Abstract Details
(2020) An Internally Consistent Framework for the Global and Regional Chemical Variability of Parental Arc Magmas
Turner S & Langmuir C
https://doi.org/10.46427/gold2020.2640
No presentation has been uploaded by the author.
05d: Room 2, Friday 26th June 22:21 - 22:24
Stephen Turner
View abstracts at 7 conferences in series
Charles Langmuir View all 5 abstracts at Goldschmidt2020 View abstracts at 7 conferences in series
Charles Langmuir View all 5 abstracts at Goldschmidt2020 View abstracts at 7 conferences in series
Listed below are questions that have been submitted by the community that the author will try and cover in their presentation. To submit a question, ensure you are signed in to the website. Authors or session conveners approve questions before they are displayed here.
Submitted by Stephen Turner on Wednesday 17th June 18:29
If the above slideshow does not play, please find my presentation at http://stephenjturner.net/
If the above slideshow does not play, please find my presentation at http://stephenjturner.net/
Submitted by Lucy McGee on Thursday 25th June 08:27
Hi Steve - this is a compelling argument! I looked at the reference you cite regarding U-series characteristics - I see that the main idea there surrounds the fractionation of U and Th in accessory phases in marine sediments. How about the generally positive correlation between both U and Ra excesses in arc basalts with TE ratios such as K/La, Ba/Th and Sr anomalies? It would be interesting to see what oceanic crustal melts look like in the U-series space. Lucy
Thanks for the question Lucy! I emphasized the pre-existing high Ba/Th of certain sediments here, but also maintain that the sediment component is a melt, and that the sediment melt generally has higher Ba/Th and U/Th than its source. With the exception of the Marianas volcaniclastic sediment, the other trace-element-focused sediment melting experiments I'm aware of (Johnson and Plank, Skora, Hermann and Rubatto) all preferentially retain Th in residual monazite or apatite. The Martindale et al volcaniclastic sediment is an outlier, and provides an explanation unique to the Marianas for why the high Ba/Th and U-series excess becomes muted on the islands close to subducting seamount chains. The very high volcaniclastic TiO2 contents also lead to greater residual rutile and drive up Th/Nb. Melting experiments on natural ocean crust (Carter et al., 2015; Sisson and Kelemen, 2018) retain Th in either epidote or apatite, and are also characterized by high U/Th and Ba/Th (and presumably Ra/Th), so in the presence of ubiquitous ocean crust melts we should also expect U and Ra excess in arcs to be common. In the Carter et al. experiments, for example, U/Th in the AOC melts is about an order of magnitude greater than the starting material. In fact, it's these very same residual minerals which would produce high K/La and Sr/Nd!
Hi Steve - this is a compelling argument! I looked at the reference you cite regarding U-series characteristics - I see that the main idea there surrounds the fractionation of U and Th in accessory phases in marine sediments. How about the generally positive correlation between both U and Ra excesses in arc basalts with TE ratios such as K/La, Ba/Th and Sr anomalies? It would be interesting to see what oceanic crustal melts look like in the U-series space. Lucy
Thanks for the question Lucy! I emphasized the pre-existing high Ba/Th of certain sediments here, but also maintain that the sediment component is a melt, and that the sediment melt generally has higher Ba/Th and U/Th than its source. With the exception of the Marianas volcaniclastic sediment, the other trace-element-focused sediment melting experiments I'm aware of (Johnson and Plank, Skora, Hermann and Rubatto) all preferentially retain Th in residual monazite or apatite. The Martindale et al volcaniclastic sediment is an outlier, and provides an explanation unique to the Marianas for why the high Ba/Th and U-series excess becomes muted on the islands close to subducting seamount chains. The very high volcaniclastic TiO2 contents also lead to greater residual rutile and drive up Th/Nb. Melting experiments on natural ocean crust (Carter et al., 2015; Sisson and Kelemen, 2018) retain Th in either epidote or apatite, and are also characterized by high U/Th and Ba/Th (and presumably Ra/Th), so in the presence of ubiquitous ocean crust melts we should also expect U and Ra excess in arcs to be common. In the Carter et al. experiments, for example, U/Th in the AOC melts is about an order of magnitude greater than the starting material. In fact, it's these very same residual minerals which would produce high K/La and Sr/Nd!
