Two cases of ionospheric storms at the equatorial anomaly G.A. Mansilla a, * , M. Mosert b a Laboratorio de Ionosfera, Departamento de Fı ´sica, Universidad Nacional de Tucuma ´ n – CONICET, Av. Independencia 1800, 4000 San Miguel de Tucuma ´ n, Argentina b CASLEO, Av. Espan ˜a 1512 (sur), CP 5400, San Juan, Argentina Received 15 December 2005; received in revised form 26 June 2006; accepted 24 October 2006 Abstract This paper presents an evaluation of the IRI-2000 model, which includes the storm-time correction (STORM model), to predict foF2 during two intense geomagnetic storms (peak Dst < 100 nT) in the equatorial anomaly region. To this end, both IRI model outputs (with the STORM model included and without the STORM model) are compared with ionosonde data from a chain of stations located in the 278°–295° longitudinal range. The results show that the STORM model, even though it is not designed for low latitudes in general captures the variations of the measured values. But there is an overestimation of the foF2 data during the main phase and first stage of the recovery of the storms. Ó 2006 COSPAR. Published by Elsevier Ltd. All rights reserved. Keywords: Geomagnetic storm; STORM model; Equatorial anomaly 1. Introduction The knowledge of the critical frequency of the F2-layer, foF2 (proportional to the square root of the peak electronic density of F2-layer, NmF2) is essential for the prediction of ionospheric behaviour. Several empirical and semi-empiri- cal models (e.g., Barghausen et al., 1969; Anderson, 1973; Llewellyn and Bent, 1973; Bent et al., 1976; Anderson et al., 1987) have been developed to predict ionospheric parameters during quiet conditions. One of the most widely empirical models used to describe the ionosphere during magnetically quiet conditions is the International Reference Ionosphere, IRI (Bilitza, 1986, 1990, 2001) which provides median values of electron density, electron temperature, and ion composition as a function of height for a given loca- tion, time and sunspot number. This model is being contin- uously revised and updated through an international cooperative effort sponsored by the Committee on Space Research and the International Union of Radio Science. The latest version of the IRI contains the empirical storm-time ionospheric correction (STORM model) designed by Fuller-Rowell et al. (2000, 2001). It dependents on the intensity of the storm, and is a function of latitude and season (Araujo-Pradere et al., 2002a,b). A number of recent studies have compared the IRI- STORM predictions with ionosonde data at mid-latitudes during storm time periods (Araujo-Pradere and Fuller- Rowell, 2003; Araujo-Pradere et al., 2004; Mansilla et al., 2004). In general, the results indicate a significant improve- ment of IRI-2000 with the STORM model over IRI-2000 without the STORM model. In this paper, the performance of IRI-2000 with the correction of the STORM model included is evaluated in the equatorial anomaly (EA) region during two intense geomagnetic storms (peak Dst < 100 nT) by comparing ionosonde data with predictions obtained from the IRI model. The database used in this analysis, provided by the Space Physics Interactive Data Resource (SPIDR), are hourly values of foF2 from a chain of stations located in the 278°–295° longitudinal sector. The stations and their locations are listed in Table 1. 0273-1177/$30 Ó 2006 COSPAR. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.asr.2006.10.022 * Corresponding author. E-mail addresses: gmansilla@herrera.unt.edu.ar (G.A. Mansilla), mmosert@casleo.gov.ar (M. Mosert). www.elsevier.com/locate/asr Advances in Space Research 39 (2007) 668–673