Re: aperture or focal length most relevant to good/bad seeing ?

Expand Messages
• I often see people in astro forums confidently stating that their seeing is a certain value. But, having observed what they go through at Palomar to derive
Message 1 of 21 , Aug 20, 2013
I often see people in astro forums confidently stating that their seeing is a certain value. But, having observed what they go through at Palomar to derive seeing, I often wonder how the average amateur astronomer can be so positive of what they're getting. I often check FWHM values on my downloaded images, mainly to verify good focus, but I never see FWHM values anywhere close to 1 arc sec. As to what the Night Assistant reports at the 200-inch Hale Telescope, the 1.2 average figure I cited for the night of Aug 18-19 is pretty typical for this time of year. Sometimes it's a bit better (down to, say, 0.9), sometimes quite a bit worse, especially in spring. My site here in Anza, CA, is similar enough to Palomar that my experience of "good" vs "poor" seeing usually parallels what they get, but I can only see it after the fact ("Oh, they had a lousy night too!")

--- In ccd-newastro@yahoogroups.com, "Stan" <stan_ccd@...> wrote:
>
> --- "CurtisC" <calypte@> wrote:
> > as a practical matter, how do you measure seeing...
>
> Setting aside serious measuring machines (scintillation monitors, DIMM and such) a "practical" method is to really know your instrument and site thru experience and measurements.
>
> A key measurement is the FWHM for the better short exps on the best night. This value can be adjusted and used as the instrument's resolution limit. The instruments resolution limit can then be used to calculate the seeing on any other night by quadratically subtracting it from the FWHM of short exps.
>
> short exp FWHM = sqrt(instrument^2 + seeing^2)
> thus
> instrument = sqrt(exp^2 - seeing^2)
>
> In the absence of impendent measurements, it is necessary to make some reasonable assumptions. Seeing (measured in any way) is very rarely below 0.5" and most non-professional sites rarely experience less than 1". So it is not unreasonable to assume that a very good night has base seeing <=1"; be warmed that there are some chronically bad sites that never approach this. So use 1" as a first approximation:
>
> Instrument limit = sqrt(exp^2 - 1)
> Examples:
> if best night FWHM = 1.5" then the instrument limit approx = 1.1"
> if best night FWHM = 2.0" then the instrument limit approx = 1.7"
> if best night FWHM <= 1" then the seeing was less than 1" and the instrument is very good.
>
> Once you have some experience and confidence with the instrument then lesser seeing can be easily estimated via short exps:
>
> seeing = sqrt(fwhm^2  instrument^2)
>
> examples:
>
> if instrument = 1" and short exp FWHM = 2.0 then seeing = 1.7"
> if instrument = 1" and short exp FWHM = 2.5 then seeing = 2.3"
> if instrument = 1" and short exp FWHM = 3.0 then seeing = 2.8"
>
> if instrument = 2" and short exp FWHM = 2.5 then seeing = 1.5"
> if instrument = 2" and short exp FWHM = 3.0 then seeing = 2.3"
> if instrument = 2" and short exp FWHM = 3.5 then seeing = 2.9"
>
> Estimation error increases as (exp - instrument) decreases.
>
> Stan
>
• ... Very few amateurs do! Amateur DS FWHM = 1.0 was a holy grail for many years. I know of only a mere handful of credible achievements, most of them
Message 2 of 21 , Aug 21, 2013
--- "CurtisC" <calypte@...> wrote:
> I often see people in astro forums confidently stating that
> their seeing is a certain value...
> but I never see FWHM values anywhere close to 1 arc sec.

Very few amateurs do! <g>

Amateur DS FWHM = 1.0" was a holy grail for many years. I know of only a mere handful of credible achievements, most of them stretching the definition of "amateur" or "deep space" (e.g. short exps from professional grade meter class RC on mountain top observatory).

Benoit Schillings and I combined our knowledge, skills and equipment to struggle for this goal over a very long time and finally broke thru that barrier with room to spare. Using Benoit's home-built 20" Newtonian (operating at f/20) and my home-assembled photon intensified camera at a good seeing camp site, we achieved FWHM near 0.5" for the core of M13:

http://www.stanmooreastro.com/IntensifiedAstronomicalImaging.htm

Since then I have been able to get near 1.0" on a few more occasions using C-14 with DS lucky imaging. (I need to update my web site with newer pics). So I have a pretty good idea of what it takes. It is not easy...

Stan
• I d like to try this out, Stan. But I have two questions: 1) Your second equation seems to mix units of seconds (time) with arc-seconds. The usual sidereal
Message 3 of 21 , Aug 21, 2013
I'd like to try this out, Stan. But I have two questions:

1) Your second equation seems to mix units of seconds (time) with arc-seconds. The usual "sidereal" conversion factor doesn't seem to make sense here - what's the deal?
2) How long is a "short" exposure - something like 1 second?
3) Would it make sense to take, say, twenty 1 second exposures and use an average?

Thanks.

