CAJ | 학술논문

通过对北京地区天气现象组网试验资料的可用性分析以及在精细化预报业务中的应用，发现自动观测天气现象资料连续时长多于20 min以上时，数据可信度较高，对预报员有较高的参考价值。自动观测系统对雨、雾、霾现象的识别结果正确率总体较好，其高频次的探测数据对低频次的人工观测起到补充作用。但在天气现象识别算法、质控、雷暴和降水性质识别、现象识别连续性等方面还存在不足，需要进一步改进。
통과대북경지구천기현상조망시험자료적가용성분석이급재정세화예보업무중적응용，발현자동관측천기현상자료련속시장다우20 min이상시，수거가신도교고，대예보원유교고적삼고개치。자동관측계통대우、무、매현상적식별결과정학솔총체교호，기고빈차적탐측수거대저빈차적인공관측기도보충작용。단재천기현상식별산법、질공、뇌폭화강수성질식별、현상식별련속성등방면환존재불족，수요진일보개진。
Based on the analysis on the data from the automatic weather phenomena observing network and the application on refined forecasting, it is found that the automatic observation data, considering the temperal span binned as 20 minutes, could have a high reference value for the forecasters. The accuracy rate of the identified weather phenomena which is complementary to manual observation including rain, fog and haze, is totally high. Also the results show that the defects on the identified algorithm, the identification of thunder and precipitation phase, the continuity still exist, and should be further improved.