Automatic rainfall station abnormal value screening method based on rainfall measuring radar data

Abstract

The invention discloses an automatic rainfall station abnormal value screening method based on rainfall measuring radar data. Abnormal value judgment and quality screening are carried out on a ground automatic rainfall station based on national Doppler radar body scanning reflectivity and jigsaw products. Obvious fault points are monitored and screened based on long-sequence time sequence station data, reflectivity three-dimensional change curved surfaces of different height layers are obtained based on a multi-layer Doppler radar reflectivity product, a rainfall precipitation distribution stereogram generated by an automatic precipitation station is combined, and a singular change value in the time sequence monitoring data of a ground rain gauge is obtained by using a curved surface similarity matching algorithm, and suspected abnormal values of an automatic rainfall station are extracted. An extreme value judgment set, a peripheral station judgment set and a radar reflectivity judgment set of the abnormal value stations are established, the similarity between the normalized time series data of the rainfall stations and the time series data of the peripheral stations and the radar reflectivity grid is obtained based on a time series data similarity mining algorithm, and the suspected station abnormal values are automatically studied and judged according to a similarity standard. and therefore, a screening result of the massive abnormal values of the automatic rainfall station is obtained, and a preliminary judgment result of station fault types is given according to different characteristics of time sequences of the abnormal station.

Description

technical field [0001] The present invention relates to the field of rainfall radar data application and ground automatic rainfall station quality control, specifically a method for abnormal screening of automatic rainfall station based on rainfall radar data, a three-dimensional free-form surface based on multi-layer radar reflectivity data and ground station measured values The similarity, radar reflectivity grid and single-point rainfall station time series similarity are used to judge station anomalies, which are mainly used for abnormal value screening of massive automatic rainfall stations and large-scale and high-precision small watershed surface rainfall acquisition, so as to forecast storm and flood disasters Alerts provide precise input. Background technique [0002] Flood disaster caused by heavy rain is one of the main natural disasters in my country, and it is also a long-term difficult problem in disaster prevention and mitigation in my country. The central co...

AI Technical Summary

Problems solved by technology

[0004] Automatic observation stations are generally built in towns and towns, with dense distribution, and can detect small and medium-scale weather systems. However, it is extremely difficult to analyze the abnormal rainfall data of automatic stations. Precipitation is sudden and uneven in time and space. Find out whether the rainfall data is abnormal during the process

At present, our country still mainly adopts the time-space difference and extreme value inspection method of the rainfall data in the analysis process. On the one hand, it is difficult to realize the automatic identification of large-scale and massive automatic stations. On the other hand, the automatic stations are compared with their own monitoring data and have not Surrounding rainfall monitoring is correlated, and precipitation data are correlated in both time and space dimensions, and cannot be analyzed in isolation, so misjudgments and missed judgments often occur

Therefore, for the national rainstorm and flood risk monitoring and early warning platform, the existing conventional automatic station rainfall anomaly analysis methods cannot provide high-time and accurate site anomaly screening results, and thus cannot achieve high-precision rainstorms across the country. Flood risk forecast and early warning

Images