Article
Aerospace Engineering · Orbital Mechanics · ⏱ ~12 min read · Last updated: 9 July 2026

Atmospheric Drag and Orbital Decay — Why Nothing in LEO Stays Up Forever

Space is not truly empty, even 400 km up. A whisper-thin residual atmosphere constantly rubs against every satellite in low Earth orbit, stealing a tiny sliver of orbital energy on every pass. That theft compounds: altitude drops, density rises, drag increases, and the decay accelerates — until, months or years later, what began as an imperceptible drift ends in a fiery reentry.

TL;DR: Residual air at 200-700 km altitude constantly drags on satellites, and because density rises exponentially as altitude drops, decay accelerates over time — slow for years, then just days near the end. Ballistic coefficient (area-to-mass ratio) and the 11-year solar cycle both strongly change how fast any given satellite decays and reenters.

1. The Residual Atmosphere in LEO

Between roughly 200 and 700 km altitude, Earth's thermosphere and exosphere still contain a measurable, though extremely thin, gas density. This is thin enough that satellites orbit for years, but not so thin that drag becomes negligible over a mission lifetime.

Approximate atmospheric density by altitude (quiet Sun): 200 km: ρ ≈ 2.5×10⁻¹⁰ kg/m³ 400 km: ρ ≈ 2-6×10⁻¹² kg/m³ (varies 3-5× with solar activity) 600 km: ρ ≈ 1-5×10⁻¹³ kg/m³ 800 km: ρ ≈ 1-3×10⁻¹⁴ kg/m³ (negligible decay over decades) Rule of thumb: below ~250 km, decay happens in days to weeks Above ~600 km, decay can take centuries — geostationary orbit (35786 km) is effectively immune to atmospheric drag entirely

2. The Drag Equation and Energy Loss

Drag force opposes velocity and depends on the satellite's "ballistic coefficient" — its area-to-mass ratio weighted by drag coefficient — exactly the parameter explored in the ballistic coefficient article for reentry vehicles, but here acting slowly over thousands of orbits instead of one violent pass.

Drag force: F_d = ½ρv²Cd·A (opposes velocity direction) Specific orbital energy: ε = −μ/(2a) Energy loss rate per orbit (circular orbit approximation): Δε per orbit ≈ −(F_d / m) · v · T_orbit Since F_d/m = ½ρv²(Cd·A/m), a LOW area-to-mass ratio decays SLOWER (this is why dense objects like solid rocket bodies decay slower than large lightweight objects like deployed solar arrays or expended upper stages with big surface area)

3. Estimating Decay Rate

A useful analytic approximation (King-Hele's classic formula) relates altitude loss per orbit directly to local density and the satellite's ballistic parameter, assuming a near-circular orbit.

Altitude loss per orbit: Δh ≈ −2π·(Cd·A/m)·ρ(h)·a² (a = semi-major axis, ρ(h) = local atmospheric density) Example: 500 kg satellite, A = 4 m², Cd = 2.2, at 400 km: Cd·A/m = 2.2 × 4 / 500 = 0.0176 m²/kg At ρ(400km) ≈ 4×10⁻¹² kg/m³, a ≈ 6771 km: Δh per orbit ≈ −2π × 0.0176 × 4×10⁻¹² × (6.771×10⁶)² ≈ −5.1 m per orbit (≈ 92 min orbital period) → About 80 m/day, or roughly 30 km/year at this altitude — consistent with the ISS's observed decay before reboosts

4. The Solar Cycle Effect

Solar extreme-ultraviolet (EUV) radiation heats the thermosphere, causing it to expand and puff outward — dramatically increasing density at any fixed altitude during solar maximum compared to solar minimum, roughly on an 11-year cycle.

Solar minimum

Thermosphere contracted; density at 400 km can be 5-10× lower than during solar maximum — satellites decay slowly.

Solar maximum

Thermosphere expanded; density at any given altitude spikes, dramatically shortening satellite lifetimes and forcing more frequent reboosts.

Skylab (1979)

Reentered years earlier than planned partly because solar activity was higher than predicted, expanding the atmosphere and increasing drag.

Space weather forecasting

Satellite operators must incorporate solar-cycle predictions into conjunction analysis and end-of-life deorbit planning.

