the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 14 Nov 2022
Multiple Early Holocene eruptions of Katla produced tephra layers with similar composition to the Vedde Ash
Abstract. The Vedde Ash, first described in Norway and dated to ~12000 cal a BP, has been taken to represent tephra derived from a large eruption of the Katla volcano in Iceland and dispersed across the North Atlantic and Europe. However, evidence for tephra layers with similar composition to the Vedde Ash, but of different ages, questions the utility of isolated Vedde-like tephra layers as reliable and independent age control. Here, we report three stratigraphically separated Early Holocene Katla tephra layers from the lake Torfdalsvatn, in north Iceland, that have bimodal chemical composition similar to the Vedde Ash. By using previously published conventional 14C ages and revised calibration curves, we provide new ages for these tephra layers of ~11375, 11360, and 11200 cal a BP – all substantially younger than the Vedde Ash. Torfdalsvatn’s record stands as an important reminder that repeated explosive eruptions of Iceland’s major volcanos during the deglacial cycle have produced multiple tephra plumes with similar geochemistry that may span 1000s of years. As a result, we urge caution when using isolated Icelandic tephra layers in distant regions as precise geochronometers without supporting age control.
This preprint has been withdrawn.
-
Withdrawal notice
This preprint has been withdrawn.
-
Preprint
(3524 KB)
-
Supplement
(391 KB)
-
This preprint has been withdrawn.
- Preprint
(3524 KB) - Metadata XML
-
Supplement
(391 KB) - BibTeX
- EndNote
Interactive discussion
Status: closed
-
CC1: 'Comment on gchron-2022-26', Jan Mangerud, 17 Nov 2022
Very interesting manuscript. But:
- I would like to see counting of glass chards between and below the Vedde-like ash layers in order to better evaluate the possibility of redeposition. In western Norway we find a tail of redeposited Vedde ash well into the Holocene.
- I am surprised that they did not obtain new radiocarbon ages.
Citation: https://doi.org/10.5194/gchron-2022-26-CC1 - AC1: 'Reply on CC1', David Harning, 28 Nov 2022
-
RC1: 'Comment on gchron-2022-26', Anonymous Referee #1, 22 Nov 2022
I recently reviewed a similar manuscript by the same authors for another journal. I had raised several critical comments, and based on my review, the editor decided to reject the ms. I was happy to see that the authors decided to submit the ms again to GChron, but my main criticism remains after reading through the text. I summarise my main critical comments here. See also graphs in the supplement.
1) It is feasible that several eruptions of Katla occurred in the early Holocene, producing Vedde-type tephras. The results presented here, however, do not support this conclusion. The age-depth model is based on radiocarbon dates performed more than 30 years ago with standard deviations of several hundred years. Age-depth modelling of such old dates cannot give reliable ages and new samples for radiocarbon dating should have been submitted. The second lowest date (Ua-1888) is potentially an outlier, and by removing it from the model, all depths would get older ages in line with the interpretation of Björck et al. (1992)
2) There is no discussion about the possibility of reworking of Vedde shards in the early Holocene. There are many examples from western Norway of reworking many thousand years into the Holocene. I agree with Mangerud's comments that a tephra count graph would be helpful here.
3) The lowest tephra layer, Tv-1 has a Hekla-like geochemistry and is believed to be one of the oldest basaltic layers from Hekla. The authors, however, do not mention that Tv-1 has been correlated with a tephra found in the NGRIP ice-core, NGRIP1519-1, dated to c 12,646 b2k (Mortensen et al., 2005; JQS). New data from NGRIP and GRIP confirms this correlation and firmly places the Tv-1/NGRIP1519-1 in the early part of Younger Dryas/GS-1 (Cook et al., 2022; QSR). The attached graph (Fig. 1 in the supplement) shows an alternative age model for the Torfdalsvatn core based on ice-core ages of the Tv-1 and Tv-4 tephras (Mortensen et al. 2005; Cook et al. 2022). It suggests that the Tv-2 layer (Vedde in Björck et al's paper) is firmly placed in the YD and not in early Preboreal. Biplots (Fig. 2 in the supplement) show that there is a generally good agreement between the Tv-1 tephra from Torfdalsvatn and the NGRIP1519.1/GRIP1654.05 layer from Greenland ice. There is more spread in the ice core samples which might be due to smaller shard that were analysed, but all major elements overlap.
4) Previous investigations at Torfdalsvatn by Rundgren (1995, QR) are mentioned in the text, but rather surprisingly, are not discussed in any detail. The pollen- and lithostratigraphy presented by Rundgren suggest an YD age for Tv-2 and an early YD age for Tv-1, in agreement Björck's paper, see also Fig. 3 in the supplement file.
- AC3: 'Reply on RC1', David Harning, 28 Nov 2022
-
CC2: 'Comment on gchron-2022-26', Simon Larsson, 23 Nov 2022
Tephra studies are becoming increasingly complex as we are refining methodology and analysis and thereby detecting hitherto unknown deposits, reinterpreting old ones, and having more and more trouble separating some of them (chemically and chronologically). It is good to see a study with the intention to highlight a possible problem with a well-used isochron such as the Vedde Ash, but there are some major issues with this manuscript that would probably have to be remedied for the conclusions to be accepted.
