|
|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
Converting RA/Dec to earth centered coordinates?
In article ,
"W. eWatson" writes: My xy plane is in the plane of the equator, and its projection into the sky represents the celestial equator. Declination is measured +/- from the celestial equator to each pole along great circle lines that pass through each pole. Somewhere on the celestial equator is point from where RA is measured. To go from RA/Dec to an x,y,z unit vector is just simple trignometry: x = cos(dec)*cos(RA) y = cos(dec)*sin(RA) z = sin (dec) Make sure to convert RA/dec to degrees or radians (whatever units your calculator or program uses). It's easy to forget to multiply hours by 15 to get degrees. The zero point of RA is the place where the ecliptic and celestial equator intersect with the equinox heading north. Both this zero point and the location of the celestial poles changes with respect to the stars because of precession. If you want a _current_ x,y,z unit vector, you need to start from current RA/Dec coordinates rather than coordinates at a standard "ecliptic and equinox" (B1950 or J2000). The Meeus book will tell you how to do that calculation. Ah, the obliquity of the ecliptic (e) is what I need. Don't see why you need that. Did you mean ecliptic coordinates rather than celestial? -- Help keep our newsgroup healthy; please don't feed the trolls. Steve Willner Phone 617-495-7123 Cambridge, MA 02138 USA |
#2
|
|||
|
|||
Converting RA/Dec to earth centered coordinates? (Correction)
On 5/2/2011 2:28 PM, Steve Willner wrote:
In , "W. writes: My xy plane is in the plane of the equator, and its projection into the sky represents the celestial equator. Declination is measured +/- from the celestial equator to each pole along great circle lines that pass through each pole. Somewhere on the celestial equator is point from where RA is measured. To go from RA/Dec to an x,y,z unit vector is just simple trignometry: x = cos(dec)*cos(RA) y = cos(dec)*sin(RA) z = sin (dec) Make sure to convert RA/dec to degrees or radians (whatever units your calculator or program uses). It's easy to forget to multiply hours by 15 to get degrees. The zero point of RA is the place where the ecliptic and celestial equator intersect with the equinox heading north. Both this zero point and the location of the celestial poles changes with respect to the stars because of precession. If you want a _current_ x,y,z unit vector, you need to start from current RA/Dec coordinates rather than coordinates at a standard "ecliptic and equinox" (B1950 or J2000). The Meeus book will tell you how to do that calculation. Ah, the obliquity of the ecliptic (e) is what I need. Don't see why you need that. Did you mean ecliptic coordinates rather than celestial? Whoops, I posed the question backwards. I have the x,y,z coordinates of a vector and I want to convert them to ra/dec. In a unit sphere I think of z as pointing through the north pole, x pointing south through 1,0,0, and y pointing east through 0,1,0. In my case, precession does not enter into matters. I'm constructing a simulation that is mostly grounded in az/el and lat/lng. I wrote a program that produces the path of a fake meteor moving in a straight line. Time is not yet useful as a consideration yet. The direction of the line points to the radiant point in the sky. Meteors lie on a great circle, hence pass their plane passes through the earth's center (spherical earth). My program has not yet needed ra/dec, which is usually the measure of the radiant point in the sky and is given in ra/dec. However,I have the data from a similar program, and they provide the radiant as ra/dec. internally, my program seems sound when I sort of run it backwards. I get agreeable results. I want to see if the independent source and I agree. |
#3
|
|||
|
|||
Converting RA/Dec to earth centered coordinates? (Correction)
In article ,
"W. eWatson" writes: Whoops, I posed the question backwards. I have the x,y,z coordinates of a vector and I want to convert them to ra/dec. So just invert the equations I gave you. Dec = arcsin(z) RA = atan2(x,y) You might have to use (y,x) in the second one, depending on your atan2 function. Don't forget to convert to degrees/hours from whatever units your calculator or program uses. In a unit sphere I think of z as pointing through the north pole, x pointing south through 1,0,0, and y pointing east through 0,1,0. Something is confused here. If +z is north, south will be -z, not x. Anyway, check a few cases that you know the answer for and make sure there's no sign inversion or 90-degree error. -- Help keep our newsgroup healthy; please don't feed the trolls. Steve Willner Phone 617-495-7123 Cambridge, MA 02138 USA |
#4
|
|||
|
|||
Converting RA/Dec to earth centered coordinates? (Correction)
Steve Willner wrote:
In article , "W. eWatson" writes: Whoops, I posed the question backwards. I have the x,y,z coordinates of a vector and I want to convert them to ra/dec. So just invert the equations I gave you. Dec = arcsin(z) Better: Dec = atan2 (z, sqrt(x^2 + y^2)), which will work for all vectors (not just unit vectors) and will give accurate results near the poles. RA = atan2(x,y) You might have to use (y,x) in the second one, depending on your atan2 function. Don't forget to convert to degrees/hours from whatever units your calculator or program uses. It's universally atan2(y,x) as far as I know. You do have to make sure you put the y coordinate in the numerator. In a unit sphere I think of z as pointing through the north pole, x pointing south through 1,0,0, and y pointing east through 0,1,0. Something is confused here. If +z is north, south will be -z, not x. Anyway, check a few cases that you know the answer for and make sure there's no sign inversion or 90-degree error. |
