| The surface automatic weather stations (AWS) data with the characteristics of high spatial resolution and short time interval has been one of the important observation data by the meteorological operation department daily observation, and a important source to provide a more accurate initial field for the numerical weather prediction. It’s hard to ensure its quality because of the deviation of instrument at automatic meteorological station, the influence of surrounding environment and various kinds of man-made factors, so the application in numerical weather prediction is limited. Therefore, improving the quality of AWS data and taking full advantage of them to improve the level of numerical weather forecast are worth to explore and research.Firstly, the AWS data from 151 national AWS and 2600 regional AWS in Jiangsu and Anhui provinces are selected. Based on the hourly AWS data from 2012 to 2014, the missing rates of various elements are estimated respectively. A systematic and sophisticated quality control (QC) scheme is designed for selecting out accurate information and rejecting abnormal information. It turns out that the missing rates of regional AWS data of various elements are apparently higher than the national ones according to the statistics. Almost all the data passes the climate limit check, the suspicious rate of relative humidity data is the highest in regional stations in the climate extreme value check, the error rate of wind field data in regional station is far more than that in national stations in the internal consistency check, the fail rates and error rates of temperature, station surface pressure and relative humidity in national stations are significantly lower than that in regional stations in the second iterated space consistency check, the data not passing the check are almost suspicious data and both results are similar in the time consistency check, the suspicious rates of various elements in regional stations are apparently higher than that in national stations and the suspicious rate of relative humidity data is the highest in national stations in the continuous check, and the error rates of wind field in regional station are far higher than that in national stations and the quality of temperature and pressure data is relative to the best among various elements in the comprehensive decision-making algorithm. What’s more, the suspicious stations of regional and national stations are respectively marked according to the result of quality control.Secondly, using WRF mesoscale numerical model and the three-dimensional variational data assimilation system, a heavy rainfall weather process during 4th-6th July 2013 in the Yangtze and Huaihe river basins is simulated through surface AWS data assimilation. It turns out that It turns out that surface AWS data assimilation mainly adjust the middle and low level of the initial field in the model and obviously improve structure of the rain belt and the intensity of the rainfall in the 12-hour accumulated precipitation simulation. The altitude difference should not be ignored because of the simulation effect of data which after terrain correction assimilation is better than that which without correction. Furthermore, the simulation effect of data assimilation is relative to the best when all of the observation data being corrected combining with the model background information using Guo-scheme.In consequence, it is beneficial to the improvement of the real-time data quality and the maintenance and correction about instruments in time of stations if the results of QC scheme, especially the information of error data and suspicious stations, can be provided to the corresponding station as timely feedback. Furthermore, the assimilation of surface AWS data is helpful to improve the model prediction and realize the AWS data be made full use of. |
PDF Full Text Request