1] A new GNSS Solar Flare Activity Indicator (GSFLAI) is presented, given by the gradient of the ionospheric Vertical Total Electron Content (VTEC) rate, in terms of the solar-zenithal angle, measured from a global network of dual-frequency GPS receivers. It is highly correlated with the Extreme Ultraviolet (EUV) photons flux rate at the 26-34 nm spectral band, which is geo-effective in the ionization of the mono-atomic oxygen in the Earth's atmosphere. The results are supported by the comparison of GSFLAI with direct EUV observations provided by SEM instrument of SOHO spacecraft, for all the X-class solar flares occurring between 2001 and 2011 (more than 1000 direct comparisons at the 15 s SEM EUV sampling rate). The GSFLAI sensitivity enables detection of not only extreme X-class flares, but also of variations of one order of magnitude or even smaller (such as for M-class flares). Moreover, an optimal detection algorithm (SISTED), sharing the same physical fundamentals as GSFLAI, is also presented, providing 100% successful detection for all the X-class solar flares during 2000-2006 with registered location outside of the solar limb (i.e., detection of 94% of all of X-class solar-flares) and about 65% for M-class ones. As a final conclusion, GSFLAI is proposed as a new potential proxy of solar EUV photons flux rate for strong and mid solar flares, presenting high sensitivity with high temporal resolution (1 Hz, greater than previous solar EUV irradiance instruments), using existing ground GNSS facilities, and with the potential use as a solar flare detection parameter.

This paper aims to compare the performance of three different artificial neural network techniques for tourist demand forecasting: a multi-layer perceptron, a radial basis function and an Elman network. We find that multi-layer perceptron and radial basis function models outperform Elman networks. We repeated the experiment assuming different topologies regarding the number of lags used for concatenation so as to evaluate the effect of the memory on the forecasting results. We find that for higher memories, the forecasting performance obtained for longer horizons improves, suggesting the importance of increasing the dimensionality for long-term forecasting.

Purpose – This study aims to apply a new forecasting approach to improve predictions in the hospitality industry. To do so, the authors developed a multivariate setting that allows the incorporation of the cross-correlations in the evolution of tourist arrivals from visitor markets to a specific destination in neural network models.\ud Design/methodology/approach – This multiple-input-multiple-output approach allows the generation of predictions for all visitor markets simultaneously. Official data of tourist arrivals to Catalonia (Spain) from 2001 to 2012 were used to generate forecasts for one, three and six months ahead with three different networks.\ud Findings – The study revealed that multivariate architectures that take into account the connections between different markets may improve the predictive performance of neural networks. Additionally, the authors developed a new forecasting accuracy measure and found that radial basis function networks outperform the rest of the models.\ud Research limitations/implications – This research contributes to the hospitality literature by developing an innovative framework to improve the forecasting performance of artificial intelligence\ud techniques and by providing a new forecasting accuracy measure.\ud Practical implications – The proposed forecasting approach may prove very useful for planning purposes, helping managers to anticipate the evolution of variables related to the daily activity of the\ud industry.\ud Originality/value – A multivariate neural network framework has been developed to improve forecasting accuracy, providing professionals with an innovative and practical forecasting approachPeer ReviewedPostprint (author's final draft

Although vertical total electron content (VTEC) forecasting is still an open subject of research, the use of predictions of the ionospheric state at a scale of several days is an area of increased interest. A global VTEC forecast product for two days ahead, which is based exclusively on actual Global Positioning System (GPS) data, has been developed in the frame of the International Global Navigation Satellite Systems (GNSS) Service (IGS) Ionospheric Working Group (IGS Iono‐WG). The UPC ionospheric VTEC prediction model is based on the Discrete Cosine Transform (DCT), which is widely used in image compression (for instance, in JPEG format). Additionally, a linear regression module is used to forecast the time evolution of each of the DCT coefficients. The use of the DCT coefficients is justified because they represent global features of the whole two‐dimensional VTEC map/image. Also, one can therefore introduce prior information affecting the VTEC, for instance, smoothness or the distribution of relevant features in different directions. For this purpose, the use of a long time series of final/rapid UPC VTEC maps is required. Currently, the UPC Predicted product is being automatically generated in test mode and is made available through the main IGS server for public access. This product is also used to generate two days ahead preliminary combined IGS Predicted product. Finally, the results presented in this work suggest that the two days ahead UPC Predicted product could become an official IGS product in the near future.

The medium‐scale traveling ionospheric disturbances (MSTIDs) constitute the most frequent ionospheric wave signatures. We propose a method for detecting the number of simultaneous MSTIDs from a time series of high‐pass‐filtered Vertical Total Electron Content (VTEC) maps and their parameters. The method is tested on the VTEC map corresponding to a simulated realistic scenario and on actual data from dual‐frequency Global Positioning System (GPS) measurements gathered by +1200 GPS receivers of the GPS Earth Observation Network (GEONET) in Japan. The contribution consists of the detection of the number of independent MSTIDs from a nonuniform sampling of the ionospheric pierce points. The problem is set as a sparse decomposition on elements of a dictionary of atoms that span a linear space of possible MSTIDs. These atoms consist of plane waves characterized by a wavelength, direction, and phase on a surface defined, the part of the ionosphere sounded by the GEONET (i.e., 25°N to 50°N of latitude and 125°E to 155°E of longitude). The technique is related to the atomic decomposition and least absolute shrinkage and selection operator. The geophysical contribution of this paper is showing (a) the detection of several simultaneous MSTIDs of different characteristics, with a continuous change in the velocity; (b) detection of circular MSTID waves compatible by time and center with a specific earthquake; (c) simultaneous superposition of two distinct MSTIDs, with almost the same azimuth; and (d) the presence at nighttime of MSTIDs with velocities in the range 400–600 m/s.