Controls
Presets
Parameters
Model checks
The science
What is measured. Apollo 11, 14 and 15 and two Soviet Lunokhod rovers left arrays of corner-cube retroreflectors on the Moon. An observatory fires a short 532 nm laser pulse through its telescope, a cube array sends a few photons straight back, and the round-trip time gives the distance, d = c τ / 2. The distance is not constant: the orbit is an ellipse, so the Moon swings between about 356 and 407 thousand km (the mean orbit used here, with e = 0.0549, gives 363 to 406 thousand km; the real extremes include solar perturbations).
Orbit. Mean anomaly M = M0 + 2πt/T with T = 2π√(a³/G(ME+m)) (27.3 days); Kepler's equation M = E - e sin E is solved by Newton iteration and r = a(1 - e cos E). The page does not include the solar perturbations, so it is a two-body orbit.
Photon budget. The uplink beam is a Gaussian with FWHM θup = √((1.22 λ/D)² + seeing²), so on the Moon its standard deviation is σ = θup r / 2.355 (a few km). The fraction landing on an array of equivalent radius ra is 1 - exp(-ra²/2σ²). Each cube (38 mm) diffracts the return into θret = √((1.22 λ/d)² + θva²), about 18 microradians including velocity aberration, so the return footprint on Earth is several km wide while the telescope is metres; the fraction collected is again a Gaussian encircled energy. Both fractions fall as r-2, hence the famous r-4 law: the expected photoelectrons per shot are n = (E/hν) pup pdn η with η the product of transmit and receive optics, detector quantum efficiency, array efficiency and two-way atmospheric transmission exp(-2τ0/sin(elevation)). With n well below 1 most shots return nothing, which is why a real station fires 20 pulses a second for minutes.
Detection. The number of photoelectrons in a shot is Poisson distributed, so the chance of at least one return is 1 - e-n; background counts in the range gate add a noise rate (sky and sunlit lunar surface, counts per second times gate width). The range precision of N returns is the timing jitter (200 ps = 30 mm) divided by √N.
Why the Moon recedes. The Moon raises tidal bulges on Earth; because Earth turns faster than the Moon orbits, the bulge leads and pulls the Moon forward. The torque is N = (3/2)(k2/Q) G m² RE5/a6. The orbital angular momentum L = m√(G ME a) grows by exactly what Earth's spin loses, so da/dt = 2aN/L = K a-11/2. The effective k2/Q = 0.0258 (ocean tides included) reproduces the measured recession of about 3.8 cm per year; Earth's day then lengthens by about 2 ms per century from the lunar tide.
What the checks compare. (1) The Kepler solver against perigee and apogee a(1 -/+ e), vis-viva, and the speed obtained from finite differences of the solved positions. (2) The Gaussian encircled-energy formula against a Monte Carlo of 200,000 photon landing points. (3) A Monte Carlo of 100,000 Poisson laser shots against 1 - e-n and the mean count. (4) The r-4 scaling between perigee and apogee, and for a doubled range. (5) A 500-million-year RK4 integration of the recession against the closed form a13/2 = a013/2 + (13/2)Kt, an angular-momentum ledger between Earth spin and orbit, and a hand-derived rate per unit k2/Q.
Simplifications, honestly. The array efficiency, atmospheric extinction, detector and optics numbers are plausible illustrative values, not those of any real station, so the absolute photoelectron rates are not predictions of a real observatory. The Moon's libration, the finite pulse length, speckle, scintillation, pointing errors and the three-body perturbations are ignored. The epoch slider extrapolates the present tidal model with a constant k2/Q, which is wrong over geological time (the real tidal response depended on the ocean basins); it only shows the form of the recession law. The scene is not to scale and the orbit's eccentricity is exaggerated in the pictures.