EVALUATION OF LAYERED INTEGRATED WATER VAPOUR OVER WEST AFRICA USING RADIOSONDE AND SATELLITE DATA

Abstract

The Layered Integrated Water Vapour (IWV) has been studied using yearly averages of air temperature and relative humidity at twenty five selected weather stations which were further grouped into four climatic regions (Coastal, Derived, Guinea and Sahel) over West Africa for the years 2010 to 2014 retrieved from ERA-Interim. It uses a December 2006 version of the European Centre for Medium-Range Weather Forecast (ECMWF) Reanalysis Integrated Forecast Model (IFS Cy31r2) which covers dates from 1 January 1979 to present, at spectral resolution of about 80km. Radiosonde data for the same parameter, as era interim, were also retrieved for Abuja of the year 2014 from Nigerian Institute of Meteorology, Oshodi, Lagos. The IWV studied was however classified into Upper, Middle and Lower layers for the purpose of the study. It was observed that maximum values of IWV at the different layers considered were obtained in the Guinea Savannah region especially at Abuja with values 0.12 g/cm3, 0.14 g/cm3 and 0.18 g/cm3 in the Upper layer, Middle and Lower layer respectively while minimum values were obtained at Dakar with values of 0.0005 g/cm3 in the Upper layer, 0.03 g/cm3 in the Middle layer and 0.04 g/cm3 in the Lower layer. In comparison, Era-Interim data at the Lower and Middle layer has considerable relative values with that of Radiosonde and hence it could be used instead of Radiosonde data but reverse is the case for the Upper layer as radiosonde values are quite negligible to that observed in Era-Interim. Also, a Multivariate Linear Regression Model which deals with modeling and analyzing several variables when the focus is on the relationship between a dependent variable and one or more independent variables was developed for each of the station and regions. Here, IWV has been modeled using air temperature and relative humidity. The statistical indicators such as Coefficient of determination (R2), Mean Percentage Error (MPE), Mean Bias Error (MBE), and Root Mean Square Error (RMSE) with confidence level of 95% for R2 and desirable lower values of MPE, MBE and RMSE were calculated to monitor the efficiency of the developed model. It was found out that IWV was well monitored across the stations and climatic regions.