Data Quality Evaluation of Chinese Wind Profile Radar Network in 2018

In 2018, a total of 98 wind profile radars are stable and upload data to CMA. Based on the ” Data quality control and evaluation system of wind profile radar network (V1.0)” independently designed and developed by Meteorological Observation Center of CMA, the quality of data is evaluated and analyzed with the following contents: detection capabilities and data accuracy. The main conclusions are as follows: (1) The effective detection height of over 80% of the wind profile radars in the whole network has met the design indicators. The main reasons causing non-compliance of the effective detection height are system failure and low system performance. (2) The data of wind profile radars is compared with the data of GRAPES forecasting field, more than 90% of the data participating in the evaluation is relatively accurate. After data quality control, the standard deviations of U and V components are both within 3.05 m/s, compared with the non-quality control, they are decreased by 19.3 % and 19.5 %,furthermore, they are decreased by 4.1 % and 2.9 % compared with 2017. The reasons leading to the poor accuracy of data mainly are system failure, system observation parameter setting errors and data non-compliance. Based on the above two assessments, 19 sites have poor data availability throughout the year, accounting for 19.4% of all the wind profile radars, and 18 sites have poor data availability during part time of the year. The factors that affect data quality have the following characteristics: the number of sites with low system performance has increased year by year; the system observation parameter setting errors and the system failure that can not be repaired for a long time remain; due to signal interference shutdown, the lack of observation data is serious. In the future, China will improve data quality and enhance the efficiency of data application by strengthening management of business operation at assessment sites and the equipment operation at non-assessment sites, upgrading equipments with low system performance, and improving data quality control algorithms and so on.