Data Fixing Algorithm in Radiosonde Monitoring Process

Abstract

Earth surface monitoring can give information that may be used in complex analysis of the air conditions, temperature, humidity etc. Data from a vertical profile of the atmosphere is also essential for accurate thunderstorm forecasting. That data is collected by radiosondes – telemetry instruments carried into the atmosphere usually by balloons. Sometimes, due to the hostile conditions of upper troposphere, incorrect data can be generated by radiosonde sensors. In this paper, a new algorithm is developed for fixing the incorrect data, i.e. missing or out of specific range values. The proposed algorithm was tested both on benchmarks and real data generated by radiosondes. About 70% of significantly damaged test data volume was recovered. Up to 100% of real data was fixed.