The Characteristics Of Hourly Precipitation During Warm Seasons And Its Forecast Evaluation In Mountain Tai Area

Mount Tai is a typical isolated topography with small spatial scale in North China Plain.The temporal and spatial scale of its influence on precipitation are quite small,and the precipitation in this area has the characteristics of suddenness and locality,so it is difficult to forecast and has important research value.In this paper,the characteristics of hourly precipitation(including diurnal variation,spatial scale,duration of rainfall)in Mount Tai and its surrounding area were investigated by using hourly precipitation observation data.Based on hourly precipitation characteristics,the precipitation forecast products of RMAPS-ST experimental system at 3 km horizontal resolution are evaluated thoroughly,and the differences of rainfall forecast products starting at 20 LST and 08 LST were compared.The main conclusions are as follows:The amount of precipitation in Mount Tai and its surrounding area is concentrated in the early morning and afternoon,and the rainfall frequency shows the characteristics of the main peak in the early morning and the secondary peak in the afternoon,while the intensity of precipitation is equivalent to the peak value in the early morning and afternoon.The enhancement on rainfall of Mount Tai is remarkable and the diurnal variation of precipitation enhancement also shows bimodal peaks in the morning and afternoon.Short-duration and small-scale precipitation tends to peak in the afternoon,while long-duration and large-scale precipitation mostly reach its peak in the early morning.Compared with other stations in surrounding areas,the rainfall in Taishan Station lasts longer and covers smaller area,especially in the early morning.In addition,Taishan Station is prone to a kind of local weak precipitation in the early morning.RMAPS-ST basically reproduced the spatial distribution characteristics of hourly precipitation in Mount Tai and its surrounding area,but the predicted rainfall amount of stations on the southwest side(Taishan Station and stations on its northeast side)of the mountain area was lower(higher)than observations.There was a large deviation of the predicted precipitation of early morning and afternoon.Taking Taishan Station as an example,RMAPS-ST underestimated the rainfall frequency of early morning,which was related to the fewer predicted rainfall events and shorter predicted rainfall duration.The predicted intensity of early morning rainfall at Taishan Station was different in the forecast results with different starting times.In the forecast results starting at 20 LST,the intensity of early morning rainfall at Taishan Station was significantly overestimated,which led to the fact that its predicted precipitation was higher than observations,while the overestimation of rainfall intensity was not obvious in the forecast results starting at 08 LST.RMAPS-ST overestimated the frequency of afternoon rainfall at Taishan Station in the prediction results starting at 08 LST,which was related to more predicted rainfall events with short duration in the afternoon.The main sources of the deviation of precipitation forecast were the false alarm and omission of rainfall.RMAPS-ST was most likely to miss precipitation in Taishan Station,and the missed rainfall was mainly weak rainfall in the early morning.There were two main types of rainfall processes that caused omission: one was the nocturnal systematic precipitation.The prediction deviation of its development and evolution led to the omission in the statistical results.The other was the local,weak precipitation in the early morning in Taishan Station that the model completely missed.RMAPS-ST was prone to have false alarm of rainfall in the afternoon,but the area where false alarm was mostly likely to occur differed in forecast products with different starting time.More than half of the precipitation process that caused the false alarm was the predicted rainfall happened in the mountainous area in the afternoon,while there was no precipitation in the observation results.This was related to the prediction deviation of thermal field and water vapor field.The unstable state of atmosphere in the model and the strong southwest air flow could easily lead to the precipitation.