Premium
A Deep Neural Network Model of Global Topside Electron Temperature Using Incoherent Scatter Radars and Its Application to GNSS Radio Occultation
Author(s) -
Hu Andong,
Carter Brett,
Currie Julie,
Norman Robert,
Wu Suqin,
Zhang Kefei
Publication year - 2020
Publication title -
journal of geophysical research: space physics
Language(s) - English
Resource type - Journals
eISSN - 2169-9402
pISSN - 2169-9380
DOI - 10.1029/2019ja027263
Subject(s) - millstone hill , ionosphere , radio occultation , gnss applications , incoherent scatter , electron density , total electron content , electron temperature , tec , earth's magnetic field , remote sensing , middle latitudes , scale height , electron , environmental science , physics , atmospheric sciences , meteorology , geology , geophysics , satellite , astronomy , magnetic field , quantum mechanics
The goal of this study is to present a new model for global topside electron temperature ( T e ) using a deep neural network (DNN) that is trained using measurements from incoherent scatter radars (ISRs). This study is also an investigation into whether this model can be used to generate the electron temperature in the topside ionosphere using GNSS ionospheric radio occultation (GNSS‐IRO) data as the input. ISR is one of the most reliable and long‐term sources to measure topside ionospheric electron density and plasma temperature information simultaneously. However, a drawback of ISR databases is the relatively poor spatial coverage due to the low number of ISR stations around the world. In contrast, GNSS‐IRO can be used to measure the global distributed electron density, butT einformation is not directly detected. The relationship between the electron density and the electron temperature has been investigated by many researchers, but these studies have not explicitly considered the parameters that are known to influence the electron temperature, such as solar and geomagnetic activity level, and the features of electron density profile ( h m F 2 , N m F 2 , and scale height). This study uses a DNN technique to create a new global topside electron temperature model from three submodels that have been trained using data from three ISR stations: Arecibo (low latitude), Millstone Hill (midlatitude), and Poker Flat (high latitude). This global model is trained using electron density profile information (e.g., vertical scale height [VSH], h m F 2, and N m F 2) and solar and geomagnetic activity ( F 10 . 7and K p , respectively) in addition to traditional spatial and temporal variables (e.g., local time, month, and latitude) as the independent variables. After theT emodel is developed,T einformation can be generated from the GNSS‐IRO electron density profiles using a newly createdN e − T emodel. This model's outputs are assessed with regard to out‐of‐sample ISR data and compared to the latest International Reference Ionosphere model. It is found that the electron temperature profiles from the DNN have a root‐mean‐square deviation of 259 K in the low‐latitude region (i.e., against Arecibo data), 254 K in the midlatitude region (i.e., against Millstone Hill data), and 314 K in high‐latitude region (i.e., against Poker Flat data), and all of them are smaller than the root‐mean‐square deviation from International Reference Ionosphere. An additional comparison between the model results versus the Thermosphere‐Ionosphere‐Electrodynamics General Circulation Model outputs is also conducted. A statistical analysis of the diurnal electron temperature profiles obtained from GNSS‐IRO is shown to agree with the Thermosphere‐Ionosphere‐Electrodynamics General Circulation Model outputs. The full codes and outputs can be found at the Zenodo ( 10.5281/zenodo.3637617 ).