Submitted by Cin-Ty Lee on Friday 26th June 00:35
nice work steve. so if these arcs have a positive Sr anomaly, do we see lithololigies with negative Sr anomalies? like eclogites?
Good question Cin-ty, Balancing the Sr/Nd budget of the Earth is tricky, as Sr often ends up sequestered strongly into certain phases/types of rock.
Chrondritic Sr/Nd is 15.5-16. The estimate of the upper continental crust from Rudnick and Gao, is ~12, presumably reflecting dominance of material that has undergone plagioclase fractionation and weathering. That number is based on an average of various estimates, some of which do not include Carbonates (rich in Sr), so the value is undoubtedly too low, but we don't know by how much. The net result is a bulk continental crust estimate of 16. If true, then there has to be a recycled continental reservoir with high Sr/Nd in addition to a recycled slab reservoir with low Sr/Nd. There is a lot of uncertainty in both upper and lower crustal estimates though.
If by eclogites you are referring to exhumed eclogites that may be representative of subducted rocks -- we should have no expectation of these as no exhumed eclogites to my knowledge record conditions above the solidus. What do arclogites look like? Most estimates of the lower continental crust are high, and if they form beyond the stable range for plag would presumably retain arc-like Sr/Nd.
But this all avoids the question of the required low/high Sr/Nd reservoirs. I have thought about this a little bit but don't have a definitive answer, but here's some food for thought:
Above is a plot of high-MgO Mid-Atlantic Ridge basalt that has been filtered based on Eu/Eu*. This is necessary to remove the pervasive signal of plagioclase fractionation and accumulation in MORB samples. The low 143/144 samples are all from the Southern section of the ridge which is influenced by hotspots thought to be linked to the LLSVP. The other end-member is then highly depleted MORB, which appears to have highly subchondritic Sr/Nd. Could the Mid-Atlantic ridge array be formed from a mixture of these recycled reservoirs?!
nice work steve. so if these arcs have a positive Sr anomaly, do we see lithololigies with negative Sr anomalies? like eclogites?
Good question Cin-ty, Balancing the Sr/Nd budget of the Earth is tricky, as Sr often ends up sequestered strongly into certain phases/types of rock.
Chrondritic Sr/Nd is 15.5-16. The estimate of the upper continental crust from Rudnick and Gao, is ~12, presumably reflecting dominance of material that has undergone plagioclase fractionation and weathering. That number is based on an average of various estimates, some of which do not include Carbonates (rich in Sr), so the value is undoubtedly too low, but we don't know by how much. The net result is a bulk continental crust estimate of 16. If true, then there has to be a recycled continental reservoir with high Sr/Nd in addition to a recycled slab reservoir with low Sr/Nd. There is a lot of uncertainty in both upper and lower crustal estimates though.
If by eclogites you are referring to exhumed eclogites that may be representative of subducted rocks -- we should have no expectation of these as no exhumed eclogites to my knowledge record conditions above the solidus. What do arclogites look like? Most estimates of the lower continental crust are high, and if they form beyond the stable range for plag would presumably retain arc-like Sr/Nd.
But this all avoids the question of the required low/high Sr/Nd reservoirs. I have thought about this a little bit but don't have a definitive answer, but here's some food for thought:
Above is a plot of high-MgO Mid-Atlantic Ridge basalt that has been filtered based on Eu/Eu*. This is necessary to remove the pervasive signal of plagioclase fractionation and accumulation in MORB samples. The low 143/144 samples are all from the Southern section of the ridge which is influenced by hotspots thought to be linked to the LLSVP. The other end-member is then highly depleted MORB, which appears to have highly subchondritic Sr/Nd. Could the Mid-Atlantic ridge array be formed from a mixture of these recycled reservoirs?!
Sign in to ask a question.