Improving Forecasting Ability of GITM Using Data‐Driven Model Refinement

At altitudes below about 600 km, satellite drag is one of the most important and variable forces acting on a satellite. Neutral mass density predictions in upper atmosphere are therefore critical for (a) designing satellites; (b) performing adjustments to stay an intended orbit; (c) collision avoidance maneuver planning. Density have great deal uncertainty, including model biases misrepresentation atmospheric response energy input. These may stem from inaccurate approximations terms Navier-Stokes equations, unmodeled physics, incorrect boundary conditions, or parameterizations. Two commonly parameterized source thermal conduction eddy diffusion. Both components transfer heat thermosphere. Determining how well major constituents (N2, O2, O) as conductors will effects temperature changes source. This work shows effectiveness using retrospective cost refinement (RCMR) technique at removing bias caused by different sources within Global Ionosphere Thermosphere Model. Numerical experiments, Challenging Minisatellite Payload Gravity Recovery Climate Experiment data during real events used show that RCMR can compensate both parameterizations drivers. eliminating before storm allows more accurate throughout storm.