摘要：This work describes several important improvements made to the International Reference Ionosphere UPdate (IRI UP) method, and a careful validation of its performances under disturbed conditions.The IRI UP method has been improved developing an algorithm capable to properly filter wrongly autoscaled ionosonde data to be assimilated, avoiding the use of these in the assimilation process.Furthermore, the preliminary quality check used to choose the variogram model in the Universal Kriging method has been replaced with a new quality check routine (NQCR), based on statistical tests carried out using the variables Q 1, Q 2, and cR , built on variogram’s residuals.NQCR objectively identifies the best variogram model from which to get more reliable effective indices maps to be ingested in the IRI model to obtain updated fo F2 and hm F2 maps.IRI UP has been applied on 30 different time intervals, between January 1, 2004, and December 31, 2016, characterized by moderate, strong, and severe geomagnetic conditions, over the European region.A statistical comparison between IRI UP and IRI at the truth sites located at Fairford (51.7°N, 1.5°W, UK) and San Vito (40.6°N, 17.8°E, Italy), for fo F2 and hm F2, has been performed.From the statistical validation clearly emerges how IRI UP, for fo F2, performs significantly better than IRI, for each of the 30 geomagnetic storms considered.Regarding hm F2, IRI UP performances are lower than those for fo F2, although still better than IRI ones.In the light of the results achieved in this investigation, the IRI UP method represents an interesting approach to Space Weather forecast in the ionospheric domain.
关键词：IRI UP; Universal Kriging; Ionospheric data assimilation; Nowcasting maps
