BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:STI Seminar - Spatio-temporal variabilities of ionospheric feature
 s observed using University of Calcutta VHF Radar and GNSS receivers: Case
  studies and modeling
DTSTART;VALUE=DATE-TIME:20260505T090000Z
DTEND;VALUE=DATE-TIME:20260505T100000Z
DTSTAMP;VALUE=DATE-TIME:20260519T015135Z
UID:indico-event-11339@ictp.it
DESCRIPTION:\n	Equatorial ionospheric irregularities intersect and interfe
 re with transionospheric radio signals often resulting in serious degradat
 ion of their performance including that of satellite-based systems and ser
 vices. While such phenomena are associated with geomagnetic disturbed cond
 itions in the mid and high latitudes\, they could occur even under geomagn
 etic quiet conditions in the equatorial and low latitudes. The radio frequ
 ency signals affected also varies across a wide spectrum extending from VH
 F to L-band and often S-band as well. Some novel coordinated experimental 
 observations were recorded during February-March 2024 using University of 
 Calcutta VHF radar (CUVR) operating at 53 MHz at the Ionosphere Field Stat
 ion\, Haringhata (22.93°N\, 88.37°E geographic\; 35°N magnetic dip) and
  GPS L-band measurements from North Bengal University (NBU) (26.71°N\, 88
 .35°E geographic\; 42°N magnetic dip). On March 18\, 2024\, the VHF rada
 r backscattered signals noted a number of irregularity patches over a comm
 on ionospheric volume with the GPS observations from NBU more-or-less arou
 nd the same $me interval. Similar analyses have also been done for other d
 ates in February and March 2024. These novel coordinated measurements vali
 date the simultaneous coexistence of ionospheric irregularities of varying
  scale sizes ranging from centimetres to metres using the VHF radar and GP
 S. However\, an outstanding issue to consider is the occurrence of ionosph
 eric irregularities even at L-band at a magnetic dip of 40°N during late 
 evening and midnight hours. This brings forth the question whether these i
 rregularities could be attributed to equatorial plasma bubbles or to some 
 remnants of transitional mid- latitude plasma density structures. This coo
 rdinated approach offers a comprehensive understanding of the spatial and 
 temporal characteristics of these phenomena\, which are critical for impro
 ving communication and navigation systems in the low-latitude region. Some
  of the background low-latitude ionospheric features were measured using t
 he MEDEA receiver operational at the Institute of Radio Physics and Electr
 onics (IRPE)\, University of Calcutta as part of the ESA AMIC project\, wh
 ich exhibited depletions in Total Electron Content (TEC) associated with t
 he occurrence of ionospheric irregularities. However\, in order to demonst
 rate the day-to-day variabilities of the occurrence of irregularities\, io
 nospheric reconstruction using GNSS data was used. This technique requires
  accurate measurement of TEC which may be provided from a reliable model. 
 Accurate short-term prediction of ionospheric TEC is essential for mitigat
 ing positioning errors in GNSS\, particularly over low-latitude regions ch
 aracterized by strong nonlinear variability. This study proposes HTR- X++\
 , a physics-informed hybrid deep ensemble framework that integrates tempor
 al deep learning with gradient-boosted regression. The model supports mult
 i-horizon forecasting (1-hour and 3-hour ahead) and incorporates thermosph
 eric composition (O/N2) to account for ionosphere–thermosphere coupling.
  The framework is evaluated using GNSS-derived TEC data from IRPE across c
 ontrasting seasonal regimes and further validated at an independent statio
 n NBU. Results demonstrate substantial improvement over persistence baseli
 nes\, achieving up to ~40% reduction in RMSE and correlation values approa
 ching unity (R ≈ 0.98–0.99). These results establish HTR-X++ as a robu
 st\, interpretable\, and operationally viable framework for real-time TEC 
 forecasting.\n\n//indico.ictp.it/event/11339/
LOCATION:E. Fermi Building Marconi Lab
URL://indico.ictp.it/event/11339/
END:VEVENT
END:VCALENDAR