Scott Shambaugh, Leonid Space
https://doi.org/10.64861/CCDF7573
Keywords: orbital lifetime, atmospheric drag, ballistic coefficient, space weather, solar cycle, Monte Carlo, orbit propagation, space debris, post-mission disposal, LEO
Abstract:
Space weather has upended low Earth orbit operations this solar cycle. Atmospheric densities have been 2-3x higher than pre-cycle consensus predictions, invalidating the lifetime assumptions in current business cases and national security missions. Multiple operators including SpaceX, Capella Space, and Planet have reported significant impact from these elevated drag levels, and NASA’s Swift observatory is the target of a $30M rescue mission to save it from an unexpectedly early reentry. Despite this operational urgency, the tools and data available to operators and regulators for predicting satellite lifetimes do not operate on real-time orbit data, and have not been validated under realistic predictive conditions where the future is not yet known._x000D_
_x000D_
This paper presents a comprehensive validation of Leonid Space’s satellite lifetime prediction pipeline, backtested against over 1000 satellites that deorbited between 1961 and 2025, spanning six solar cycles including the ongoing Solar Cycle 25 maximum. Our pipeline combines a custom semianalytic orbit propagator, automatic ballistic coefficient (CB) estimation from two-line element (TLE) data, atmospheric density corrections via data assimilation of space debris trajectories, NASA Marshall Space Flight Center solar cycle forecasts, and calibrated Monte Carlo analysis. We simulate the final year of flight for each satellite and score predicted deorbit dates against actuals using the continuously ranked probability score (CRPS) and calibration curves scored with the Cramer-von Mises criterion._x000D_
_x000D_
Our validation methodology progressively removes hindsight through three scenarios. First, a Perfect Knowledge scenario validates the end-to-end toolchain using known space weather and CB extracted from the actual TLEs over the last year of life. Second, a Historical Conditions scenario tests our ability to extrapolate drag parameters by estimating CB from only the 30 days of flight data prior to the simulation start. Third, a Fully Predictive scenario uses forecasted space weather and estimated CB, mimicking the realistic operational case of evaluating a spacecraft flying today. At each stage we calibrate Monte Carlo dispersions to capture unmodeled process noise and ensure the output probability distributions match the true range of outcomes with no additional uncertainty padding required._x000D_
_x000D_
To our knowledge, this is the first large-scale validation of long-term satellite lifetime predictions that reports quantitative results for the fully predictive case. Prior studies in this space have either used known space weather, examined small satellite sets, or focused on short-horizon reentry events. We qualify “large-scale” as greater than 100 satellites and “long-term” as prediction horizons beyond 30 days, which screens out the large body of work focused on parametric model evaluations, small selected cohorts, or short-duration conjunction assessments and reentry location predictions. Rather than point estimate metrics such as median absolute error, this work is also one of the first to apply rigorous probabilistic scoring to lifetime predictions. Our scoring framework – CRPS for prediction accuracy, rank percentile calibration curves for distribution quality, and the Cramer-von Mises criterion for comparing dispersed configurations – provides a reproducible methodology for evaluating and comparing probabilistic lifetime forecasts. We present this as a template that the community can adopt for benchmarking lifetime prediction tools against each other and against future improvements._x000D_
_x000D_
Orbit-derived ballistic coefficient estimation is a critical enabler. Our approach extends accurate lifetime prediction to the over 98% of tracked debris objects that lack mass and area estimates in the existing DISCOS database. Once CB is properly characterized, we show that solar cycle forecasting dominates all other error sources – higher fidelity propagators and atmosphere models provide negligible improvement at lifetime-relevant timescales. This has direct implications for where the SSA community should invest effort: better space weather forecasting and robust on-orbit drag characterization are the only inputs that will further improve operational lifetime prediction._x000D_
_x000D_
These results establish a quantitative performance baseline for operational satellite lifetime prediction. The demonstrated accuracy supports mission planning, regulatory compliance with FCC and ESA deorbit requirements, constellation revenue forecasting, and operational decision-making for LEO operators navigating an increasingly dynamic atmospheric environment._x000D_
Date of Conference: September 15-18, 2026
Track: Atmospherics/Space Weather