Research On Snow Storm Minitoring And Early Warning Method Based On GNSS And ECMWF

In recent years,the winter snowstorm has been frequent in Shandong Peninsula,the snowstorm has caused huge losses to the Shandong Peninsula region near the southern Bohai Sea.It is very important to use effective means to monitor snowstorm and make timely disaster warning.Using GNSS data to GNSS water vapor results with high spatial and temporal resolution,low-cost and free from the influence of weather changing.Therefore,this technology has been widely used in the meteorological department.The European Mesoscale Numerical Weather Prediction Model(ECMWF)data provides global stratified grid meteorological data for regional water vapor inversion.Based on the PPP(Precise points Positioning)and ECWMF(European Centre for Medium-Range Weather Forecasts)data water vapor extraction technology,this paper systematically studies the calculation of GNSS PWV and ECWMF PWV and application of snowstorm monitoring based on two ways.Based on Sat Ref data,the convergence time of atmospheric PWV(precipitable water vapor)and the accuracy of GNSS,ECWMF and radiosonde were analyzed.Based on the Bohai Sea area GNSS data of the China Coastal GPS Observing Network and ECWMF data,the characteristics of water vapor and meteorological elements before and after the winter snowstorm in Shandong Peninsula were analyzed in detail.Main contents and conclusions are as follows:1.With the Sat Ref HKSC station data,ECWMF data and radiosound data,the bilateral filtering approach of PPP was introduced,based on ECMWF result and radiosound result,the PWV obtained by bilateral filtering is compared and analyzed.The results show the average PWV convergence time from one-way filtering of PPP is about 30 minutes,and the corresponding average elevation error is 11 cm.The bilateral filtering method of PPP can overcome the problem of the water vapor convergence at the initial stage of one-way filtering,and the bilateral filtering result is obviously better than the one-way filtering result.The average deviation and root mean square(RMS)error from the bilateral filtering relative to ECMWF are reduced by 25% and 10%,respectively.Relative to the radiosonde data,the deviation and RMS error from the bilateral filtering are reduced by 15% and 7%.2.Combined with the heavy snowfall in Weihai on December 7-8,2014,based on ECWMF,Tm and ZWD were calculated,and difference between the Bevis-Tm and ECWMF-Tm is about 5K.The surface PWV has risen and descended,and then rises and falls.PWV in the snowfall area before the snowfall has an increase of 5mm.The water vapor shifts to the northeast direction and moves along the northeast to Weihai.The complex changes of surface temperature,relative humidity and specific humidity are firstly increased,decreased,and increased.The relative humidity and specific humidity have obvious regional changes.3.Based on GNSS PWV,PWV analysis is carried out before and after the snowstorm.Compared with ECMWF PWV,the GPS inversion PWV can reproduce more accurately the change of water vapor.ECWMF PWV of 850 h Pa accounts for nearly 70% of the water vapor.The water vapor is distributed form the height of 850 h Pa to the ground,indicating that the distribution of water vapor is dry at high altitude and wet on surface.With meteorological data of 850 h Pa,when the temperature difference between the upper and lower layers gradually becomes larger,and the temperature at the height of 850 h Pa drops to about-12 ?,and the relative humidity and specific humidity of 850 h Pa reach a certain condition,that is,the thickness of the unstable high-humidity layer increases greatly.It is believed that the snowstorm of the Shandong Peninsula is about to happen.