ShotgunNinja
commented
7 years ago Note on solar wind occlusion by celestial bodies It will be very hard to determine occlusion in the proper way, because the 'parker-spiral-arm' path would have to be walked, and for each step all celestial bodies position at the time would have to be calculated from their orbits, then sun visibility evaluated. This could probably be also done analytically, but it will be even more ugly.
A better solution is to ignore celestial bodies occlusion when evaluating the solar wind intensity at a point, and instead only using it when evaluating the radiation reaching a particular vessel. That is, ignore the real solar wind path and assume a line-of-sight one instead. Some corner cases may be built for which this result will not match the proper one, but in general the two results should be the same.
YaarPodshipnik
PiezPiedPy
DaveTSG
lordcirth
This outline a proposed re-implementation of space weather mechanics, based on a simple idea. That idea came out of private discussions with David Hunter, that is knowledgeable on the subject. Thank you David!
At the moment, space weather is only represented by symbolic 'CME events' originating at random from the sun, toward a planetary system. After the CME symbolically hit the planetary system, the whole of it enter what I define as a 'solar storm' state, for some time. In that state, radiation is increased (if you are not inside a magnetopause), and communications are blacked out (if you are inside a magnetopause). This system is okay (if not a bit bland). There is no graphical representation of the CME.
The proposal is to scrap the existing system, and replace it with a new one.
Space weather primer The sun emit charged particles in all directions, usually called 'the solar wind'. The emission is not uniform over the photosphere, nor it is constant over time. At some point in time, areas of the sun photosphere emitting less or more than average are called 'dark spots' or 'bright spots'. These cause 'discernible events' both on the photosphere, above the photosphere, and in the path that the related emissions follow through the solar system. These 'events' have many names, that are beyond the scope of this gentle introduction.
A 11 years cycle has been observed on the level of charged particles emitted from the sun. Therefore, the frequency of dark and bright spots change over the course of this 11 years cycle. Intuitively, dark spots are associated to lower emission levels while bright spots are associated with higher emission levels. Cycles start with low emission level, high number of dark spots and low number of bright spots (the solar minimum), reach an apex of high emission level, high number of bright spots and low number of dark spots (the solar maximum), then return to the starting condition.
The charged particles are emitted radially, and as the sun photosphere rotate the stream emitted from a particular spot will form a spiral as they travel outward from the sun. This is called the 'Parker Spiral'.
The charged particles are emitted with variable velocities. The set of velocities observed seem to cluster around two particular ones. These have been called the 'fast solar wind' and the 'slow solar wind'.
All of this is a simplication. The reasons for the sun cycle and the anisotropy of emission intensity and velocity are irrelevant for our purpose.
The idea For any point in space (in the sun frame of reference), the associated 'parker spiral arm' passing through the point is deduced. Then, the origin of the 'parker spiral arm' is traced back to a specific point on the surface of the sun, at a specific point in the past.
The surface point and the moment in time are then used as coordinates to lookup in a map (in the mathematical sense), that associate all possible ones to a level of charged particle emission.
This map is implemented procedurally, in such a way as to model the dark and bright spots, with a frequency that follow a single cycle. A simple modulo operation on the time coordinate cause the cycle to repeat infinitely. The procedural representation requires no data storage requirements and has infinite resolution.
What gives It is possible to evaluate the solar wind intensity at any point, and these intensities vary with time and are 'parker-spiral-like'. This space weather model is a decent representation of reality, and is fast to evaluate. Now the question arise what to do with this data.
Intuitively, it could map to radiation directly so that increasingly intense emissions cause increasingly intense radiation. What is now the CME 'solar storm' will then be simply a particularly intense emission.
Maybe they should also interfere with communications, in a more generic way than the current 'blackout'. Albeit I didn't explore this aspect at all, at this moment.
It would be nice to add a new experiment part, that act as sensor for solar wind intensity data (like SWEPAM). But this require either me learning how to model and texture in Blender, or somebody else contributing the model.
I'm out of ideas on what else the solar wind intensity could be used for. It is interesting to note that this data will be exposed by the API, so it could be used by other mods (one that add electric sails for example).
Graphical representation of solar wind A lot of particles are rendered (eg: 100k), centered on the sun, using an uniform distribution. The particles are static, that is, they don't move in the sun frame of reference. They are also sparse enough so that the overall 3d field can be seen.
The color of each particle will be determined from the solar wind intensity at that point, at current time. This will happen by re-implementing the algorithm mentioned above inside a shader.
To aid in visualizing the 3d field, depth of each particle will have to be conveyed somehow.
This is only rendered in map-view or tracking-station.
Effect on performance should be trivial.
Graphical representation of sun photosphere and sun corona events Ideally the bright and dark spots should be represented on the photosphere, albeit this is secondary. The simplest way would be to replace the existing sun shader with a custom one, that sample the solar wind 'map' at current time. But this could conflict with custom planet packs and/or light flare effects, logos forbid.
Even more secondary is representing the plasma filaments along very dynamic magnetic field, occurring over bright spots usually, and with intensity proportional to the underlying emission intensity. I have some ideas on this, that involve an ad-hoc physical simulator using verlet integration (nowadays called 'position based dynamics') and using spline attractors for the magnetic lines. So assume this will not happen.