Improved Forecasting of Operational Solar and Geomagnetic Indices

Improved thermospheric density forecasts are a critical need identified by the Space Weather Operations, Research, and Mitigation (SWORM) Working Group, a Federal interagency coordinating body [National Space Weather Strategy and Action Plan, 2019]. This work will improve thermospheric density prediction by reducing forecast error up to 65% during large geomagnetic storms. We will validate two replacement forecast drivers used by operational models.nbsp;Given the increased debris environment in Low Earth Orbit (LEO) there is strong interest by U.S. agencies, companies, and international organizations for managing hazards from debris collision. There are two areas in which we will improve forecast neutral thermosphere densities. First, we will validate an existing observation-based solar irradiance index, i.e., the S10 produced by National Solar Observatory (NSO) SIFT model, that can accurately extend operational forecasts out to 7 days. Second, we validate a combined solar wind Gated Recurrent Unit (GRU) networks using solar wind observations algorithm (fromnbsp;Hu) plus a solar feature data-driven, machine-learned (ML) algorithms (Logit) (from Swiger) that together createnbsp;a forecast Disturbance storm time (Dst) geomagnetic index for shortnbsp;(hours)nbsp;and longnbsp;(days) time frames, respectively. The SEFT S10 index will improve 3-7 day predictions for the debris collision maneuver and reentry windows.nbsp;The Hu+Swiger Dst forecast capability will help resolve the single largest problem in thermosphere density forecasting, i.e., large error in densities during geomagnetic storms resulting from poorly forecast storm magnitude and timing by Anemomilos Dst.nbsp;SET will validate SIFT S10 as well as thenbsp;GRUnbsp;and Logit Dst data outputs in Phase I by comparing them to existing SET S10 and SETnbsp;Anemomilos Dst that are the operational baselines used for the past decade. This projectnbsp;will design a method to automatically feed our newnbsp;Dst predictions to the CME Scoreboard in coordination with CCMC.