5. JavaScript Orbital Decay Simulator

// King-Hele style orbital decay estimator (circular-orbit approximation)
function exponentialDensity(h, rho0 = 3.9e-12, h0 = 400000, H = 60000) {
  // H is local scale height near LEO, varies with altitude and solar activity
  return rho0 * Math.exp(-(h - h0) / H);
}

function simulateDecay(h0Alt, bcInv, days) {
  // bcInv = Cd*A/m in m²/kg (inverse ballistic coefficient)
  const muEarth = 3.986004418e14, rEarth = 6371000;
  let h = h0Alt;
  const history = [];
  for (let day = 0; day < days; day++) {
    const a = rEarth + h;
    const rho = exponentialDensity(h);
    const orbitalPeriod = 2 * Math.PI * Math.sqrt(a ** 3 / muEarth);
    const orbitsPerDay = 86400 / orbitalPeriod;
    const dhPerOrbit = 2 * Math.PI * bcInv * rho * a * a;
    h -= dhPerOrbit * orbitsPerDay;
    history.push({ day, altitudeKm: h / 1000 });
    if (h < 120000) break; // effectively reentering
  }
  return history;
}

// A 500 kg satellite (Cd*A/m ≈ 0.0176) starting at 400 km altitude
const trace = simulateDecay(400000, 0.0176, 3000);
console.log(`Days to reenter: ${trace.length}`);

6. Real Satellites and Their Fates

International Space Station

Requires periodic reboosts (usually via visiting cargo vehicles) to counteract roughly 2 km/month of decay at ~400-420 km altitude.

Starlink satellites

Deliberately operate at relatively low altitudes (~550 km) partly so that a failed satellite decays and burns up within roughly 5 years — a deorbit-by-design mitigation for space debris.

Tiangong / Skylab

Both experienced uncontrolled or partially controlled reentries after operational life ended, illustrating how quickly decay accelerates below ~200 km.

CubeSats

Very high area-to-mass ratio (large solar panels, tiny mass) gives them naturally fast decay — often just 1-5 years even from 400-500 km, useful for automatic debris mitigation.

Frequently Asked Questions

Why does the ISS need periodic reboosts?

At its roughly 400 km altitude, the ISS still experiences residual atmospheric drag from the exosphere, which continuously removes orbital energy and lowers its altitude by around 2 km per month on average. Without periodic reboost burns from visiting spacecraft, the station would decay and reenter within a few years.

Why does orbital decay speed up as a satellite gets lower?

Atmospheric density increases exponentially as altitude decreases, so drag force grows exponentially too, while the satellite's orbital energy loss rate is proportional to that drag force times velocity. This creates a runaway feedback: lower altitude means denser air means faster decay means even lower altitude, which is why the final descent from around 200 km to reentry can take just days after years spent slowly decaying from 400-500 km.

How does solar activity affect satellite decay rates?

Solar EUV and X-ray output heats and expands the thermosphere, increasing atmospheric density at satellite altitudes by a factor of 5-10 between solar minimum and solar maximum. This means the same satellite can decay several times faster during peak solar activity than during a quiet Sun period, which is why long-term orbital lifetime predictions must account for the roughly 11-year solar cycle.

Why do lightweight objects with large surface area decay faster?
Decay rate depends on the area-to-mass ratio (technically Cd·A/m). A large, lightweight object like a spent upper stage with a big flat panel, or a fragment of an exploded satellite, has a high Cd·A/m ratio and therefore experiences proportionally larger deceleration per orbit than a dense, compact object of the same mass — this is why debris fragments from the same breakup event often decay at very different rates.
Is atmospheric drag used deliberately to deorbit satellites?
Yes. Many small satellites and constellations are designed to naturally decay and burn up within 25 years of end-of-mission (an international space-debris mitigation guideline), sometimes assisted with a deployable drag sail or balloon that increases the effective area-to-mass ratio to speed up the natural decay process without using propellant.
How accurately can we predict when a satellite will reenter?
Predictions weeks or months in advance carry large uncertainty because of unpredictable solar activity swings and the satellite's changing attitude (tumbling changes its effective drag area). Uncertainty shrinks dramatically in the final hours before reentry, though the exact reentry location typically remains uncertain to within a large swath of the ground track until just a couple of orbits before impact.
What is different about decay at geostationary altitude?
At 35786 km, atmospheric density is so vanishingly small that drag-induced decay is completely negligible over any realistic mission timescale — geostationary satellites are instead retired to a "graveyard orbit" a few hundred kilometers above GEO using their remaining propellant, rather than being deorbited through the atmosphere like LEO satellites.
🛰️ Open Orbital Mechanics →