简介：Inthispaper,wedescribetheestimationoflow-altituderefractivitystructurefromsimulationandrealground-basedGPSdelays.Theverticalstructureoftherefractiveenvironmentismodeledusingthreeparameters,i.e.,ductheight,ductthickness,andductslope.Therefractivitymodelisimplementedwithaprioriconstraintsontheductheight,thickness,andstrength,whichmightbederivedfromsoundingsornumericalweather-predictionmodels.Araypropagationmodelmapstherefractivitystructureintoareplicafield.Replicafieldsarecomparedwiththesimulationobserveddatausingasquarederrorobjectivefunction.Aglobalsearchforthethreeenvironmentalparametersisperformedusingageneticalgorithm.Theinversionisassessedbycomparingtherefractivityprofilesfromtheradiosondestothoseestimated.Thistechniquecouldprovidenear-real-timeestimationoftheductingeffect.Theresultssuggestthatground-basedGPSprovidessignificantatmosphericrefractivityinformation,despitecertainfundamentallimitationsofground-basedmeasurements.Radiosondesaretypicallylaunchedjustafewtimesdaily.Consequently,estimatesoftemporallyandspatiallyvaryingrefractivitythatassimilateGPSdelayscouldsubstantiallyimproveover-estimatescausedbyusingradiosondedataalone.