APOD: The Flash Spectrum of the Sun (2016 Mar 12)

[APOD Robot]

The Flash Spectrum of the Sun

Explanation: In a flash, the visible spectrum of the Sun changed from absorption to emission on March 9 during the total solar eclipse. That fleeting moment, at the beginning the total eclipse phase, is captured by telephoto lens and diffraction grating in this image from clearing skies over Ternate, Indonesia. At left, the overwhelming light from the Sun is just blocked by the lunar disk. The normally dominant absorption spectrum of the solar photosphere is hidden. What remains, spread by the diffraction grating into the spectrum of colors to the right of the eclipsed Sun, are individual eclipse images. The images appear at each wavelength of light emitted by atoms along the thin visible arc of the solar chromosphere and in an enormous prominence extending beyond the Sun's upper limb. The brightest images, or strongest chromospheric emission lines, are due to Hydrogen atoms that produce the red hydrogen alpha emission at the far right and blue hydrogen beta emission to the left. In between, the bright yellow emission image is caused by atoms of Helium, an element only first discovered in the flash spectrum of the Sun.

<< Previous APOD

This Day in APOD

Next APOD >>

[/b]

[geckzilla]

I was wondering if the spikes of light to the right of the normal visible spectrum were infrared or just some internal lens reflections of some sort. I'm leaning toward reflections, but I'm not sure.

[Ann]

I was wondering if the spikes of light to the right of the normal visible spectrum were infrared or just some internal lens reflections of some sort. I'm leaning toward reflections, but I'm not sure.

There is no reason why infrared flashes would look purple in a "true-color" image like this, I think. So I believe they are reflections.

I find the APOD quite beautiful and fascinating. Imagine the Sun as an emission-line star!

I love how the red Hα and the bluish Hβ arcs both sport the same prominence at top. The helium emission arc, which is clearly orange and not yellow in color, shows the same prominence. They resemble the cartoon character Tintin!

The Balmer series.

I think there might be some smeared-out prominences in the smeared-out purple arcs, too. As the image at left from www.britannica.com shows, the emission lines of the hydrogen Balmer series are ever more closely spaced the farther you get into the purple and violet part of the spectrum. We should expect a bit of blurriness here.

But although the Sun produces a lot of green light, the green emission arcs are very faint, and lack prominences. I wonder if one of the faint green arcs closest to the Hβ arc could be OIII. If it is, then it is clear that the Sun does not produce copious amounts of OIII emission. Ah Sun, your time will come!

Ann

[geckzilla]

I was wondering if the spikes of light to the right of the normal visible spectrum were infrared or just some internal lens reflections of some sort. I'm leaning toward reflections, but I'm not sure.

There is no reason why infrared flashes would look purple in a "true-color" image like this, I think. So I believe they are reflections.

I find the APOD quite beautiful and fascinating. Imagine the Sun as an emission-line star!

What gives me pause is that they are lined up perfectly with the rest of the diffraction. Often, I notice reflections are inverted. As for the color, the camera's sensor may not be designed to pick up infrared at all. The color won't necessarily make sense. Anyway, I honestly only have about half a clue how this was made, so it's just a thought.

[Ann]

I see your point, but look at the shape of the arcs in the infrared part of the spectrum. The first one, the one closest to the red Hα arc, is very blurry, just like the violet arc farthest to the left in the visible spectrum. The next arc is a bit sharper, like the blue arc in the visible spectrum, and the third arc is quite sharp, like the Hβ arc in the visible spectrum.

I'd say these arcs to the right of the Hα arc are reflections.

Ann

[Chris Peterson]

I was wondering if the spikes of light to the right of the normal visible spectrum were infrared or just some internal lens reflections of some sort. I'm leaning toward reflections, but I'm not sure.

Diffraction gratings produce orders of spectra- that is, the spectra repeat at increasingly wide angles as the ruling spacing matches integral multiples of the wavelength of the light. The main spectrum we're seeing here is the first order spectrum (the zeroth order is the direct light of the Sun itself). I think the purple and blue bands to the right are the much dimmer second order spectrum. So, not IR and not a reflection.

