Open AccessModelling weighted mean temperature in the West African region: implications for GNSS meteorology
Author(s) -
Isioye Olalekan Adekunle,
Combrinck Ludwig,
Botai Joel
Publication year - 2016
Publication title -
meteorological applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.672
H-Index - 59
eISSN - 1469-8080
pISSN - 1350-4827
DOI - 10.1002/met.1584
Subject(s) - radiosonde , gnss applications , environmental science , mean squared error , precipitable water , depth sounding , linear regression , meteorology , outlier , root mean square , statistics , global positioning system , mathematics , geography , computer science , water vapor , cartography , telecommunications , engineering , electrical engineering
Weighted mean temperature ( T m ) is a critical parameter in the estimation of precipitable water vapour from ground‐based Global Navigation Satellite System ( GNSS ) receivers. In the present study, three models of T m are developed for GNSS meteorological applications in the West African region in general and Nigeria in particular. The first is a least squares linear regression model based on National Centers for Environmental Prediction/National Center for Atmospheric Research reanalysis II data for Nigeria from 2010 to 2012. A regression of 37 330 data pairs of surface temperature and T m calculated from the vertical profiles of temperature above each grid node of the reanalysis model was used to produce the Nigerian Weighted Mean Temperature Equation‐I ( NWMTE‐I ) model after outlier data were removed. By using the same approach, NWMTE‐II was obtained from 11 433 radiosonde profiles from 24 sounding stations in the West African region for the period 2009–2013. NWMTE‐III was produced from a combination of the data used to build NWMTE‐I and NWMTE‐II . To evaluate the accuracy of these three models, they were compared with four global models based on T m obtained by using the integral method, which was the reference model for this study. The normalized mean absolute error, root‐mean‐square error, model efficiency, reliability index and correlation co‐efficient were used as performance indicators. The performances of the developed models were promising compared with the global models, justifying the importance of using measured meteorological parameters when estimating T m and fine tuning T m to specific areas or regions. Finally, NWMTE‐III is recommended for Nigerian users and NWMTE‐II for users in the West African region.