- As already commented, the study is missing a complete tephrostratigraphic description and has only sampled visible tephra layers (I'm assuming - this is not explicitly stated. While these can be argued to be "stratigraphically separated", there is a high likelihood of redeposition in such an environment as that of the study site and a complete tephra count for the sequence is, in my opinion, required to motivate the layers as being separate and primary deposits.
- The study is missing a complete description of the creation of the age models. It is specified that it is Bayesian and that IntCal20 was applied to create the age model for the older study's sediment sequence, but no other specifics and no reference for software used is included. The creation of the age model for the present study's sediment sequence is not described explicitly (should it be assumed that the same procedure is applied as for the older?).
- As already commented, the study re-uses quite old radiocarbon dates with wide error margins, recalibrating them to create an age model for the old sediment sequence. This is interesting for comparison purposes but should be interpreted with a great deal of care. The suggested ages for the three tephra layers are based on linear interpolation between two other tephra dates spaced >60 cm apart, which is in my opinion not too robust of an age model. The study would greatly benefit from, if not require, a more complete chronology.
- There is little-to-no lithostratigraphic description of the sample cores. This should be provided and expanded upon when comparing the sediment sequences of the older study and the current one, and hopefully this could provide better motivation for the assumptions being made about the tephra findings in the current study correlating to those in the older study.
Citation: https://doi.org/10.5194/gchron-2022-26-CC2 - AC2: 'Reply on CC2', David Harning, 28 Nov 2022
-
RC2: 'Comment on gchron-2022-26', Anonymous Referee #2, 12 Jan 2023
Ultimately, sharp contacts and pristine glass shards do not provide robust evidence that an ash layer is primary. Moreover, since the tephra layers identified here are compositionally identical and in sediments younger than the Vedde ash, other lines of evidences are essential to categorically prove these are not redeposited Vedde ash. Therefore I find the conclusions of the paper here are unsupported. Numerous other high-resolution lake records have shown that it is extremely difficult to discriminate reworked ash from older (primary) events, and that ash can be redeposited, producing discrete visible and cryptic layers with perfect pristine glass shards (even over tens of thousands of years after the original eruption). Rigorous morphological, geochemical and sedimentological work has not been demonstrated, and may not be able to rule out the possibility of redeposition.
The authors also need to show the integrity of the lake sediments, showing the tephrostratigraphy over a longer period of sedimentation and show the compositions of glass shards incorporated within the sediments over different timescales. I would suggest presenting the complete tephrostratigraphy of the lake sediments and including the work in this manuscript alongside a discussion of interpreting taphonomic issues like these that are commonly faced by the tephrochronological community working with sedimentary records.
Citation: https://doi.org/10.5194/gchron-2022-26-RC2 - AC4: 'Reply on RC2', David Harning, 13 Jan 2023
Interactive discussion
Status: closed
-
CC1: 'Comment on gchron-2022-26', Jan Mangerud, 17 Nov 2022
Very interesting manuscript. But:
- I would like to see counting of glass chards between and below the Vedde-like ash layers in order to better evaluate the possibility of redeposition. In western Norway we find a tail of redeposited Vedde ash well into the Holocene.
- I am surprised that they did not obtain new radiocarbon ages.
Citation: https://doi.org/10.5194/gchron-2022-26-CC1 - AC1: 'Reply on CC1', David Harning, 28 Nov 2022
-
RC1: 'Comment on gchron-2022-26', Anonymous Referee #1, 22 Nov 2022
I recently reviewed a similar manuscript by the same authors for another journal. I had raised several critical comments, and based on my review, the editor decided to reject the ms. I was happy to see that the authors decided to submit the ms again to GChron, but my main criticism remains after reading through the text. I summarise my main critical comments here. See also graphs in the supplement.
1) It is feasible that several eruptions of Katla occurred in the early Holocene, producing Vedde-type tephras. The results presented here, however, do not support this conclusion. The age-depth model is based on radiocarbon dates performed more than 30 years ago with standard deviations of several hundred years. Age-depth modelling of such old dates cannot give reliable ages and new samples for radiocarbon dating should have been submitted. The second lowest date (Ua-1888) is potentially an outlier, and by removing it from the model, all depths would get older ages in line with the interpretation of Björck et al. (1992)
2) There is no discussion about the possibility of reworking of Vedde shards in the early Holocene. There are many examples from western Norway of reworking many thousand years into the Holocene. I agree with Mangerud's comments that a tephra count graph would be helpful here.
