the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 09 Aug 2023
Explaining the green volcanic sunsets after the 1883 eruption of Krakatoa
Christian von Savigny
Anna Lange
Christoph Hoffmann
Alexei Rozanov
Abstract. Volcanic sunsets are usually associated with extended and enhanced reddish colours typically complemented by purple colours at higher elevations. However, many eyewitnesses reported remarkably clear and distinct green twilight colours after the eruption of Krakatoa (Sunda Strait, Indonesia) on August 27, 1883. To our best knowledge, no earlier studies exist providing an explanation for this unusual phenomenon. In the current work we employ simulations with the SCIATRAN radiative transfer model to investigate the processes leading to green volcanic sunsets. Green sunsets can be simulated based on plausible assumptions by anomalous scattering on stratospheric sulfate aerosols. We investigate the sensitivity of the twilight colours to relevant parameters such as aerosol optical depth, the parameters of the particle size distribution and the amount of ozone. The main requirements for the occurrence of green twilights are a sufficiently large aerosol optical depth combined with particle radii of about 500 – 700 nm (assuming stratospheric sulfate aerosols) and a preferably narrow aerosol particle size distribution. The occurrence of green twilights after historic eruptions provides important constraints on the particle size of volcanic aerosols.
- Preprint
(1746 KB) - Metadata XML
- BibTeX
- EndNote
Christian von Savigny et al.
Status: final response (author comments only)
-
EC1: 'Comment on egusphere-2023-1669', Timothy Garrett, 26 Aug 2023
There is a similar phenomenon of green sunsets that is seen for the case of storm clouds as discussed by Bohren and Fraser (1993) https://journals.ametsoc.org/view/journals/bams/74/11/1520-0477_1993_074_2185_gt_2_0_co_2.xml. Specifically, Bohren argues that the color of such sunsets owes to a blue-shift of red sunset light due to a combination of scattering and a preferential absorption by condensed water in the red. Can this explanation be excluded for the case described here? A substantial amount of condensed water would be required, but of course Krakatoa was an exceptional event.
Citation: https://doi.org/10.5194/egusphere-2023-1669-EC1 -
AC3: 'Reply on EC1', Christian von Savigny, 30 Oct 2023
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1669/egusphere-2023-1669-AC3-supplement.pdf
-
AC3: 'Reply on EC1', Christian von Savigny, 30 Oct 2023
-
RC1: 'Comment on egusphere-2023-1669', Filip Vanhellemont, 29 Sep 2023
Article: “Explaining the green volcanic sunsets after the 1883 eruption of Krakatoa “, by von Savigny et al.
The authors provide an explanation for the observed green twilight observations that were reported by eye witnesses after the eruption climax of Krakatoa in 1883. They use the well-known SCIATRAN radiative transfer code to simulate radiances, based on an educated assumption on the aerosol density profile shape. The perceived color is then estimated by converting radiances to chromaticity values using the CIE color matching functions. The color is studied as function of particle size distribution parameters, total ozone column and aerosol optical depth, for a number of solar zenith angles representative of twilight. The authors arrive at the clear conclusion that green twilight can be simulated for sufficiently large aerosols from a narrow size distribution and sufficiently large optical depth.
General Comments
This article is perfectly suited for publication in ACP. I am not aware of any other publication that presents an explanation of volcanically related green twilight, so the obtained conclusions are important. Furthermore, while it is difficult to obtain precise numbers on e.g. particle size from observations that are subjective (visual color perception), clear constraints/thresholds are obtained on particle size and optical depth. These are new numerical results on an important past volcanic event, and should be published.
The applied methodology is clearly explained, and it should be possible for other researchers to reproduce the results. The article title and the abstract perfectly describe the content. The body text is very well written and needs almost no adaptation. Plenty of references are provided.
I really don’t see any major problem with this paper; I have a few minor remarks that should be taken into account (see below), so I strongly recommend the publication of this paper.
Specific comments
- It is not clear to me at which wavelength Aerosol Optical Depth (AOD) is evaluated in the entire paper. This should be specified.
- The finding that green twilight is associated with narrow size distributions suggests that these green colors are only observed in the early stages of the stratospheric aerosol evolution (say, a few months after the event of August 27). Coagulation of particles shifts the distribution to higher size values, but also tends to widen the distribution, but this process takes quite some time. It might explain why no green twilight is reported at later times after the eruption (as far as I can tell from the paper). Perhaps you can add a small comment if you agree with this.
- In Fig. 3, the phenomenon of purple light that is often observed after large volcanic eruptions is also simulated, but it is not mentioned in the paper. A small description (one sentence or so) of this finding would enhance the credibility of the method even more.
Technical Corrections
- Line 3: volcanic activity of Krakatau started already several months before August 27. It is perhaps better to speak about the eruption climax of August 27, or something similar.
- Line 70: This sentence seems to be incorrect. Suggestion: ‘Please note that …, while the displayed …’.
- Figure 3: the caption seems to be wrong. Panel (b) shows Total Ozone Column, Panel (c) Aerosol Optical Depth, while the caption indicates the reverse.
Citation: https://doi.org/10.5194/egusphere-2023-1669-RC1 -
AC2: 'Reply on RC1', Christian von Savigny, 30 Oct 2023
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1669/egusphere-2023-1669-AC2-supplement.pdf
-
RC2: 'Comment on egusphere-2023-1669', Anonymous Referee #2, 04 Oct 2023
The paper describes a small study which explains why there can be green sunsets after volcanic eruptions. Radiative transfer model simulations are performed for various different aerosol assumptions to explain the green color and to also constrain stratospheric aerosol parameters, i.e. size distribution and aerosol optical thickness. The detailed discussion shows that the found parameters are reasonable. The authors suggest that the method could be used to constrain aerosol properties of historic eruptions, however I suppose that this would be difficult because there are not many reports about the sky colors available and of course also no quantitative observations.
The paper is clearly presented and since it explains the phenomenon of green sunsets to my knowledge for the first time I recommend to publish the paper after very minor corrections (see below).
Comments:
- Often, the term "anomalous scattering" is used. I think that this is misleading because the dependece of the scattering coefficient on wavelength is just a result of well-known Mie theory and not "anomalous".
- It is not mentioned how the optical properties of the aerosols were computed. I assume Mie theory?
- The term "volcanic sunset" in the title sounds strange. To me "Explaining the green sunsets after the 1883 volcanic eruption of Krakatoa" sounds better.
Citation: https://doi.org/10.5194/egusphere-2023-1669-RC2 -
AC1: 'Reply on RC2', Christian von Savigny, 30 Oct 2023
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1669/egusphere-2023-1669-AC1-supplement.pdf
-
AC1: 'Reply on RC2', Christian von Savigny, 30 Oct 2023
Christian von Savigny et al.
Christian von Savigny et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 351 | 108 | 26 | 485 | 9 | 10 |
- HTML: 351
- PDF: 108
- XML: 26
- Total: 485
- BibTeX: 9
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1