The evolution of the SaRIF space weather forecasting system using new satellite data and improved methods for defining boundary conditions.

Since 2019, the SaRIF webpages on the ESA space weather portal have been used as a tool to visualise current conditions in the electron radiation belt and forecasts up to 24 hours ahead. SaRIF collects a variety of real-time data from satellites and geomagnetic indices and uses this information to determine current risk factors for satellites operating within the radiation belts around earth. Timely and accurate warnings of space weather events are crucial for satellite operators to minimise risk to fundamental services such as navigation, communications and science operations.

Over the lifetime of this tool, satellites have been used to provide data for the modelling systems used in SaRIF. In particular, the GOES satellites have provided electron flux data for the British Antarctic Survey Radiation Belt Model (BAS-RBM) that produces the SaRIF forecasts. As of April 2025, however, GOES-16 has been decommissioned and is no longer providing real-time data. We have analysed its successors, GOES 18 & 19, to determine which is a more suitable replacement for use in the SaRIF forecasting system. This includes an investigation into data quality, availability across the different energy channels and long-term data trends. We show that GOES-18 is the optimal choice, due to higher availability at the highest energy channels, crucial for calculation of the BAS-RBM boundary conditions, as well as its location closest to the equator, therefore measuring fluxes with a pitch angle closest to 90°.

Additionally, we have explored improved methods for calculating boundary conditions for the BAS-RBM thus enhancing the accuracy of the modelling underpinning our forecasting system. To determine the outer L* boundary conditions, we fit a kappa-type distribution to the electron flux data from GOES, allowing the flux to be calculated at any energy value, and then convert to phase-space density for providing the final boundary conditions.  Finally, we present a new approach to providing the low energy boundary conditions for the RBM, using Van Allen Probes data to determine the average profile along the low energy boundary for different levels of activity. Using a continuous data period of ³6 months, we demonstrate that using these activity dependent profiles improve our simulation results.