#5
|
|||
|
|||
Converting RA/Dec to earth centered coordinates? (Correction)
On May 6, 1:34*am, Bill Owen wrote:
Steve Willner wrote: In article , *"W. eWatson" writes: Whoops, I posed the question backwards. I have the x,y,z coordinates of a vector and I want to convert them to ra/dec. So just invert the equations I gave you. Dec = arcsin(z) Better: Dec = atan2 (z, sqrt(x^2 + y^2)), which will work for all vectors (not just unit vectors) and will give accurate results near the poles. RA = atan2(x,y) You might have to use (y,x) in the second one, depending on your atan2 function. *Don't forget to convert to degrees/hours from whatever units your calculator or program uses. It's universally atan2(y,x) as far as I know. *You do have to make sure you put the y coordinate in the numerator. In a unit sphere I think of z as pointing through the north pole, x pointing south through 1,0,0, and y pointing east through 0,1,0. Something is confused here. *If +z is north, south will be -z, not x. Anyway, check a few cases that you know the answer for and make sure there's no sign inversion or 90-degree error. All the doctorates here with not enough sense to look at a telescope tracking a circumpolar star at all times during a year where the turning telescope creates its own axis rather than following the attributes of planetary geometry,in short,you are not dealing with the convenience of calendar based geocentric coordinates but pure homocentricity. Let me spell it out for you and for the people who pay your salary,Ra/ Dec is not a geocentric system hence it does not translate into planetary dynamics directly even though you and your colleagues have this idea that the daily and orbital motion of the Earth can be squeezed into right ascension.This requires something even more plain for the reader,when you assume there is an imbalance between the amount of daylight/darkness cycles over a 4 year orbital period,and this is what Ra/Dec applied to the Earth's planetary dynamics does,you are in trouble with the most basic cause and effect of all - the correspondence between one 24 hour rotation and one day/night cycle. Don't know how any of you can do it,no matter how hard you try to skip over or ignore what is a fundamental astronomical cause and effect or treat it like trivia,the more your empirical agenda looks inconsequential and while the calendar based Ra/Dec is a convenience it seems you just don't have the ability to interpret that any telescope turning on its mount gives a homocentric picture and not a geocentric one. |
#6
|
|||
|
|||
Converting RA/Dec to earth centered coordinates? (Correction)
On May 5, 8:59*pm, oriel36 wrote:
All the doctorates here with not enough sense to look at a telescope tracking a circumpolar star at all times during a year where the turning telescope creates its own axis... ...it seems you just don't have the ability to interpret that any telescope turning on its mount gives a homocentric picture and not a geocentric one. Set that telescope on a star tonight, any star at all, and turn the drive off, no tracking needed. About 23:56:04 will pass before the same star is centered up again in the telescope, and it will continue to do so, every night, night after night, in the exact same time frame. What could this mean? Clearly, with respect to that star, rotation takes about 23:56:04... what other possible conclusion could a rational person come to? Where does your own superior ability to interpret lead you, given this incontrovertible, easily observable fact? |
#7
|
|||
|
|||
Converting RA/Dec to earth centered coordinates? (Correction)
On May 5, 3:43*pm, (Steve Willner) wrote:
In article , *"W. eWatson" writes: Whoops, I posed the question backwards. I have the x,y,z coordinates of a vector and I want to convert them to ra/dec. So just invert the equations I gave you. Dec = arcsin(z) RA = atan2(x,y) You might have to use (y,x) in the second one, depending on your atan2 function. *Don't forget to convert to degrees/hours from whatever units your calculator or program uses. If you have the ATAN2 function, that would mean you are using FORTRAN, which uses radians as its units... I would have thought. However, in C, math.h also includes an atan2 function. The C atan2 function is described as being atan2(y,x) and returning the arctangent of y/x, set in the appropriate quadrant. Thus, given the signs of y and x, the result is in the range: y - positive, x - positive: [0, pi/2] y - positive, x - negative: [pi/2, pi] y - negative, x - positive: [-pi/2, 0] y - negative, x - negative: [-pi, -pi/2] Thus, positive angles are measured counterclockwise from the positive x-axis where x is the second argument. This is the same convention as FORTRAN used. John Savard |
#8
|
|||
|
|||
Converting RA/Dec to earth centered coordinates? (Correction)
On May 6, 6:52*am, palsing wrote:
On May 5, 8:59*pm, oriel36 wrote: All the doctorates here with not enough sense to look at a telescope tracking a circumpolar star at all times during a year where the turning telescope creates its own axis... ...it seems you just don't have the ability to interpret that any telescope turning on its mount gives a homocentric picture and not a geocentric one. Set that telescope on a star tonight, any star at all, and turn the drive off, no tracking needed. About 23:56:04 will pass before the same star is centered up again in the telescope, and it will continue to do so, every night, night after night, in the exact same time frame. What could this mean? Clearly, with respect to that star, rotation takes about 23:56:04... what other possible conclusion could a rational person come to? Any person here who concludes there are 366 1/4 rotations in a year through this late 17 the century junk is either