Matt Wilson wrote:
This is an artifact of supporting the "what do you want" Basic world-building sequence, and it produces a gradation of planetary size with distance from the star that has no physical basis.
Can you expand on that a bit? I understand MMWR and how the farther out you get, the easier it is for a world to hold on to an atmosphere. Is there something else in your comment here suggesting that worlds are more/less likely to be a certain physical size at a certain distance?
GURPS Space has two world-building sequences, Basic and Advanced. The Basic Sequence (steps 1 to 14, on pp.73–98) starts with "world type" (a combination of atmospheric type and surface material/conditions in fifteen categories not counting gas giants) and progresses through atmospheric composition and mass, ocean coverage, surface temperature, density/mass/gravity, resources and habitability, settlement type, tech level, population, unity, government/society/mode of production (muddled together, because GURPS does that), control rating (law level), income and trade, bases and installations. It is designed so that you can roll for each step or make a choice of what is possible given results so far. The Advanced Sequence (steps 15–39) starts with the number of stars in the system, then progresses through their masses, the system's age, the stars' luminosity classes and spectral types, surface temperature, and luminosity, orbits of the companion stars (in the case of a multiple system); then for each star the existence and arrangement of its gas giants, the orbits of it "first gas giant" (the one defining the arrangement of its system) and any designed planet to be inserted, other orbits, what planets are in them, what moons those planets have, and then for each planet and moon, what type it is. And at that point the advanced system recapitulates the basic system, Steps 26 to 29 referring to Steps 3, 4, 5, and 6. Steps 30 and 31 introduce day lengths and tidal braking. Steps 32 to 39 refer to Steps 7 to 14. Tidal locking and spin-orbit resonance are dealt with as afterthoughts.
The Basic Sequence starts with world type because that's the fundamental question that restricts everything else when you are using it as a design system. Then it proceeds through a bunch of other design choices (that can be rolled) to determine a bunch of other things — surface temperature, atmospheric composition and column mass, ocean cover, surface temperature — that go together to determine its albedo and greenhouse effect, which gives you its black body temperature, which tells you (in step 21) where it has to go in a system, when you come to designing a system. That is really slick. I'm in awe. It was so slick that I could implement the whole thing, including letting the user design a planet and inserting it into the right place in the system, as an Excel workbook with no macros.
The problem arises when the Advanced Sequence "re-uses code" from steps 3 to 6 as steps 26 to 29. It saves word-count, explanation, and complexity, and it is what makes the system so slick; but it introduces a subtle bias. In the Advanced Sequence you place the orbits, then you calculate the black-body temperatures at those orbits and then you roll on a table of world types that are possible at those BB temperatures. That means that the ratio of tiny:small:standard:large worlds remains the same right across each band of temperatures. But those aren't world sizes, that are categories of minimum molecular weight retained. The planets in each category get smaller as you go further out from the star. What ought to happen is that 'tiny' worlds (i.e. ones that won't retain nitrogen) get less common (because the largest of them become 'small' at low temperature). But instead of getting less common they get smaller. And since the largest 'large' planets don't start getting replaced by small gas giants you get a phenomenon of smaller planets at low temperatures than at high ones, which is not physical. The distribution of masses ought to remain the same, and the world types ought to slide over that distribution.
It's one of my main reasons for wanting to replace the GSpace world design sequence with my own.
Curious to know what you'd change/keep. Something more like 2300 where you determine diameter/density first and then find out what that gets you?
Well, I'm still in doubt about how to place the gas giants. I'd prefer not to have to place them by a separate preliminary process and squeeze the terrestrials in afterwards, and I feel very unconfident about having to simulate migrational history. But assuming that that problem is solved I'd like to go to each orbit in turn and determine randomly its initial endowment of iron and siderophilic material, of stone and lithophilic material, of chalcophiles, of ices, and of gasses. Calculate mass and radius, then determine black body temperature when the star enters the main sequence. Figure what boils off and remove it. Generate an initial rotational rate. Then generate moons to the same stage. Calculate tidal forces, move moons accordingly (including removing moons that spiral in to the Roche limit and ones that spiral out to the Hill Sphere radius, and adjust the planet's rotation rate. I hope to be able to do that well enough with a single step, but there is a problem with the geostationary orbit radius moving out as the planet's rotation slows. I might have to do that by a series of steps. Any way, once I have game-date orbits and rotation rates, turn the star up to its current luminosity, calculate the energy budgets (insolation + tidal kneading + radiogenic geothermal + energy of crystallisation for the cooling metallic core + outgoing thermal radiation) and the blackbody temperature. Then re-calculate Jeans escape. Then work through the volatiles in order of increasing boiling point to work out the albedo and greenhouse effect of after each one comes into the atmosphere, and whether that produces a temperature and pressure high enough to melt and then to boil the next. When I get to water, if CO2
is sufficient to melt it and there is enough geothermal heat flux to suppose vulcanism, reduce CO2
to a point where a carbonate-silicate cycle stabilises climate or zero, whichever is lower. Calculate the surface temperature and pressure. If water boils, assume runaway greenhousing (i.e. return all the CO2
to the air. Figure out how long the planet has had liquid water at the surface (in face of solar brightening) and the flux of light suitable to drive photosynthesis, and figure out whether it has passed its oxygen catastrophe. Estimate ocean and ice cover and vegetation, calculate albedo, recalculate climate….
It's an algorithm to be executed by a computer, obviously.