Skip to content

Space environment and orbits

solid pillar 1 · Physical modeling and domain theory

In one minute

  • Two-body motion gives conic sections and six orbital elements. Everything real is a perturbation on that.
  • \(J_2\) is not a small correction for mission design — it drives nodal regression, which is what makes sun-synchronous orbits possible.
  • Delta-v budgets are the currency of mission design, and the rocket equation makes them exponentially expensive.
  • Drag in LEO varies by an order of magnitude with solar activity; orbit lifetime predictions inherit that uncertainty.
  • Differential charging, not total dose, causes most spacecraft anomalies — the discharge, not the accumulation.

Key results

Two-body motion:

\[ r = \frac{h^2/\mu}{1 + e\cos\theta}, \qquad v = \sqrt{\mu\left(\frac{2}{r} - \frac{1}{a}\right)}, \qquad T = 2\pi\sqrt{\frac{a^3}{\mu}} \]

with \(\mu_\oplus = 3.986\times10^{14}\ \mathrm{m^3/s^2}\), \(R_\oplus = 6378\) km.

Hohmann transfer between circular orbits \(r_1 \to r_2\):

\[ \Delta v_1 = \sqrt{\frac{\mu}{r_1}}\left(\sqrt{\frac{2r_2}{r_1+r_2}}-1\right), \qquad \Delta v_2 = \sqrt{\frac{\mu}{r_2}}\left(1-\sqrt{\frac{2r_1}{r_1+r_2}}\right) \]

Plane changes are brutally expensive: \(\Delta v = 2v\sin(\Delta i/2)\). A 30° change at LEO velocity costs about 3.9 km/s — comparable to reaching orbit in the first place. Do plane changes at apoapsis where \(v\) is low, or not at all.

\(J_2\) perturbation. Earth's oblateness causes secular drift of the node and argument of perigee:

\[ \dot{\Omega} = -\frac{3}{2}\frac{J_2 R_\oplus^2\sqrt{\mu}}{a^{7/2}(1-e^2)^2}\cos i, \qquad \dot{\omega} = \frac{3}{4}\frac{J_2R_\oplus^2\sqrt{\mu}}{a^{7/2}(1-e^2)^2}\left(5\cos^2 i - 1\right) \]

\(J_2 = 1.0826\times10^{-3}\). Two consequences worth memorising: a sun-synchronous orbit chooses \(i\) (about 98° at LEO) so \(\dot\Omega = 0.9856°\)/day, matching Earth's orbit around the Sun; and \(\dot\omega = 0\) at the critical inclination \(i = 63.4°\), used by Molniya orbits to keep apogee over the northern hemisphere.

Drag:

\[ \dot{a} = -\frac{\rho C_D A}{m}\sqrt{\mu a} \]

Thermospheric density at 400 km varies by roughly an order of magnitude between solar minimum and maximum. Ballistic coefficient \(m/(C_D A)\) is what distinguishes a cubesat that reenters in months from one that lasts years.

Environment by orbit:

Hazard Where Effect
Atomic oxygen LEO material erosion
Trapped radiation (Van Allen) MEO, transits total dose, SEUs
GCR and solar particle events all, worse beyond LEO SEUs, latch-up, crew dose
Surface/deep charging GEO, auroral LEO ESD, the leading anomaly cause
Micrometeoroid and debris LEO especially impact, puncture
Thermal cycling all fatigue at every eclipse entry/exit

Mental model

Orbital mechanics is energy and angular momentum bookkeeping. Semi-major axis is energy, eccentricity and inclination shape it, and every manoeuvre is a purchase paid for in delta-v with the rocket equation setting the exchange rate. Because that rate is exponential, mission design is mostly an exercise in not manoeuvring — choosing an orbit whose natural perturbations do the work for you. Sun-synchronous orbits are the cleanest example: let \(J_2\) precess the orbit for free rather than spending propellant.

Numerics / practice

  • Propagate with the right fidelity. SGP4 for TLEs (and only with TLEs — mixing TLE data into a high-precision propagator is meaningless), numerical integration with a geopotential model, drag and third-body terms for anything serious.
  • Budget margin on drag. Solar activity forecasts carry large uncertainty; lifetime estimates should be quoted as ranges.
  • Use variational or symplectic integrators for long propagations so energy error does not accumulate — see astrophysics › N-body.
  • Check the epoch and frame. ECI vs ECEF, and TLE epoch age, cause more errors than the dynamics.
Failure modes

Assuming Keplerian motion over long spans. \(J_2\) alone precesses a LEO node by several degrees per day. A two-body propagation is wrong within one orbit for any ground-track or coverage question.

Plane change budgeted as if it were small. \(2v\sin(\Delta i/2)\) at LEO makes even modest inclination changes cost kilometres per second. Launching into the right plane is nearly always cheaper than changing it.

TLEs used with a non-SGP4 propagator. TLE elements are mean elements defined by the SGP4 model itself. Feeding them into a numerical propagator as osculating elements introduces kilometre-scale errors immediately.

Drag modelled at a single density. Using a mean atmosphere over a solar cycle can be off by 10× at 400 km. Reentry-date predictions built on it are not predictions.

Total dose treated as the whole radiation story. Single-event upsets and latch-up depend on particle flux and energy spectrum, not accumulated dose, and a part can meet its dose budget and still be destroyed by one heavy ion.

Surface charging conflated with deep dielectric charging. They have different timescales, different environments and different mitigations. Grounding everything conductive helps the first; only shielding and bleed paths help the second. Differential charging between isolated surfaces is what actually discharges.

Eclipse thermal cycling omitted. A LEO spacecraft sees about 5,500 thermal cycles per year. Fatigue, not peak temperature, sizes many structures and solar-array hinges.

Worked example

Sun-synchronous inclination falls straight out of \(J_2\):

import numpy as np

mu, Re, J2 = 3.986004418e14, 6378.137e3, 1.08263e-3
rate_needed = np.deg2rad(360.0 / 365.2422) / 86400.0     # rad/s, to track the Sun

for alt_km in (400.0, 700.0, 1000.0):
    a = Re + alt_km * 1e3
    coef = -1.5 * J2 * Re**2 * np.sqrt(mu) / a**3.5
    i = np.degrees(np.arccos(rate_needed / coef))
    print(f"altitude {alt_km:6.0f} km  ->  sun-synchronous inclination = {i:.2f} deg")

altitude    400 km  ->  sun-synchronous inclination = 97.03 deg
altitude    700 km  ->  sun-synchronous inclination = 98.19 deg
altitude   1000 km  ->  sun-synchronous inclination = 99.48 deg

Slightly retrograde, and rising with altitude — the familiar ~98° for Earth-observation orbits, obtained for free from a perturbation rather than paid for in propellant.

Connections

Sources

  • Vallado, Fundamentals of Astrodynamics and Applications.
  • Tribble, The Space Environment.
  • Archive: Aerospace/AERSP597I_SpaceEnvInteraction.