Saturday, April 27, 2013

XHIP counts

So, my census results for the XHIP catalog, for a 100 parsec wide cube centered on Sirius.

In the cube: 10,117 stars (plus sol, not in the file)
  • 811 of those are multiple star systems, but not variable stars
  • 3835 are listed is variable, but do not list components
  • 2101 are both variable and multiple - of course, that may include eclipsing binaries.
There are 50,760 stars more outside the cube, and additionally 57,078 with no distance information.

Monday, April 22, 2013

Firming up the sector model

First, just to note the XHIP computed heliocentric galactic-coordinates position for Sirius is (X,Y,Z) is -1.8,-1.9,-0.4.

Second, why am I so hung up on getting as wide a space as possible?  Fundamentally, because while ships move outwards in a linear way, stars pile up with the cube of the distance. with the rules we have discussed so far, and technologies where 10-20 milli-Gs (a few cm S^-2) is a lot of acceleration we are talking 3 months or so to transit from one average star to the next - a distance of a couple of parsecs.

So, if 100 parsecs is 50 jumps (optimistic, not a straight line) and for each jump we spend 3 months traversing the system to position for the next jump (reasonable in terms of time, but optimistic because we will have to "stop for gas") that's 150 months or more than 12 years to reach the edge of the map.  So I am worrying about nothing, it would seem.  Cut it in half - an 8th the number of stars - and you still have 6 years to the edge.  And a lot of those systems are going to be junk not worth visiting - we can probably store an adequate amount of information about it in a dozen bytes.

Another note about 100 parsecs - the galactic lens is only about 300 parsecs thick, so the 100 parsec radius would be 2/3rds the thickness of the galaxy!.

My next census will be a 50 parsec radius cube (so still 10^6  cubic parsecs) divided into sectors of radius 10 (so 125 cubes 20 parsecs across) centered on Sirius. 

Sunday, April 21, 2013

Stellar density and star catalogs

I've done a little analysis of XHIP that illustrates the issues with any current parallax catalog.

I took the stars with good distances in the new reduction, and determined which 5-parsec thick "shell" the fell into around sol.  While the number of stars in each shell is stable, of course the volume of each shell is greater, and the density drops off considerably.

This should not come as a surprise to anyone, of course.  Visual magnitude is a key factor in being able to make any observation about a star, and that (of course) drops off with square of distance.

With the current state of the art, that means that any catalog I produce which attempts to reproduce the "real" density of stars in local space will require a fair amount of guesswork.  Keep in mind, my objective is to produce something suitable for a game or for an SF story.  Story lines can be preserved by emphasizing the things we know, and playing down what what is made up.  For a game, the MST3K Mantra comes to mind.

How many stars do we need to backfill?  Well, taking the local density as "correct" - which is a poor base of itself, and will need refinement, we should have 361179 stars out to 105 parsecs, and only have distances for 24183.  This means that we will need to generate (or at least assign distances for) 336996 stars - overall, 93% or so of objects. 

Graphically by distance this looks like:

All of which will be invalidated by the ESA in about five years, but I'll survive.

A side note: if I want my space to be a 200 parsec  cube centered on Sirius, I should re-work this to 180 parsecs radius or so.

Saturday, April 20, 2013

Jump distance rules

I've been considering an interesting way to limit distance of jumps (for games and SF of course)  for a while  - something simple and not particularly concerned with units.

A simple approach would be to use  a relationship like:

Where M is the mass of the destination, and r is in parsecs.

Looking at a few examples, this would let us jump to Sirius (mass of primary+companion roughly 3 solar masses) over about 5.5 parsecs.  One could jump to our Sol  system from as far away as 3.2 parsecs.  To reach a A small M main sequence star like Gliese 185 one would need to be within 1.4 parses.

We are not all that close to any very massive stars; however this limit does have an interesting asymmetry; there will be many jumps that are possible in one direction buy not the other.

Saturday, April 13, 2013

Thoughts on ships, stars and science fiction

I've been trying to think of ways to combine maneuverability and fuel capacity in a single ship.  Here's a concept that might work.  It assumes that the only propulsion system that really does the job is some sort of very high exhaust velocity "atomic rocket".  Also, its clear that our travel system will need a lot of fuel; but size (length particularly) is opposed to maneuverability.  To be maneuverable, you want a sphere with no more in it than you need to fight.  To travel for long distances and maintain the craft you need a lot of impedimenta.

Here's my compromise:
The main propulsion is our probably-radioactive "rocket".  The combat module has the main engineering systems, significant fuel, main weapons, and enough habitat and life support for enough crew to fight the ship in minimal comfort.