3) The lowest tephra layer, Tv-1 has a Hekla-like geochemistry and is believed to be one of the oldest basaltic layers from Hekla. The authors, however, do not mention that Tv-1 has been correlated with a tephra found in the NGRIP ice-core, NGRIP1519-1, dated to c 12,646 b2k (Mortensen et al., 2005; JQS). New data from NGRIP and GRIP confirms this correlation and firmly places the Tv-1/NGRIP1519-1 in the early part of Younger Dryas/GS-1 (Cook et al., 2022; QSR). The attached graph (Fig. 1 in the supplement) shows an alternative age model for the Torfdalsvatn core based on ice-core ages of the Tv-1 and Tv-4 tephras (Mortensen et al. 2005; Cook et al. 2022). It suggests that the Tv-2 layer (Vedde in Björck et al's paper) is firmly placed in the YD and not in early Preboreal. Biplots (Fig. 2 in the supplement) show that there is a generally good agreement between the Tv-1 tephra from Torfdalsvatn and the NGRIP1519.1/GRIP1654.05 layer from Greenland ice. There is more spread in the ice core samples which might be due to smaller shard that were analysed, but all major elements overlap.
4) Previous investigations at Torfdalsvatn by Rundgren (1995, QR) are mentioned in the text, but rather surprisingly, are not discussed in any detail. The pollen- and lithostratigraphy presented by Rundgren suggest an YD age for Tv-2 and an early YD age for Tv-1, in agreement Björck's paper, see also Fig. 3 in the supplement file.
- AC3: 'Reply on RC1', David Harning, 28 Nov 2022
-
CC2: 'Comment on gchron-2022-26', Simon Larsson, 23 Nov 2022
Tephra studies are becoming increasingly complex as we are refining methodology and analysis and thereby detecting hitherto unknown deposits, reinterpreting old ones, and having more and more trouble separating some of them (chemically and chronologically). It is good to see a study with the intention to highlight a possible problem with a well-used isochron such as the Vedde Ash, but there are some major issues with this manuscript that would probably have to be remedied for the conclusions to be accepted.
- As already commented, the study is missing a complete tephrostratigraphic description and has only sampled visible tephra layers (I'm assuming - this is not explicitly stated. While these can be argued to be "stratigraphically separated", there is a high likelihood of redeposition in such an environment as that of the study site and a complete tephra count for the sequence is, in my opinion, required to motivate the layers as being separate and primary deposits.
- The study is missing a complete description of the creation of the age models. It is specified that it is Bayesian and that IntCal20 was applied to create the age model for the older study's sediment sequence, but no other specifics and no reference for software used is included. The creation of the age model for the present study's sediment sequence is not described explicitly (should it be assumed that the same procedure is applied as for the older?).
- As already commented, the study re-uses quite old radiocarbon dates with wide error margins, recalibrating them to create an age model for the old sediment sequence. This is interesting for comparison purposes but should be interpreted with a great deal of care. The suggested ages for the three tephra layers are based on linear interpolation between two other tephra dates spaced >60 cm apart, which is in my opinion not too robust of an age model. The study would greatly benefit from, if not require, a more complete chronology.
- There is little-to-no lithostratigraphic description of the sample cores. This should be provided and expanded upon when comparing the sediment sequences of the older study and the current one, and hopefully this could provide better motivation for the assumptions being made about the tephra findings in the current study correlating to those in the older study.
Citation: https://doi.org/10.5194/gchron-2022-26-CC2 - AC2: 'Reply on CC2', David Harning, 28 Nov 2022
-
RC2: 'Comment on gchron-2022-26', Anonymous Referee #2, 12 Jan 2023
Ultimately, sharp contacts and pristine glass shards do not provide robust evidence that an ash layer is primary. Moreover, since the tephra layers identified here are compositionally identical and in sediments younger than the Vedde ash, other lines of evidences are essential to categorically prove these are not redeposited Vedde ash. Therefore I find the conclusions of the paper here are unsupported. Numerous other high-resolution lake records have shown that it is extremely difficult to discriminate reworked ash from older (primary) events, and that ash can be redeposited, producing discrete visible and cryptic layers with perfect pristine glass shards (even over tens of thousands of years after the original eruption). Rigorous morphological, geochemical and sedimentological work has not been demonstrated, and may not be able to rule out the possibility of redeposition.
The authors also need to show the integrity of the lake sediments, showing the tephrostratigraphy over a longer period of sedimentation and show the compositions of glass shards incorporated within the sediments over different timescales. I would suggest presenting the complete tephrostratigraphy of the lake sediments and including the work in this manuscript alongside a discussion of interpreting taphonomic issues like these that are commonly faced by the tephrochronological community working with sedimentary records.
Citation: https://doi.org/10.5194/gchron-2022-26-RC2 - AC4: 'Reply on RC2', David Harning, 13 Jan 2023
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 763 | 290 | 40 | 1,093 | 173 | 18 | 21 |
- HTML: 763
- PDF: 290
- XML: 40
- Total: 1,093
- Supplement: 173
- BibTeX: 18
- EndNote: 21
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Cited
1 citations as recorded by crossref.
Thor Thordarson
Áslaug Geirsdóttir
Gifford Miller
Christopher Florian
This preprint has been withdrawn.
- Preprint
(3524 KB) - Metadata XML
-
Supplement
(391 KB) - BibTeX
- EndNote