| Continuous,real-time,and highly accurate acquisition of the spatial and temporal distribution of water vapor,as well as the changing characteristics could provide guarantee for accurately controlling the evolution of the weather system.The coastal areas in China cover wide latitudes from the north and the south,showing combined vapor characteristics of sea and land,and the climate is complex and variable.Meanwhile,Marine disasters also occur frequently in thses areas.Therefore,strengthening study on monitoring methods and application of water vapor information in coastal areas will help to understand climate change in China,and it is also essential to prevent and reduce marine hazards.Based on the data of the Chinese coastal GNSS observation network and other observational equipment,this paper analyzed the accuracy of PWV retrieved from GPS/GLONASS data in the coastal areas,and evaluated the applicability of reanalysis surface meteorological data form multiple sets of Numerical Weather Prediction(NWP)reanalysis data and GPT2 model in ground-based GNSS water vapor retrieval,respectively.On this basis,the spatial-temporal variation of PWV and water vapor transport in the coastal areas were studied,then a method for short-term prediction of typhoon precipitation using the change rate of water vapor and surface temperature.The main research aspects and conclusions of this paper included:(1)Combining the GNSS data and ground meteorological data of the Chinese coastal GNSS observation network in January 2014,the precision of the PWV retrieved from GPS/GLONASS data uder different elevation mask were discussed,using the precise point positioning(PPP)and double difference positioning,respectively.The study indicated that,at the elevation mask of 10°,the zenith tropospheric delay(ZTD)estimated from combined GPS/GLONASS double difference was better than that from GPS double difference and combined GPS/GLONASS PPP.There is no obvious systematic bias in the extraction results of each method.Satellite cut-off angle setting has a great influence on the zenith tropospheric delay accuracy.The ZTD extracted by the double-difference network solution GPS single system has the best accuracy under 300 cut-off elevation angle,and its precision is significantly reduced compared to the results of the 10° cut-off angle.(2)As to the problem of the large number of existing and historical GNSS observation stations without collocated ground meteorological equipments,the method that interpolates surface pressure and temperature from ERA-Interim,NCEP FNL,and JRA-55 reanalysis data and GPT2 model to any GNSS station was assessed using 1-year term and 25 stations evenly distributed in the coastal areas of china.Compared with the measured meteorological observation data in 2014,the accuracy of ground meteorological data interpolated from ERA-Interim reanalysis data is slightly better than that of other three data interpolated results.The average Bias and RMSE between the GNSS PWV obtained by interpolating surface meteorological data from ERA-Inteirm reanalysis data and the measured data are 0.03 mm and 0.31 mm,respectively,and between 0 and 1.2 mm from the long-term results of radiosonde data.The corresponding RMSE is between 1.8-3.2 mm and satisfies the precision requirements of the water vapor application.Under the conditions of mesoscale extreme weather such as typhoon,the interpolation of surface meteorological data can not meet the accuracy requirement for inversion of ground-based GNSS water vapor information.(3)Using the GNSS PWV from 30 observation stations of China's coastal GNSS observation network for the period from 2010 to 2016,the spatial-temporal variation and periodic characteristics of water vapor in coastal areas were systematically analyzed.The distribution of water vapor in coastal areas is mainly affected by latitudes and is also affected by land-sea distribution and water vapor channels.Water vapor has obvious seasonal,monthly,and daily variations.Influenced by summer season,the maximum monthly PWV is different in different regions.The daily variation of water vapor changed with latitude location,season,topography and local climate.The main reason for the daily variation is surface evaporation and local thermal circulation.The water vapor daytime series has a remarkable annual period,half year period,daily period and half daily period.The annual period amplitude is 16.54-22.09mm and decreases from mid-latitudes to high latitudes and equatorial regions.The half-year period amplitude is 0.42?6.32mm and there is no obvious correlation with the latitude.The daily period amplitude is 0.12?0.61mm and the half daily period amplitude of each station is 0.18?0.62mm.(4)Using GNSS PWV and ground meteorological observation data,the relationship between water vapor change and surface meteorological elements was analyzed.The study shows that seasonal variation of water vapor has a good correlation with atmospheric temperature,and their correlation with surface air temperature depends on the ground temperature level.The daily variation of hourly cumulative precipitation frequency is more significant than that of hourly cumulative precipitation,which has obvious relationship with PWV daily variation in summer.The daily variation of PWV is related to the change of temperature and specific humidity day and the evolution of surface wind,which has no obvious correlation with the diurnal variation of surface pressure and sea surface temperature.(5)The relationship between the variation of atmospheric water vapor and ground meteorological elements and the precipitation of typhoon is analyzed by using the measured data of GNSS and the ground meteorological hourly data of the coastal GNSS observation network during 2010-2014 years when typhoon transit,and a short-term method of forecasting the typhoon precipitation is proposed with the using water vapor and ground temperature changes.Studies have shown that the GNSS PWV and ground temperature have short-term trends before the occurrence of typhoon and they play a strong role in short-term prediction of rainstorm.Using the change rate of the zenith precipitable water vapor content during the rising period and the change rate of ground temperature during the decline period,the typhoon precipitation can be short and impending forecast.This method can predict more than 85%of the strong precipitation events,the error forecast rate is low,and has certain guiding significance for the typhoon precipitation forecast. |
PDF Full Text Request