Bruce

--- In ccd-newastro@yahoogroups.com, "Stan" <stan_ccd@...> wrote:
>
> --- "CurtisC" <calypte@> wrote:
> > as a practical matter, how do you measure seeing...
>
> Setting aside serious measuring machines (scintillation monitors, DIMM and such) a "practical" method is to really know your instrument and site thru experience and measurements.
>
> A key measurement is the FWHM for the better short exps on the best night. This value can be adjusted and used as the instrument's resolution limit. The instruments resolution limit can then be used to calculate the seeing on any other night by quadratically subtracting it from the FWHM of short exps.
>
> short exp FWHM = sqrt(instrument^2 + seeing^2)
> thus
> instrument = sqrt(exp^2 - seeing^2)
>
> In the absence of impendent measurements, it is necessary to make some reasonable assumptions. Seeing (measured in any way) is very rarely below 0.5" and most non-professional sites rarely experience less than 1". So it is not unreasonable to assume that a very good night has base seeing <=1"; be warmed that there are some chronically bad sites that never approach this. So use 1" as a first approximation:
>
> Instrument limit = sqrt(exp^2 - 1)
> Examples:
> if best night FWHM = 1.5" then the instrument limit approx = 1.1"
> if best night FWHM = 2.0" then the instrument limit approx = 1.7"
> if best night FWHM <= 1" then the seeing was less than 1" and the instrument is very good.
>
> Once you have some experience and confidence with the instrument then lesser seeing can be easily estimated via short exps:
>
> seeing = sqrt(fwhm^2  instrument^2)
>
> examples:
>
> if instrument = 1" and short exp FWHM = 2.0 then seeing = 1.7"
> if instrument = 1" and short exp FWHM = 2.5 then seeing = 2.3"
> if instrument = 1" and short exp FWHM = 3.0 then seeing = 2.8"
>
> if instrument = 2" and short exp FWHM = 2.5 then seeing = 1.5"
> if instrument = 2" and short exp FWHM = 3.0 then seeing = 2.3"
> if instrument = 2" and short exp FWHM = 3.5 then seeing = 2.9"
>
> Estimation error increases as (exp - instrument) decreases.
>
> Stan
>
• I ll throw-in my 2-cents worth. I m not an expert in seeing, and my method does not give a definite seeing value. It s also affected by mount alignment, worm
Message 4 of 21 , Aug 22, 2013
I'll throw-in my 2-cents worth. I'm not an expert in seeing, and my method does not give a definite seeing value. It's also affected by mount alignment, worm accuracy/PE. But since I align my mount to the same accuracy for each use (<1 arc-min), and my Mach1 is a fairly accurate mount, I figure that it's my night-to-night comparisons should be fairly accurate.

When I guide, using Maxim, I see the macro effects of seeing, by looking at the guider corrections on the scope. I can't really say with certainty that my seeing is X, but I can see definite differences between good nights and poor nights. I approximate the seeing by the variation in my guiding graph. What I concern myself with is that my "guider seeing" is better than what I need for my scope/camera combo.

Since I shoot with an FSQ-106ED and an FLI11K CCD, my image scale is 3.5 arc-sec per pixel. So I figure that as long as my guiding chart indicates less than +/- 2 px in guiding, I should be near 1 pixel in image scale resolution. I also save each night's guiding, so if I see problems in post-processing, I can review the graph to see if there was any anomalies.

With this technique, my "guiding seeing" is usually around 1-2 arc-sec. On the rare evening, the atmosphere calms to where my guiding graph is less than +/- 1/2 Arc-sec.

Maybe a rough way to look at it, but at least I have fairly repeatable way to see the night-to-night variations in atmospherics.

Again, a rough technique, but one that seems to fit my wide-field imaging requirements.

