to a geostationary orbit remote sensor and optimized using extended Fourier amplitude sensitivity test to study optimal channel combination in measuring concentration.Through the comparison of ozone concentration products temporal variation trend from ground station monitoring,sounding stations and ozone concentration products from remote sensing,spatial-of ground-level ozone concentration variation and products correlation difference were obtained to further analyze the affecting factors in inversing ground-level ozone concentration through remote sensing imagery.The relationship between weighting function and measurement channel sensitivity were derived.With indications from the eFast global sensitivity index,ultraviolet response sensitivity to ozone concentration along all atmosphere stratifications was obtained and then the optimal sampling number and search curve are determined.In ideal circumstances,the regulatory factors to the average kernel function were analyzed.And combining with multispectral sensitive features,a two-step separation algorithm for surface ozone concentration inversion was proposed.Compared with the traditional algorithm,the new algorithm showed significant improvement on the inversion accuracy of lower tropospheric ozone concentration.Given the geostationary orbit UV sensor's spatial coverage and view geometry,the sensor's reasonable resolution spatial,spectral and radiation resolution as well as the signal-to-noise ratio can be designed and calculated.After analyzing the precision of TSS inversion algorithm along variations of sun and view geometry,the required range of spatial-,temporal-resolution and seasonality can be derived.1)eFast global sensitivity analysis found each measurement channel's sensitivity index in between the threshold range.Weighting function matrix,priori contour of covariance matrix and measurement error covariance matrix jointly influence the detestability of remote sensing in measuring atmospheric ozone,which showed the signal freedom.After determining the best sampling numbers and searching curves,each measuring channel's sensitivity can be quantified.A significant sensitive characteristic was found in spectrum less than 300nm when the height was above stratosphere while remarkable sensitive characteristics were found in the spectrum more than 300 nm when the in atmosphere layers of stratification.All these characteristics provide essential support to the atmospheric ozone concentration retrieving accuracy in between and above stratification.Some sensitive measuring channels were also found in the troposphere,especially in height at ground-level though the relatively smooth sensitivity curves in these atmosphere layers.2)Taking advantage of the important role of weighting functions in the ozone concentration inversion,Optimization method(OE)introduces prior knowledge of ozone profile information and measurement error information,improving the reliability and applicability of the inversion results.The OE method comes with a well-equipped inversion technology system and is able to find out different error source's influence on inversion results,making it useful in improving algorithm precision and optimization of input parameters.A two-step separation(TSS)algorithm was proposed after studying the tuning mechanism of weighting functions,prior profile covariance,and measurement error covariance to inversion results and combining sensitivity difference of measuring channels in UV band.The TSS algorithm took advantage of the preferred spectra measuring channels when atmosphere on and above the stratosphere,hence avoiding the uncertainty associated caused by insensitive channels.The prior knowledge parameter information is then updated,which in turn improved the inversion precision of.the troposphere and ground-level ozone concentration inversion.Although traditional algorithm(TR)are quicker in completing the reiteration process and has a slightly more precise result in ozone concentration above stratosphere,the TSS algorithms give more precise results in ozone concentration inversion in troposphere and ground-level.This means that TSS algorithm would be better applied within troposphere.TR improved the group-level ozone product result to some point since the stratosphere and above atmosphere also take part in the inversion,however,this improvement would not outweigh the loss of precision that caused by TR in troposphere and ground-level atmosphere.3)In the upper troposphere,the inversion accuracy of ozone profile can be significantly improved by using TSS algorithms.Smoothing error is the main error source.Current remote sensors have inadequate capability in accurate detecting ozone concentration in the troposphere,especially in the ground-level atmosphere.Current remote sensing data are still primarily used to derive coarse-resolution global scale ozone concentration;the remote sensing products of ozone concentration vertical profile in troposphere as well as ozone concentration in ground-level have large deviations from the measured values.Data from group monitoring stations has high correlation with data from air sound stations,both of which suggest that ground surface type has an important influence upon the horizontal spatial distribution of surface ozone concentration.The surface ozone concentration ranges extracted from the OMI profile are close to measured value from ground stations,have low correlation with time variation,and low correlation with atmosphere NO2 and sun geometry.The inversion precision of satellite ozone concentration and the average kernel function suggest that to get better ozone concentration inversion results,a reasonable combination of channels together with different atmosphere stratification is recommend.4)The study has proved reasonable observation spatial coverage and time range to satellite remote sensors when inverse ozone concentration using TSS algorithm.According to retrieval ozone concentration in lower troposphere,geostationary orbit satellite remote sensor's spatial resolution,spectral resolution and radiometric resolution were concluded.As the variation of solar incident angle changes the radiance on the sensor observations direction,and hence the apparent modification to sensor SNR,which varies with different sensor scanning angle and measuring channels.These influences SNR from the sun incidence angle can't be neglected and should be taken seriously when designing the UV sensors in future.Calculation result showed that with the increase of sun zenith angle,detecting channel freedom both decreases in troposphere and above stratospheres with the former apparently decrease while the latter slightly decrease happening.Also with the solar zenith angle increases,the freedom attenuation increases gradually.