Cislunar Astrography

Jonathan McDowell, 2026 Jun 14

The report defines `astrographic regions' which are analogous to the boundaries I have used in my own work, notably McDowell 2018, 'The Edge of Space: Revisiting the Karman Line', Astra Astronautica 151, 668 and in McDowell 2020, the General Catalog of Space Objects (https://planet4589.org/space/gcat/, hereafter GCAT).

This note compares their choices with mine. In general I like their approach but prefer different specific numerical choices for the boundaries.

Table 1 of CC defines four regions divided by geocentric distance: (Rs(E), Rs(L) = mean surface level of Earth and Moon)

CC Generic Name	CC Cislunar Name	GCAT name		CC radius	GCAT radius
Surface environment	Terrestrial Environment	Earth		Rs(E)+100 km	Rs(E)+80 km
Surface environment	Lunar Environment	Luna		Rs(L)+100 km	Not yet defined
Near-body space	Near-earth space	Near-earth space		Rs(E)+50000 km	Rs(E)+145688 km (EL1:4)
Near-body space	Near-Lunar space	Lunar space		Rs(L)+60000 km	Rs(L)+64445 km (EM Hill sphere)
Celestial neighbourhood	Cislunar space	Deep space	Deep Earth orbit	Rs(E)+1.5Mkm	Rs(E)+1.5Mkm (SE Hill sphere)
Deep space			Interplanetary space	-	Not yet defined
Deep space			Interstellar space		Not yet defined

In my work, I use "Deep space" to include both RAND's "Deep space" and their "Celestial neighbourhood". Some details:

CC introduces the term "minorbit" to represent the lowest stable orbit around a given body. I like the concept, but not the word. I suggest 'katosphere' from the Greek Κάτω (under)- which I feel is much more euphonious.
For Earth they pick 100 km; my reasons for preferring 80 km are given in my 2018 paper.
For the Moon they also pick 100 km based on the effects of mascons. However, no detailed analysis is presented for where that 100 km value comes from - I'd love to see one. Short lived lunar orbits can be as low as 10 km at their lowest point.
For Mars, they say `about 200 km' based on the atmosphere. Some day I plan to repeat the 2018 Karman analysis for a standard Martian atmosphere, but I bet the number I'll get is lower than that.

In GCAT I make extensive use of Hill spheres as the primary boundary separating different regions with different central bodies, in agreement with the approach used by CC.
However, we are seeing increasing use of orbits that linger near the Earth-Moon (EM) Hill boundaries with repeated crossings. CC uses the concept of 'Lagrange zones', which is a useful way of treating situations like this, but do not give any recipe for determining even an approximate size for these zones - another open question that I've been worrying about and warrants research.

The biggest difference between CC and GCAT is the edge of Near-earth space;they pick 50000 km and I pick 145688 km (the 1:4 lunar orbit resonance distance). My choice was influenced by my estimate of where three-body effects become important; CC also mention a so-called `Worden line' at 2.7GEO (about 110000 km) with a similar motivation.
I would argue that 50000 km is rather on the low side with many scientific satellites in the 50000 to 100000 km apogee range.

I distinguish interplanetary and interstellar space. The fraught argument about where the edge of the solar system is has mostly been about where the solar wind transitions to the interstellar medium, thanks to the Voyagers crossing that boundary, but I am more interested in the gravitational boundary between the Sun and the galactic potential, a Sun-Galaxy Hill sphere equivalent.
This could probably be calculated quite easily by someone proficient with galpy, which I am not (yet).
The outer gravitational boundaries of Milky Way space and of Local Group and Virgo Supercluster space could be modelled similarly based on the cosmographic work of Tonry and colleagues.