Prediction of tropospheric wet delay by an artificial neural network model based on meteorological and GNSS data

Mahmut Oğuz Selbesoğlu

Engineering Science and Technology an International Journal · 12 December 2019

In one sentence. A neural network predicts GNSS zenith wet delay up to six hours ahead from in-situ meteorological data, reaching 1.5 cm RMSE at two Austrian stations.

Abstract

Estimation of tropospheric wet delay is of great importance for real-time weather forecasting applications. In the last decade, based on troposphere wet delays obtained from Global Navigation Satellite System observations, high temporal and spatial resolution water vapor data can be produced for reliable and accurate weather forecasting. The main objective of this study is to investigate the accuracy of tropospheric wet delay prediction based on artificial neural network technology by the integration of Global Navigation Satellite System and meteorological data from in-situ observations of The New Austrian Meteorological Measuring Network. In the study, artificial neural network model was used to predict the wet troposphere delay up to six hour. Predicted zenith wet delay values were compared with the values estimated from Global Navigation Satellite System observations for validation. The predictions were carried out during humid (August) and dry (December) periods on two reference stations belonging to Echtzeit Positionierung Austria GNSS Network of Austria. The root mean square error of zenith wet delay prediction based on newly designed artificial neural network Model was found 1.5 cm for up to six hours.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.