[neufer]

---

[color=#00FF00][size=150]Spectral lines of helium[/size][/color]

---

<<The first evidence of helium was observed on August 18, 1868 as a bright yellow line with a wavelength of 587.56 nanometers in the spectrum of the chromosphere of the Sun. The line was detected by French astronomer Jules Janssen during a total solar eclipse in Guntur, India. This line was initially assumed to be sodium. On October 20 of the same year, English astronomer Norman Lockyer observed a yellow line in the solar spectrum, which he named the D3 Fraunhofer line because it was near the known D1 (589.59 nm)and D2 (589.00 nm) lines of sodium. He concluded that it was caused by an element in the Sun unknown on Earth. Lockyer and English chemist Edward Frankland named the element with the Greek word for the Sun, ἥλιος (helios).

In 1882, Italian physicist Luigi Palmieri detected helium on Earth, for the first time, through its D3 spectral line, when he analyzed the lava of Mount Vesuvius.

On March 26, 1895, Scottish chemist Sir William Ramsay isolated helium on Earth by treating the mineral cleveite (a variety of uraninite with at least 10% rare earth elements) with mineral acids. Ramsay was looking for argon but, after separating nitrogen and oxygen from the gas liberated by sulfuric acid, he noticed a bright yellow line that matched the D3 line observed in the spectrum of the Sun. These samples were identified as helium by Lockyer and British physicist William Crookes. It was independently isolated from cleveite in the same year by chemists Per Teodor Cleve and Abraham Langlet in Uppsala, Sweden, who collected enough of the gas to accurately determine its atomic weight. Helium was also isolated by the American geochemist William Francis Hillebrand prior to Ramsay's discovery when he noticed unusual spectral lines while testing a sample of the mineral uraninite. Hillebrand, however, attributed the lines to nitrogen. His letter of congratulations to Ramsay offers an interesting case of discovery and near-discovery in science.>>

[alter-ego]

I was wondering if the spikes of light to the right of the normal visible spectrum were infrared or just some internal lens reflections of some sort. I'm leaning toward reflections, but I'm not sure.

Diffraction gratings produce orders of spectra- that is, the spectra repeat at increasingly wide angles as the ruling spacing matches integral multiples of the wavelength of the light. The main spectrum we're seeing here is the first order spectrum (the zeroth order is the direct light of the Sun itself). I think the purple and blue bands to the right are the much dimmer second order spectrum. So, not IR and not a reflection.

My first thought was the 2nd order too, but the 2nd set does not satisfy the dispersion of a 2nd (or higher) order. The angular separations between the lines should be twice that of the 1st order. In fact, the angular separations are equal to those in the first set so that would most likely make the 2nd set a duplicate artifact of the first set, e.g. by reflection. Therefore the 2nd set is not a higher spectral order (and I agree, not IR either).

[Chris Peterson]

I was wondering if the spikes of light to the right of the normal visible spectrum were infrared or just some internal lens reflections of some sort. I'm leaning toward reflections, but I'm not sure.

Diffraction gratings produce orders of spectra- that is, the spectra repeat at increasingly wide angles as the ruling spacing matches integral multiples of the wavelength of the light. The main spectrum we're seeing here is the first order spectrum (the zeroth order is the direct light of the Sun itself). I think the purple and blue bands to the right are the much dimmer second order spectrum. So, not IR and not a reflection.

My first thought was the 2nd order too, but the 2nd set does not satisfy the dispersion of a 2nd (or higher) order. The angular separations between the lines should be twice that of the 1st order. In fact, the angular separations are equal to those in the first set so that would most likely make the 2nd set a duplicate artifact of the first set, e.g. by reflection. Therefore the 2nd set is not a higher spectral order (and I agree, not IR either).

It looks to me like the spectrum I'm identifying as second-order does, in fact, have the doubled angular separations. (Secondary on the top, aligned with primary on bottom.)

[alter-ego]

Diffraction gratings produce orders of spectra- that is, the spectra repeat at increasingly wide angles as the ruling spacing matches integral multiples of the wavelength of the light. The main spectrum we're seeing here is the first order spectrum (the zeroth order is the direct light of the Sun itself). I think the purple and blue bands to the right are the much dimmer second order spectrum. So, not IR and not a reflection.

My first thought was the 2nd order too, but the 2nd set does not satisfy the dispersion of a 2nd (or higher) order. The angular separations between the lines should be twice that of the 1st order. In fact, the angular separations are equal to those in the first set so that would most likely make the 2nd set a duplicate artifact of the first set, e.g. by reflection. Therefore the 2nd set is not a higher spectral order (and I agree, not IR either).

