February 23rd, 2018, 1:00 pm - 2:00 pm
December 7th, 1:00 pm - 2:00 pm
Leveraging distributed network of GNSS TEC measurements from India to develop ANN based TEC models
Dibyendu Sur, Narula Institute of Technology and University of Calcutta
Modern civilization has now become completely dependent on space-based navigation systems. The signals from these systems are being propagated through the ionosphere. The ionosphere is a part of earth?s atmosphere which contains the highest density of ions and free electrons. Thus ionosphere acts as the key source of error for all navigation systems such as Global Positioning System (GPS) as well as any satellite-based systems. The equatorial ionosphere is characterized by the phenomena of (a) Equatorial Ionization Anomaly (EIA) and (b) post-sunset generation of ionospheric irregularities. In response to the fountain effect, sharp latitudinal TEC gradients are developed in this region. Ionization in this region shows sharp diurnal, spatial and day-to-day variabilities. Standard ionospheric models are often unable to take into account the dynamic variabilities observed in this region. Ionospheric modeling efforts are extremely limited in India. In order to obtain accurate TEC prediction in this region, data driven Artificial Neural Network (ANN) based models have been designed along 77deg E, 88deg E and 121deg E longitudes covering wide latitudinal swaths (8deg -31deg N). These models are designed using dataset from the following stations.
(a) GNSS TEC receivers located along 88deg E, operated by Institute of Radio Physics and Electronics (IRPE), University of Calcutta over latitudinal swath of 22deg N-26deg N at four stations (i) Calcutta (22.58deg N, 88.38deg E geographic; magnetic dip 33.09deg), (ii) Baharampore (24.09deg N, 88.25deg E geographic; magnetic dip 36.35deg), (iii) Farakka (24.79deg N, 87.89deg E geographic; magnetic dip 38.15deg) and (iv) Siliguri (26.72deg N, 88.39deg E geographic; magnetic dip 41.54deg). The GPS TEC data recorded at Calcutta by IRPE is under Scintillation Network Decision Aid (SCINDA) program. Calcutta situated near the northern crest of the EIA in the Indian longitude sector.
(b) GPS-TEC receivers operated along 77deg E (latitudinal swath: 8deg -31deg N geographic) by Indian Space Research Organization (ISRO) and Airports Authority of India (AAI) under GPS and Geo-Augmented Navigation (GAGAN) program. The stations includes (i) Trivandrum (8.47deg N, 76.91deg E geographic; magnetic dip 0.9deg), (ii) Bangalore (12.95deg N, 77.68deg E geographic; magnetic dip 11.69deg), (iii) Hyderabad (17.44deg N, 78.47deg E geographic; magnetic dip 21.9deg), (iv) Bhopal (23.28deg N, 77.34deg E geographic; magnetic dip 33.95deg), (v) Delhi (28.58deg N, 77.21deg E geographic; magnetic dip 43.5deg), (vi) Shimla (31.09deg N, 77.07deg E geographic; magnetic dip 47.43deg).
(c) GNSS receives of International GNSS Service (IGS) at (i) Bangalore, (ii) Hyderabad, (iii) Lucknow (26.91deg N, 80.96deg E geographic; magnetic dip 42deg), (iv) Port Blair (11.64deg N, 92.71deg E geographic; magnetic dip 9.72deg) in India and at (v) Quezon City (14.64deg N, 121.08deg E geographic; magnetic dip 15.50deg, (vi) Hsinchu (24.80deg N, 120.99deg E geographic; magnetic dip 35.5deg) and (vii) Taoyuan (24.95deg N, 121.16deg E geographic; magnetic dip 35.76deg) along 121deg E.
These models are tested alongside standard ionospheric models such as (a) International Reference Ionosphere (IRI), (b) Parameterized Ionospheric Model (PIM) and (c) NeQuick and validated with actual measured data in order to see the applicabilities of these models in the equatorial region. Predictions from these models are then used to observe latitudinal and longitudinal TEC variabilities in this region. Latitudinal gradients of TEC estimated from the ANN TEC models are used to correlate with occurrence of scintillations near the anomaly crest in order to develop predictive capabilities. The effects of neutral wind on the performances of these models are tested by including neutral wind components as model inputs. While the initial versions of the models were developed for geomagnetic quiet periods, they have now been updated for magnetically disturbed conditions with emphasis on forecasting impact of geomagnetic storms on technological infrastructure.