Premium
A Neural Network‐Based Model for Daytime Vertical E×B Drift in the Indian Sector
Author(s) -
Chaitanya P. Pavan,
Patra A. K.
Publication year - 2020
Publication title -
journal of geophysical research: space physics
Language(s) - English
Resource type - Journals
eISSN - 2169-9402
pISSN - 2169-9380
DOI - 10.1029/2020ja027832
Subject(s) - daytime , artificial neural network , computer science , meteorology , geodesy , geology , geography , atmospheric sciences , artificial intelligence
Observations of ionospheric vertical E × B drift are limited in the Indian sector. In this paper, we present an artificial neural network‐based model for inferring ionospheric daytime vertical E × B drift using ground‐based magnetometer and radar observations in the Indian sector. We have used E × B drift estimated using 150‐km echoes observed by the Gadanki mesosphere‐stratosphere‐troposphere radar and the difference in ground magnetic fields (Δ H ) observed from Tirunelveli (magnetic latitude 0.4°N) and Alibag (magnetic latitude 13.0°N), providing a measure of electrojet current. We have used a feed forward neural network with error back propagation to train the network and estimated E × B drifts for the period 2006–2011. The model successfully reproduces the radar‐observed vertical E × B drifts. The model E × B drifts are also found to agree exceedingly well with those measured using the CINDI onboard the C/NOFS. The model seasonal mean drifts are found to differ considerably from those of Scherliess‐Fejer model. The neural network‐based model presented here is first of its kind from the Indian sector, presenting local time, day‐to‐day, and seasonal variations of daytime equatorial F region E × B drifts during magnetically quiet conditions. Results and the potential of the model for studying equatorial electrodynamics in the Indian sector are presented.