Premium
Comparing different assimilation techniques for the ionospheric F 2 layer reconstruction
Author(s) -
Gerzen Tatjana,
Minkwitz David,
Schlueter Stefan
Publication year - 2015
Publication title -
journal of geophysical research: space physics
Language(s) - English
Resource type - Journals
eISSN - 2169-9402
pISSN - 2169-9380
DOI - 10.1002/2015ja021067
Subject(s) - ionosonde , ionosphere , electron density , daytime , radio propagation , quiet , radio wave , data assimilation , physics , computational physics , atmospheric sciences , electron , geodesy , meteorology , geology , geophysics , quantum mechanics , astronomy
From the applications perspective the electron density is the major determining parameter of the ionosphere due to its strong impact on the radio signal propagation. As the most ionized ionospheric region, the F 2 layer has the most pronounced effect on transionospheric radio wave propagation. The maximum electron density of the F 2 layer, N m F 2 , and its height, h m F 2 , are of particular interest for radio communication applications as well as for characterizing the ionosphere. Since these ionospheric key parameters decisively shape the vertical electron density profiles, the precise calculation of them is of crucial importance for an accurate 3‐D electron density reconstruction. The vertical sounding by ionosondes provides the most reliable source of F 2 peak measurements. Within this paper, we compare the following data assimilation methods incorporating ionosonde measurements into a background model: Optimal Interpolation (OI), OI with time forecast (OI FC), the Successive Correction Method (SCM), and a modified SCM (MSCM) working with a daytime‐dependent measurement error variance. These approaches are validated with the measurements of nine ionosonde stations for two periods covering quiet and disturbed ionospheric conditions. In particular, for the quiet period, we show that MSCM outperforms the other assimilation methods and allows an accuracy gain up to 75% for N m F 2 and 37% for h m F 2 compared to the background model. For the disturbed period, OI FC reveals the most promising results with improvements up to 79% for N m F 2 and 50% for h m F 2 compared to the background and up to 42% for N m F 2 and 16% for h m F 2 compared to OI.