An updating of the SIRM model

The SIRM model proposed by Zolesi et al. (1993, 1996) is an ionospheric regional model for predicting the vertical -sounding characteristics that has been frequently used in developing ionospheric web prediction services (Zolesi and Cander, 2014). Recently the model and its outputs were implemented in the framework of two European projects: DIAS (Digital upper Atmosphere Server; http://www.iono. noa.gr/DIAS/) (Belehaki et al., 2005, 2015) and ESPAS (Near-Earth Space Data Infrastructure for e-Science; http://www.espas-fp7.eu/) (Belehaki et al., 2016). In this paper an updated version of the SIRM model, called SIRMPo1, is described and corresponding outputs in terms of the F2-layer critical frequency (foF2) are compared with values recorded at the mid-latitude station of Rome (41.8 degrees N, 12.5 degrees E), for extremely high (year 1958) and low (years 2008 and 2009) solar activity. The main novelties introduced in the SIRMPo1 model are: (1) an extension of the Rome ionosonde input dataset that, besides data from 1957 to 1987, includes also data from 1988 to 2007; (2) the use of second order polynomial regressions, instead of linear ones, to fit the relation foF2 vs. solar activity index R-12; (3) the use of polynomial relations, instead of linear ones, to fit the relations A(0) vs. R-12, A(n), vs. R-12 and Y-n vs. R-12,R- where A(0), An and Yn are the coefficients of the Fourier analysis performed by the SIRM model to reproduce the values calculated by using relations in (2). The obtained results show that SIRMPo1 outputs are better than those of the SIRM model. As the SIRMPo1 model represents only a partial updating of the SIRM model based on inputs from only Rome ionosonde data, it can be considered a particular case of a single -station model. Nevertheless, the development of the SIRMPo1 model allowed getting some useful guidelines for a future complete and more accurate updating of the SIRM model, of which both DIAS and ESPAS could benefit. (C) 2017 COSPAR. Published by Elsevier Ltd. All rights reserved.