plate solve precision?

[Mike Dubson]

Hi, I'm a new user with a couple of questions regarding plate solution.
Question 1: After I plate solve an image, how can I see the plate scale, that is, the arc-seconds per pixel?  I see it displayed approximately right on the image (there is a scale bar marked in arcsec or arcmin and I can count pixels), but I want to know the plate scale as precisely as possible.
Question 2: After I plate solve an image, I would like to have some information about the residuals of the solution, namely, how accurate is the solution?  I am doing astrometry on near-earth asteroids, and I would like some estimate of the uncertainty in the position determinations. I don't mean uncertainty in the centroid of the asteroid image due to photon count statistics, I mean the precision of the WCS plate solve itself. -thanks!

[karenacollins]

Hi Mike,

Go to Image_Menus->WCS->"Set pixel scale for images without WCS" to see the pixel scale. The following panel will open. When images have a WCS solution, the numbers shown represent the pixel scale from the WCS solution. I need to update the menu wording to more properly represent the current functionality when an image has WCS headers.

Once the panel opens, you can read off the numbers and then click okay without making any changes.


The plate-solve WCS solution is determined by the astrometry.net website and is not really a function of AIJ. Once the WCS solution is found by astrometry.net, the WCS headers are returned to AIJ and they are simply copied to the FITS header of the image. The Lang et al. 2010 astrometry.net journal paper (https://ui.adsabs.harvard.edu/abs/2010AJ....139.1782L/abstract) might give info on astrometric uncertainty, but I expect it depends on the number of stars available as the the basis for the plate solution, etc.

There is a component of astrometric uncertainty that is affected by AIJ, and that is due to the centroiding function used to find the X,Y coordinates of stars in the image. The AIJ derived X,Y coordinates are sent to astrometry.net and the X,Y positions are the basis of the astrometry.net plate solution. However, I think the uncertainty in the centroids would be hard to quantify and would depend on the shpes and sizes of the PSFs of the stars in the image. For instance, a sharply focused image should have much lower uncertainty in the position of the X,Y coordinates than an image that is not sharply focused, and especially if the defocused PSF is not symmetric. I would expect that images with a focus differential across the image, or images with a background gradient would cause higher uncertainties on the centroiding precision/uncertainty.

Sorry I can't be more specific on the astrometric uncertainty, but maybe the Lang et al. Astrometry.net paper will help.

Karen

On 6/16/2019 4:47 PM, Mike Dubson [via AstroImageJ] wrote:

Hi, I'm a new user with a couple of questions regarding plate solution.
Question 1: After I plate solve an image, how can I see the plate scale, that is, the arc-seconds per pixel?  I see it displayed approximately right on the image (there is a scale bar marked in arcsec or arcmin and I can count pixels), but I want to know the plate scale as precisely as possible.
Question 2: After I plate solve an image, I would like to have some information about the residuals of the solution, namely, how accurate is the solution?  I am doing astrometry on near-earth asteroids, and I would like some estimate of the uncertainty in the position determinations. I don't mean uncertainty in the centroid of the asteroid image due to photon count statistics, I mean the precision of the WCS plate solve itself. -thanks!

---

If you reply to this email, your message will be added to the discussion below:

http://astroimagej.1065399.n5.nabble.com/plate-solve-precision-tp1160.html

To start a new topic under AstroImageJ, email [hidden email]
To unsubscribe from AstroImageJ, click here.
NAML