It looks to me like the spectrum I'm identifying as second-order does, in fact, have the doubled angular separations. (Secondary on the top, aligned with primary on bottom.)

orders.jpg

You did a nice job at capturing the 2nd set for comparison (color, contrast). You're saying that the primary cyan-colored line corresponds to the last line visible in the 2nd set? It wasn't clear to me that was the case. I associated the last 2nd-set line with the orange primary, i.e. one-to-one mapping of primary and secondary lines. I haven't measured the line positions accurately, but your image comparison does support a 2nd-order spectra instead of an artifact. Sometimes first impressions are the right ones.

[Chris Peterson]

You did a nice job at capturing the 2nd set for comparison (color, contrast). You're saying that the primary cyan-colored line corresponds to the last line visible in the 2nd set?

That's how it appears to me.

[Ann]

Diffraction gratings produce orders of spectra- that is, the spectra repeat at increasingly wide angles as the ruling spacing matches integral multiples of the wavelength of the light. The main spectrum we're seeing here is the first order spectrum (the zeroth order is the direct light of the Sun itself). I think the purple and blue bands to the right are the much dimmer second order spectrum. So, not IR and not a reflection.

My first thought was the 2nd order too, but the 2nd set does not satisfy the dispersion of a 2nd (or higher) order. The angular separations between the lines should be twice that of the 1st order. In fact, the angular separations are equal to those in the first set so that would most likely make the 2nd set a duplicate artifact of the first set, e.g. by reflection. Therefore the 2nd set is not a higher spectral order (and I agree, not IR either).

It looks to me like the spectrum I'm identifying as second-order does, in fact, have the doubled angular separations. (Secondary on the top, aligned with primary on bottom.)

orders.jpg

Your interpretation seems absolutely right to me.

The colors of the second order spectrum do not correspond one-to-one to the colors of the primary spectrum. Nevertheless, even in the second order spectrum, there is a progression from violet to longer wavelength blue. The cyan-colored Hβ flash in the primary spectrum thus corresponds to a shorter-wave blue flash in the secondary spectrum. It would have been interesting to see what the yellow-orange helium flash would have looked like in the secondary spectrum. Perhaps it would have been green?

Ann

[Chris Peterson]

It looks to me like the spectrum I'm identifying as second-order does, in fact, have the doubled angular separations. (Secondary on the top, aligned with primary on bottom.)

Your interpretation seems absolutely right to me.

The colors of the second order spectrum do not correspond one-to-one to the colors of the primary spectrum. Nevertheless, even in the second order spectrum, there is a progression from violet to longer wavelength blue. The cyan-colored Hβ flash in the primary spectrum thus corresponds to a shorter-wave blue flash in the secondary spectrum. It would have been interesting to see what the yellow-orange helium flash would have looked like in the secondary spectrum. Perhaps it would have been green?

I think you might be misunderstanding the secondary spectrum a bit. The colors (really, wavelengths) of each band are identical, all that changes is the distance between them. So the yellow-orange helium flash will look identical in the secondary spectrum (except for being dimmer). It will just be twice as far from the leftmost violet line.

EDIT: here's a more carefully composed comparison between the first- and second-order spectra. And the spacing is quite close to two-to-one, within reasonable measurement error and considering some likely lens distortion.

[alter-ego]

EDIT: here's a more carefully composed comparison between the first- and second-order spectra. And the spacing is quite close to two-to-one, within reasonable measurement error and considering some likely lens distortion.

orders2.jpg

Yup. Last night had I looked closely at the violet lines, the higher resolution of the split in the 2nd set would have cinched the 2nd order dispersion requirement for me.

[neufer]

Yup. Last night had I looked closely at the violet lines, the higher resolution of the split in the 2nd set would have cinched the 2nd order dispersion requirement for me.

The Ca II [397,393 nm] doublet :

The Solar Eclipse 1999 From Hungary
EurAstro Team Szeged I (Manfred Rudolf, Alfred Jakoblich, Franz Michlmayer, Carmen Ortega)

---

In the image [at right] the emission lines of the flash spectrum have been straightened by "bending" the image along a curve. So the assignment of the spectral lines is considerably simplified.

Some prominent emission lines are:

• Hydrogen: Ha (657nm), H b (486 nm), Hg (434 nm), Hd (410 nm)
  He I: 668, 588, 502, 492, 447 nm
  Ba II: 650, 614, 493, 455 nm
  Na I: 589 nm
  Mg I: 553, 516-518 nm ("triplet")
  Sr II: 422, 408 nm
  Ca I: 586, 423 nm
  Ca II: 397,393 nm
  Fe I: many lines between 520-570 and 413-430 nm

[alter-ego]

Yup. Last night had I looked closely at the violet lines, the higher resolution of the split in the 2nd set would have cinched the 2nd order dispersion requirement for me.

