Using novel remote sensing techniques to enhance local prediction models of ionosphericresponse to space weather: a RISER project approach

Novel remote-sensing techniques can be used to inject vital "oncoming storm" data into upgraded prediction models. Here we present several possibilities being investigated under the auspices of the RISER project, whose goal is to improve space weather forecast times by up to four days, as well as their accuracy. Here, in this early stage of the project one strand is to investigate how best to integrate such new data. We describe how to leverage the capabilities of our existing IONwork software to produce GNSS-based TEC disturbance maps, and use that processed data in concert with information from the heliosphere to create remote-plus-local combined prediction models, comparing both traditional and machine learning techniques. Heliospheric information can be extracted, for example, from L1 monitors: however, within the RISER project we aim at expanding the heliospheric information available by including observations of interplanetary scintillation and corresponding tomographic reconstructions to map solar-wind or CME features structures present in the Sun-Earth interplanetary space. Such extended data will also provide longer lead times than the approximately one hour advance notice given by L1 data, making it easier to create forecasts that are both timely and more accurate.