Assessing Upper Atmosphere Models for Navigating a Crowded LEO Environment

Today there is an expanding drive to populate Low Earth Orbit (LEO) with thousands of satellites for remote sensing and global communication. For example, SpaceX is in the process of creating a 12,000-satellite Starlink constellation in LEO, with more than 4,600 SmallSats in orbit as of August 2023. The growth in the number of LEO objects directly increases the probability of unintentional collisions between objects due to accumulating space debris. This could lead to the runaway Kessler syndrome, where unavoidable cascading collisions occur, leading to a potentially unusable LEO orbital domain. In addition to a congested LEO space environment, the rapid rise of this solar cycle suggests that the predicted solar maximum between 2024-2027 could be higher than the previous solar maximum, thus causing higher perturbations due to drag from atmospheric density on LEO satellites. Despite these increasingly hazardous conditions, there is still no consensus among agencies and companies on how to quantify and predict the thermospheric environment through which these objects are orbiting. This presentation outlines current state-of-the-art thermospheric density models, describing their performance, computation time, required operational space weather input parameters, and notes for implementation. We include models that are at a technology readiness level of eight or nine, meaning that the model is currently being run on an operational system or the model has validated performance under operational conditions.

Current state-of-the-art models utilize input parameters (either proxies or indices) to generate a global density field nowcast and forecast. The current thermospheric density model used by the Combined Space Operations Center (CSpOC) in operations, the High Accuracy Satellite Drag Model (HASDM), applies corrections to an empirical density model every three hours using observations of 80+ calibration satellites. The ISO International Standard 14222 on Earth’s upper atmosphere recommends using the US Naval Research Laboratory Mass Spectrometer and Incoherent Scatter radar 2.0 (MSIS 2.0) empirical model for relative constituent abundances and Jacchia-Bowman 2008 (JB2008) model for mass densities related to satellite drag. JB2008 is the background density model applied in HASDM. It is an empirical model that uses solar and geomagnetic indices and proxies as inputs to obtain a global thermospheric density nowcast and forecast. MSIS is also an empirical model, but it utilizes information from only two space weather indices: the Ap index and F10.7 proxy. Unlike MSIS, JB2008 utilizes multiple solar irradiance parameters, S10, M10, Y10 and F10.7, to identify the energy deposition in particular layers of the thermosphere. The Drag Temperature Model 2020 (DTM2020) is an empirical density model used at multiple European space research labs and integrated into several NASA astrodynamics software packages.

Although empirical models have historically been applied in operations due to their faster and more efficient nature, as computation power becomes more accessible, the utilization of physics-based models, like the Thermosphere-Ionosphere-Electrodynamics General Circulation Model (TIE-GCM), have become more tractable. The Whole Atmosphere Model and Ionosphere Plasmasphere Electrodynamics (WAM-IPE) is another example of a physics-based model. It was developed by the National Oceanic and Atmospheric Administration (NOAA) and University of Colorado researchers, and has been operational at NOAA since July 2021. WAM-IPE was recently designated as one of the density models in the Department of Commerce’s new Space Traffic Management (STM) system being developed.

The models we will review in our presentation include, but are not limited to, DTM2020, HASDM, JB2008, MSIS2.0, TIE-GCM2.0, and WAM-IPE. An analysis is provided in which each model’s performance is compared during quiet and storm conditions during the month of April 2023. Models are evaluated globally against HASDM densities and locally against Gravity Recovery And Climate Experiment Follow-On (GRACE-FO) satellite accelerometer-derived density data. A propagation analysis is also included in which satellites are propagated through each model’s density field during the April 2023 storm and quiet conditions. Each model’s corresponding satellite trajectory is compared to the satellite trajectory generated by propagating it through the HASDM density field. The resulting LEO object trajectory prediction errors are quantified at various orbit altitudes. Results indicate that regardless of the model, recent data is needed to constrain the model and remove bias; something for which only HASDM is currently doing operationally. Overall, this presentation provides a comparison between state-of-the-art density models to identify possible areas of improvement for particular models and for thermospheric density modeling as a whole.