a complete fraud or criminally incompetent,it is not an accusation,it is not an opinion but something which far more dangerous than any act of terrorism as it undermines the actual correspondence between what the body experiences and the mind accepts as valid.You have these guys from NASA and Harvard with an ideology that sinks below a standard that the world hasn't seen before and while you may try to reason your way to justify stellar circumpolar motion using the Earth's planetary dynamics,it cannot be done . Never.never has the world had to deal with these creatures who willingly try to avoid a open debate and even when it is pointed out that their Ra/Dec reckoning is homocentric,not geocentric,they still continue to push for something which is contrary to all known astronomical principles ,either geocentric or the astronomy of planetary dynamics. A telescope tracking a star creates its own axis of rotation ,a sane person interpreting that individual rotation will ultimately conclude that Ra/Dec is based on the calendar system,no more or less.What a horrible bunch of people that go out of their way to propose a system which does not tally with cause and effect of the daylight/darkness cycle and can call themselves astronomers despite the inability to comprehend the proportion of rotations to orbital cycles as 1461 to 4 or 365 1/4 rotations to 1 orbital cycle. You think you are better than a flat Earther or a creationist by arguing against the rotation of the planet once in 24 hours and 365 1/4 times in an orbital year but you are far,far below what those people accept as fact,the difference is that the toxic strain of empiricism which ran with Ra/Dec reasoning is a dominant force in the education system via the 'scientific method'. You have moved on,let the other readers who stuck their necks out continue. Where does your own superior ability to interpret lead you, given this incontrovertible, easily observable fact? |
#9
|
|||
|
|||
Converting RA/Dec to earth centered coordinates? (Correction)
On May 6, 7:23*am, oriel36 wrote:
You think you are better than a flat Earther or a creationist by arguing against the rotation of the planet once in 24 hours and 365 1/4 times in an orbital year... I'm not arguing against anything here, in fact, I agree with you, with respect to the sun, and so does everyone else. I'm only asking for your own interpretation of the fact that an undriven telescope centered on any star of your choice will have that same star momentarily centered in the field of view every 23:56:04... an undeniable fact that is readily observable by anyone, anytime, anywhere. It doesn't have a thing to do with any calendar, but it does mean something, and I'm just asking you for your interpretation. Simple. |
#10
|
|||
|
|||
Converting RA/Dec to earth centered coordinates? (Correction)
On May 6, 5:13*pm, palsing wrote:
On May 6, 7:23*am, oriel36 wrote: You think you are better than a flat Earther or a creationist by arguing against the rotation of the planet once in 24 hours and 365 1/4 times in an orbital year... I'm not arguing against anything here, in fact, I agree with you, with respect to the sun, and so does everyone else. Empiricists,a spectacularly intellectually impotent bunch,are arguing for 366 1/4 rotations in an orbital year hence the attempt to create an imbalance between rotations and the daylight/darkness cycles so this is an issue of cause and effect experienced by the body,assign any other value than 1461 rotations and day/nights cycles across the calendar,and this is what criminally incompetent individuals try to do and the things descends to a level worse than any the world has ever known as ultimately all it requires is the ability to count. Go teach the NASA,Harvard guys,people who can't distinguish geocentric from homocentric by interpreting the tracking of a stellar circumpolar star using a GoTo telescope deserve the same ridicule as any group that tries to make something out of nothing.Go ahead and explain why Feb 29th as a 24 hour rotation and day/night cycle is required to complete the proportion of 1461 rotations to 4 orbital circuits.It is not the these intellectual freaks are repentant,much like the old Nazi were,it is the lack of support for the rotation of the Earth once in 24 hours and 365 1/4 times in an orbital year. Fundamentalists have nothing on empiricists,that much is for certain and while some like yourself can be excused as nuisances, the worst behavior has always been self-serving silence. The simple fact that the 1461 rotations spanning 4 orbital circuits accounts for the day/night cycle of Feb 29th would simply transfer to the basic proportions of 365 1/4 rotations per circuit so yes,you are arguing against something so fundamental to human understanding that I'm only asking for your own interpretation of the fact that an undriven telescope centered on any star of your choice will have that same star momentarily centered in the field of view every 23:56:04... an undeniable fact that is readily observable by anyone, anytime, anywhere. It doesn't have a thing to do with any calendar, but it does mean something, and I'm just asking you for your interpretation. Simple. |
|
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Converting RA/Dec to earth centered coordinates? | William Hamblen[_2_] | Astronomy Misc | 1 | May 3rd 11 02:49 AM |
Converting star coordinates to x,y,z | [email protected] | UK Astronomy | 4 | December 11th 04 11:28 PM |
converting star coordinates to x,y,z | [email protected] | Amateur Astronomy | 28 | December 10th 04 06:45 PM |
converting star coordinates to x,y,z | [email protected] | Astronomy Misc | 3 | December 9th 04 08:34 PM |
converting coordinates | J. Jason Fry | Amateur Astronomy | 8 | May 31st 04 06:27 PM |