The Fuel module is just what it says on the tin, and comes with rotating habitats, extended life support, and support and maintenance as well as reserve combat crew.  Landers, shuttles, expendable reloads, the "FTL space jump drive if it is a separate system"  and the like would also be part of that module.

The combat module would "park" the logistics module well away from the expected action, un-dock, and fight with in a low mass configuration with minimized moment of inertia.

I've also pulled down the All Sky catalog from Vizier; to save myself some time I decided to let the server calculate the galactic coordinates.

Now, if you can think back to reading SF in the 1970s, consider: in what decade you would have expected the above sentence to be written; perhaps what century?

Also in the realms of no longer SF, it looks like about three years from now I am going to need fibre-op and a couple of terrabytes of disk space. Cool or what?

Friday, April 12, 2013

Back to start?

The more I work with the HYG data (and star catalogs in general) the more I think I should go back to first principles with a new starting catalog, such as Kharchenko's All Sky Compiled Catalog.  It is a compilation from all the standard references, but also contains some useful additional record flags, and also goes down in some cases to 14th apparent magnitude, which will aid in the "backfill" objective. 

It does not have all the "name" cross references from HYG, but I an use the HIP catalog number to link that in.  It will doubtless have some errors, but all star catalogs have errors and I am coming to understand that (especially as you dive into the low magnitude objects) it is very difficult to say with absolute confidence that two catalogs that offer objects at a given location really are talking about the same star.

Also, since I will not be integrating catalogs myself, I can produce galactic references from a single self-consistent source. It's important to keep in mind that my ultimate objective is a fictional construct that it is as constant with the real world as I can make it, but with no regrets were it is not.  Science marches on, no matter what level of alignment with current knowledge I manage.

Something to keep in mind: If I want to go out to 100 parsecs (mostly at that extreme by photometric parallax) and get the faint end of main sequence dwarfs I need to get down to 20th apparent magnitude, which exceeds Kharchecko's low end.

Of course, using the estimate of .14 stars per parsec^3, that's over 586,000   systems.  Bit much to be useful; 50 parsecs would reduce to 73,000 stars, which is a lot more manageable. 

Sunday, April 7, 2013

More Hipparcos data

The Strasbourg Astronomical Data Centre has the Hipparcos multiple and variable data ; I'll get that merged into my own reference catalog, and see where that leads me.

Friday, April 5, 2013

Washington Double Star index included

I was surprised to find only 105 hits, but then these are visual binaries, which can very easily mean apparent binaries.  I'll try adding cross indexes on alternate catalogs, them move on to the MCS data.  A smaller dataset, but he fact that they are fully computed may make for a higher hit percentage.

Thursday, April 4, 2013

Some details on spin gravity side effects

On the spacefuture website

Thoughts on shuttle technology

Here's an interesting engine and spaceplane concept - the Skylon project from Britain. There are also some nice flight concept CGI clips floating around youtube.  The basic idea is to build velocity with an airbreathing engine before climbing to orbit with rockets; the engine itself is multi-mode.

While I am not particularly saying that this is "the" way shuttles will work in my future, it is interesting to see that research into viable reusable orbital technology is not dead.

Tuesday, April 2, 2013

And a curveball from reality

I've identified three good sources for multiple star data.  Two, importantly, have "HIP" cross references.
 The downside of these discoveries is that some of the entries I have already processed may be in those catalogs, with no notation of multiplicity in it's HYG. data line.  Such is life;  I will start the splitting by incorporating the cross reference in the names along with the critical annotations in the ongoing "rejects" data.  Incorporating these data into the re-run to exclude multiples from "simple solitary" processing will be simple enough, although I will end up with a lot more "leftovers" when I am done.

It is also time to start thinking about the file structure and database structure to capture the orbital data.

Almost 3/4 done...

but its the easy 3/4.  I have masses for 18,170  stars, leaving 6,688 catalog entries yet to solve; but I have also used up all the the simple, Astro 101 style, spectral types.  That's not the end of the world, but it does mean I need new techniques to process the single star entries, and also that it's time to start thinking about how to process the binary stars.  In fact, I probably need to look for a binary catalog to cross-index.

Monday, April 1, 2013

More progress

Well, the color index computations worked well, although I do have to extend my table for interpolation of B-V a bit.  I am doing a very simple linear interpolation; I could do something fancier with a polynomial but this is really good enough.

I do have to take a quick step back - I forgot to include a sequence number (for when we hit multiple star systems) and the computed spectral type in the final output sequence.  I'll get that in tonight I expect, the re-runs will be quick.