Large-scale rainfall monitoring abnormal site screening method

Abstract

The invention relates to a large-scale rainfall monitoring abnormal site screening method. The method comprises the following steps of 1, preliminarily judging a base station based on annual sequence rainfall observation data by adopting a Hampel method; 2, judging the base station again by adopting the Grubbs criterion; 3, after the base station is determined, a surrounding observation station analysis method is adopted to judge abnormal stations based on hourly rainfall monitoring data; and step 4, performing radar-assisted verification on the abnormal site. According to the method, anomaly identification and rapid processing of large-scale rainfall monitoring station data can be realized, the anomaly identification rate is up to 95% or above, and an accurate and reliable basis is provided for rainstorm flood risk prompting and early warning.

Description

technical field [0001] The invention relates to a large-scale rainfall monitoring abnormal station screening method, which belongs to the field of meteorology and hydrology, and is mainly used for providing accurate and reliable rainfall monitoring information for rainstorm and flood risk warning and early warning. Background technique [0002] Rainfall monitoring is an important part of hydrological monitoring, and it is the eyes, ears and staff of storm flood disaster prevention. Since the 21st century, water conservancy departments have increased their support for the construction of automatic monitoring stations, especially through the construction of mountain torrent disaster prevention and control projects, the number of automatic monitoring stations for mountain torrent disasters in the country has reached 132,000, and the average density of automatic rainfall station networks is 38km 2 / station, which is 22 times that of 2006 (6000 stations), the minimum flood report...

AI Technical Summary

Problems solved by technology

However, due to the low construction standards of some measuring stations, the operation and maintenance of measuring stations located in hilly areas is difficult, and the data quality is difficult to be guaranteed. There are often situations such as taking large numbers, missing measurements, and missing measurements, and the problems of measuring stations have a strong influence Randomness, it is impractical to completely abandon a station

[0003] In order to effectively use the monitoring data of the stations, it is necessary to find out the stations with accurate monitoring data in different periods from many rainfall monitoring stations and eliminate the stations with problematic data quality. However, most of the current identification methods are highly subjective and misjudge The possibility is very high, and there are great risks in the application of business. First, it is difficult to identify abnormal sites. Second, it is easy to mistake the correct site for abnormal site, which is contrary to the original intention of carrying out abnormal site identification.

Images