Quality control method of surface air temperature data

Abstract

The invention discloses a quality control method of surface air temperature data. A random forest algorithm optimized through a genetic algorithm (GA) is used to a meteorological observation station to carry out meteorological element quality control. According to the method, first, surface meteorological observational data in a certain range is collected; afterwards, basic quality control is carried out on the collected data; subsequently, an adjacent station site which is higher in relativity is sought according to variable importance under a random forest model by using the genetic algorithm; training is carried out on the data according to a time sequence and a spatial sequence by using a modified random forest method; a quality control model based on the modified random forest method is obtained; regression prediction is carried out by utilizing a test set; finally, the data are compared with the actual observational data of a target station; the capability that the model identifies wrong data in the observational data is observed. By using the method, valuable meteorological data information accumulated for years by China is effectively utilized; compared with a conventional method, the method is more high-efficient, and is high in accuracy and high in adaptability in a quality control aspect; the quality of the meteorological data of the meteorological observation station is improved.

Description

technical field [0001] The invention relates to the field of quality control of data collected by ground observation stations, in particular to a quality control method for temperature data. Background technique [0002] Since 1951, there have been nearly 1,800 weather stations in China, and the accumulation of meteorological data is rich. However, there may be some errors in observation, input and transmission in these meteorological data. These errors have reduced the quality of weather station observation data. Therefore, the quality control of meteorological data has become an indispensable link in the application of meteorological data. [0003] The quality control of surface meteorological station observation data is generally divided into two categories, one is the single-station quality control method, which mainly includes extreme value check, climate extreme value check, internal consistency check, and time consistency check; the other is multi-station networking ...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to overcome the above technical deficiencies, and propose a method for quality control of base surface air temperature data aimed at the deficiencies of multi-station quality control, which solves the problems of instability and low accuracy of the current multi-station networking quality control method, specifically Achieved by the following schemes:

Images