Neural Network Vertical Drift (NNVD) Model

Abstract

The Neural Network Vertical Drift (NNVD) Model, version 1.0 -------------------INTRODUCTION-------------------The NNVD model is a model of the ionospheric vertical drift across the globe within the equatorial region (+/- 5 degrees of the geomagnetic Equator). The model can be used to predict the global ionospheric vertical drift near the geomagnetic equator. For this version of the model (version 1.0), usage is encouraged to be restricted between 1 January 2000 - 22 March 2026. The model is proposed to be updated when there is need to do so. The year span of encouraged usage will be expanded in subsequent versions of the model. The model is developed by the method of artificial neural networks as detailed in the paper: https://doi.org/10.1029/2025EA004622, which is included in the model folder. -------------------USAGE-------------------IMPORTANT: Before running the program, users need to download and install the free MATLAB Compiler Runtime (MCR), version R2025b (25.2), found here: mathworks.com/products/compiler/matlab-runtime.html (this needs to be done only once) The program comes in a folder which contains all required files The main program to run is named NNVD.exeWhen launched, it generates a graphical user interface which is simple and easy to use Users can either produce diurnal profiles or spatial mapsUsers can also generate the diurnal profiles for all the days in a year by ticking the checkbox on the user interfaceWhen the program runs, it creates the figure for the user specified details, and stores the corresponding data in a folder named Output (which is created within the same folder where the NNVD program is located) -------------------REFERENCE-------------------Habarulema, J. B., Okoh, D., Yizengaw, E., Habyarimana, V., Pezzopane, M., Fagundes, P. R., et al. (2026). A new global climatological model of the equatorial ionospheric vertical E × B drift: Integrating ground‐based magnetometer, radar, and satellite data sets. Earth and Space Science, 13, e2025EA004622. https://doi.org/10.1029/2025EA004622. -------------------CORRESPONDENCE-------------------John Bosco HabarulemaSouth African National Space Agency (SANSA) Space Science, Hermanus, South AfricaEmail: jhabarulema@sansa.org.za