Ionospheric forecasts for the European region for space weather applications

Abstract

This paper discusses recent advances in the implementation and validation of the Solar Wind driven autoregression model for Ionospheric short-term Forecast (SWIF) that is running in the European Digital upper Atmosphere Server (DIAS) to release ionospheric forecasting products for the European region. The upgraded implementation plan expands SWIF's capabilities in the high latitude ionosphere while the extensive validation tests in the two solar cycles 23 and 24 allow the comprehensive analysis of the model's performance in all terms. Focusing on disturbed conditions, the results demonstrate that SWIF's alert detection algorithm forecasts the occurrence of ionospheric storm time disturbances with probability of detection up to 98% under intense geomagnetic storm conditions and up to 63% when storms of moderate intensity are also considered. The forecasts show relative improvement over climatology of about 30% in middle-to-low and high latitudes and 40% in middle-to-high latitudes. This indicates that SWIF is able to capture on average more than one third (35%) of the storm-associated ionospheric disturbances. Regarding the accuracy, the averaged mean relative error during storm conditions usually ranges around 20% in middle-tolow and high latitudes and 24% in the middle-to-high latitudes. Our analysis shows clearly that SWIF alert criteria were designed to effectively anticipate the ionospheric storm time effects that occurred under specific interplanetary conditions, e.g., cloud Interplanetary Coronal Mass Ejections (ICMEs) and/or associated sheaths. The results provide valuable input in advancing our ability in predicting the space weather effects in the ionosphere for future developments, and further work is proposed to enhance the model forecasting efficiency to support operational applications.

A new short-term forecasting model for the total electron content storm time disturbances

“…Since there are some evidence available today that indicate dependence of the ionospheric response on the solar wind drivers of the storms (e.g. Tsagouri et al, 2017) and moreover, the performance of the SWIF model had been proved to be particularly successful in forecasting the impact of intense CME-related events (Tsagouri & Belehaki, 2015), the focus in the present analysis was given in this type of storm events in order to build progress on a solid and clear basis. The distinction of the interplanetary drivers of the storms is compatible with the list of the ICMEs that is available at http:// www.srl.caltech.edu/ACE/ASC/DATA/level3/icmetable2.htm (Richardson & Cane, 2010).…”

“…Multiple onsets are detected for several of the storm events analyzed here. It should be noted that only one of the storm events, the one occurred in the time interval 16-20 March 2013, was "missed" by SWIF's ADA (see Tsagouri & Belehaki, 2015 for the evaluation of the results of the SWIF's ADA).…”

“…In Table 3, the performance of the proposed model is assessed through the estimation of several metric parameters: the Mean Error (ME), the Mean Absolute Error (MAE), the Mean Relative Error (MRE) and the Root Mean Square Error (RMSE) and the relative improvement over climatology. The latter is calculated by using the following formula (see for example Araujo-Pradere & Fuller-Rowell, 2002;Tsagouri, 2011;Tsagouri & Belehaki, 2015):…”

“…On the other hand, a full range of regional foF2 forecasting products and services up to 24 h in advance are routinely provided by the European Digital Upper Atmosphere Server (DIAS, http://dias.space.noa.gr). These are supported by the implementation of the Solar Wind driven autoregression model for Ionospheric Forecast (SWIF À Tsagouri & Belehaki, 2006Tsagouri et al, 2009;Tsagouri, 2011;Tsagouri & Belehaki, 2015) and include: warnings for upcoming storm-time ionospheric disturbances, single site forecasts for all DIAS locations and regional forecasting maps for the European area. DIAS forecasting services are also integrated with ESA/SSA/ SWE to provide ionospheric forecasts over whole Europe for HF propagation users (http://swe.ssa.esa.int/web/guest/diasfederated), as part of the European Ionosonde Service (EIS).…”

“…The storm time component is based on the empirical Storm Time Ionospheric Model (STIM, Tsagouri & Belehaki, 2006, which assumes the IMF disturbances observed at the L1 point by ACE spacecraft as drivers for upcoming ionospheric foF2 storm-time disturbances. The forecasting efficiency of the model has been assessed in several occasions and particularly successful performance has been identified under the occurrence of ionospheric disturbances associated to interplanetary coronal mass ejections (ICMEs) detected at L1 point (see Tsagouri & Belehaki, 2015 for a comprehensive description of the SWIF model and the performance tests).…”