Eric

--- In ccd-newastro@yahoogroups.com, "CurtisC" <calypte@...> wrote:
>
> Dumb question for either you, Ron, or Stan, or maybe both: as a practical matter, how do you measure seeing so as to arrive at the 1 arc sec value or whatever?
>
> --- In ccd-newastro@yahoogroups.com, Ron Wodaski <yahoo@> wrote:
> >
> > Having worked with 16", 20", and 40" telescope at this point, I would describe seeing not as a problem or a difficulty, but as appropriate or not.
> >
> > For example, if I want fine detail, then I'm going to wait until I have really good seeing. If I want to record an extremely dim object, I also need best possible seeing.
> >
> > But even with modest seeing, I can take satisfying images of many targets.
> >
> > You have to take the seeing you get. Everyone can decide for themselves when the seeing is "too bad for imaging."
> >
> > For example, I get really excited when the seeing drops below 1", because the one-meter telescope can take advantage of that.
> >
> > So "bad seeing" on that scope is anything over about an arcsecond. But if we didn't use the scope when the seeing was 1.2" or 1.5" or even 2", we wouldn't use it very much. :-)
> >
> > So I get excited when the seeing approaches the capability of a given telescope, but that's not the only time I use it!
> >
> > As I will be describing in a talk at the Australian AIC next week, there are also things you can do to control the net seeing effects at the focal plane. (Will also come out in book form later this year.)
> >
> > So that's the way I see it: just part of the job. You do what you can in terms of locating your observatory or observing location; you build your observatory to minimize seeing effects, you design (or modify) the telescope to minimize its contribution to seeing, you deal with temperature differences in the observatory, telescope, and/or optics, and so on.
> >
> > In my experience, you can gain from 0.5-1.5" of improved seeing by knowing what to do with the above variables. (More, if you can get a site at 10,000 feet, of course!) But at a given location, with the right skills, you can definitely improve your best FWHM.
> >
> > nature's part in the seeing can't be modified, but there are seeing forecasts to help you plan. :-)
> >
> > So a bigger aperture may not give you a smaller FWHM if your site can't deliver one. But it is a bigger aperture, so even if you can't get much additional detail, you are getting more light, and you can image more efficiently at the image scale that your site and equipment allow.
> >
> >
> >
> >
> > On Aug 16, 2013, at 9:56 AM, lmbuck2000 <lmbuck2000@> wrote:
> >
> > >
> > >
> > > thanks, stan. appreciate the explanation. would be interested to hear from some users of 16-20inch aperture scopes (@f/7-f/9) how much difficulty they have under "typical" seeing conditions.
> > >
> > > Lee
> > >
> > > --- In ccd-newastro@yahoogroups.com, "Stan" <stan_ccd@> wrote:
> > > >
> > >
> > >
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
>
• ... in the eq, exp is meant to refer to the FWHM of the exp ... The full answer would be a long discourse because seeing is a time based dynamic. Seeing
Message 5 of 21 , Aug 22, 2013
> 1) Your second equation seems to mix units of seconds (time)

in the eq, "exp" is meant to refer to the FWHM of the exp

> 2) How long is a "short" exposure - something like 1 second?

The full answer would be a long discourse because seeing is a time based dynamic. "Seeing wander" is often a significant component of image FWHM but it is not measured by instruments like DIMM because such devices normalize wander and only measure relative wavefront distortions. Also, wander is a function of aperture (larger apertures produce less wander) so including it in the measurement produces an aperture dependent seeing value.

To derive the seeing independently of wander via a normal scope/exp requires very short exps, such as 10-20ms. But such short exps produce varying, irregular non-Gaussian PSF that yield unreliable FWHM. A stream of a few hundred 20ms exps can be aligned and stacked to produce a virtually wanderless FWHM and I do that frequently with intensified video. But that's not practical for slow CCD and not even possible for charge transfer CCDs (shortest exp time is >100ms for many such cameras).

So 1-5 seconds is usually enough time to get a near Gaussian PSF and short enough to escape most mount issues (except vibrations), especially if you take a dozen or so exps. But it does include wander and as such may over-estimate the core seeing. However, it should be fairly consistent for that scope and can be a useful measurement.

> 3) Would it make sense to take, say, twenty 1 second
> exposures and use an average?

Yes.

Stan
• ... That is an easy and useful method of estimating conditions relative to particular equipment and practices. However, as you note it does not produce an
Message 6 of 21 , Aug 23, 2013
--- "echesak@..." <echesak@...> wrote:
> I see the macro effects of seeing, by looking at the guider
> corrections on the scope...

That is an easy and useful method of estimating conditions relative to particular equipment and practices. However, as you note it does not produce an accurate seeing FWHM that can be reliably compared with other systems and measurements.

Guide exps are usually several seconds long in order to avoid having the mount chase the seeing. Longish guide exps blur most of the seeing effects to distill pointing errors. So bad seeing usually does not produce large guide errors since this method deliberately minimizes seeing effects. Guide errors can have a seeing component but are more often due to factors such as mount, vibrations, OTA response to wind, etc.

AO guiding is another matter. AO guide exps are usually near 100ms. So most of the report error RMS is due to seeing wander. AO can remove a substantial amount of coherent seeing wander.

Stan
• ... (continued) Image wander is produced by refractions from atmospheric seeing cells and these cells have a large range of sizes and speeds. The power
Message 7 of 21 , Aug 23, 2013
--- "echesak@" <echesak@> wrote:
> I see the macro effects of seeing, by looking at the guider
> corrections on the scope...

(continued)

Image wander is produced by refractions from atmospheric seeing "cells" and these cells have a large range of sizes and speeds. The power spectrum of seeing effects occur over a broad range of time scales. Most of the disturbances are fast and will not be detected by guide errors (they simply blur into each other in a single exp) but the large cell effects have low frequencies (>=1hz) and this low end of the power spectrum can show up as a guiding error. But this is undesirable because it causes the mount to chase an imaginary deviation so if you experience guide error attributable to seeing then you should increase the guide exp time and/or lower the guider aggression.

This slow seeing wander can be attenuated with AO.

Stan
Your message has been successfully submitted and would be delivered to recipients shortly.