...

The Ca II [397,393 nm] doublet :

Sure looks to be the case:

[conemmil]

I was in Tidore from where I had excellent sky conditions and managed to capture the flash spectrum and the spectrum of totality, I was astonished to see that even on the lower exposures made for the flash spectrum, the corona emitting at FeXIV was captured at the same time! I never thought that this would be possible but I guess a high activity of the sun at the moment and good seeing conditions made this possible.

Here is the image with explanation lines

Click to view full size image

The same image without the lines

Click to view full size image

Maximum eclipse spectrum

Click to view full size image

I didn't know how the solar activity would have been this time since from Gabon 2013 it should go lower now. However it seems that from the view of the corona but also the spectrograph data, it still goes quite strong up there!

NeverForgetTheShadow!
Constantine Emmanouilidi
f/b: Infection Photography
http://www.stellar-explosions.com

[Ann]

It looks to me like the spectrum I'm identifying as second-order does, in fact, have the doubled angular separations. (Secondary on the top, aligned with primary on bottom.)

Your interpretation seems absolutely right to me.

The colors of the second order spectrum do not correspond one-to-one to the colors of the primary spectrum. Nevertheless, even in the second order spectrum, there is a progression from violet to longer wavelength blue. The cyan-colored Hβ flash in the primary spectrum thus corresponds to a shorter-wave blue flash in the secondary spectrum. It would have been interesting to see what the yellow-orange helium flash would have looked like in the secondary spectrum. Perhaps it would have been green?

I think you might be misunderstanding the secondary spectrum a bit. The colors (really, wavelengths) of each band are identical, all that changes is the distance between them. So the yellow-orange helium flash will look identical in the secondary spectrum (except for being dimmer). It will just be twice as far from the leftmost violet line.

EDIT: here's a more carefully composed comparison between the first- and second-order spectra. And the spacing is quite close to two-to-one, within reasonable measurement error and considering some likely lens distortion.

orders2.jpg

I think I understood you correctly. I just thought that the cyan color of Hβ looks different than the color of the Hβ flash in the secondary spectrum.

Ann

[Chris Peterson]

I think I understood you correctly. I just thought that the cyan color of Hβ looks different than the color of the Hβ flash in the secondary spectrum.

Well, I had to change the intensity to make the secondary spectrum more visible, and that means I changed the color as well. To really see it properly we'd need a spectroscope, not a color camera. It would be difficult to perfectly balance the primary and secondary such that the colors appear identical. But I think it's close enough to show what's going on.

[Ann]

I think I understood you correctly. I just thought that the cyan color of Hβ looks different than the color of the Hβ flash in the secondary spectrum.

Well, I had to change the intensity to make the secondary spectrum more visible, and that means I changed the color as well. To really see it properly we'd need a spectroscope, not a color camera. It would be difficult to perfectly balance the primary and secondary such that the colors appear identical. But I think it's close enough to show what's going on.

I didn't look at your picture when I concluded that the secondary Hβ flash seemed to be shortwave blue instead of cyan. It looked that way in the APOD.

I agree that a spectroscope should show that the primary and secondary Hβ flashes are exactly the same wavelength.

Ann

[geckzilla]

The second order diffraction is now the only thing that makes sense to me. I wasn't familiar with it until now.

[Len]

Having just returned from Indonesia, I am late to the discussion, and it appears the questions are satisfactorily resolved.

The secondary spectrum does appear in the frame although on my monitor requires considerable stretching to make it visible.

The lens in use, even though having some 16 elements, seems to be remarkably free of spurious reflections.

The camera is unmodified so, like the majority of 'domestic' cameras has built in UV and infrared filters; in fact the infared filter also cuts out about 60% of the Ha wavelength.

Refractive optics do not handle all wavelengths equally - Even high end apochromat refractors have difficulty bring violet to an accurate focus. This lens (Tamron 150-600 zoom with 1.4X amplifier) is an achromat so the extreme blue violet and extreme red will not come to sharp focus with the visually dominant wavelengths (yellow green) making the detail in the blue violet area appear slightly out of focus. Similarly the red end is slightly soft. It does a good job but is nowhere near perfect!

There are more spectra from this eclipse on flickr if one wishes to look: https://www.flickr.com/photos/130gt/

Len.

[Len]

Correction I mentioned the wrong lens: I used a standard Olympus Zukio 50-150 Zoom at 150mm

Len.