Abstract Details
(2020) Evaluating Groundwater Discharge Using Age-Dating Tracers
Humphrey CE, Solomon DK, Mittelstet AR & Gilmore TE
https://doi.org/10.46427/gold2020.1114
The author has not provided any additional details.
13e: Room 4, Thursday 25th June 08:09 - 08:12
C. Eric Humphrey
D. Kip Solomon View all 2 abstracts at Goldschmidt2020 View abstracts at 7 conferences in series
Aaron R. Mittelstet
Troy E. Gilmore
D. Kip Solomon View all 2 abstracts at Goldschmidt2020 View abstracts at 7 conferences in series
Aaron R. Mittelstet
Troy E. Gilmore
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 Rolf Kipfer on Wednesday 24th June 09:46
Dear Humphrey Thanks for that nice and timing presentation on how to EXPERIMENTALLY determine base flow contribution to river discharge on the catchment scale. Your observation that tracer-based age scaling exceeds hydraulic age estimates is line with many other studies showing that catchments loose more 'old' groundwaters than models predict (e.g. the paradox of old groundwater). My question: You mentioned that some 'specific' pathways might have long transit times. Such 'long flowlines' allow 4He to accumulate. Thus, are some of your tube wells enriched in total He and do these concentrations somehow scale with the C-14 activity? A follow up question - you mentioned but did not discuss the C-14 data. Please, can you shortly comment? Best regards, Rolf Kipfer (RoKi)
Hello RoKi, thank you for the insightful question. To briefly answer, while this aeolian sand depositional environment has very little uranium and thorium and therefore low terrigenic 4He, we do observe a negative, linear correlation between 4He and 14C in older waters. Below is a plot [I will show plot during the Q/A session] showing only samples with 3H less than 0.4 TU (i.e. waters older than the 1953+ “bomb peak”) which shows that linear correlation. To answer your second question regarding the 14C data, this is something which our colleagues at North Carolina State University are studying. Most standard 14C models of our samples show negative ages which is likely due to the ubiquitous pedogenic nature of subsurface carbonate in the system with non-zero 14C activity. There are additional complications, including both C3 and C4 plants in the system and still undetermined spatial recharge patterns. We used a constant initial 14C value of 0.85 PMC, however it is possible that the Ao range can be between 0.66 and 0.95PMC based on preliminary modeling. For example, there are samples with tritium less than 0.4TU and nearly zero terrigenic 4He and have 14C activities greater than 0.9 PMC. Importantly, while the 14C ages need further work, we don’t anticipate changes to the basic shape of the TTD or substantially change the mean residence time.
Dear Humphrey Thanks for that nice and timing presentation on how to EXPERIMENTALLY determine base flow contribution to river discharge on the catchment scale. Your observation that tracer-based age scaling exceeds hydraulic age estimates is line with many other studies showing that catchments loose more 'old' groundwaters than models predict (e.g. the paradox of old groundwater). My question: You mentioned that some 'specific' pathways might have long transit times. Such 'long flowlines' allow 4He to accumulate. Thus, are some of your tube wells enriched in total He and do these concentrations somehow scale with the C-14 activity? A follow up question - you mentioned but did not discuss the C-14 data. Please, can you shortly comment? Best regards, Rolf Kipfer (RoKi)
Hello RoKi, thank you for the insightful question. To briefly answer, while this aeolian sand depositional environment has very little uranium and thorium and therefore low terrigenic 4He, we do observe a negative, linear correlation between 4He and 14C in older waters. Below is a plot [I will show plot during the Q/A session] showing only samples with 3H less than 0.4 TU (i.e. waters older than the 1953+ “bomb peak”) which shows that linear correlation. To answer your second question regarding the 14C data, this is something which our colleagues at North Carolina State University are studying. Most standard 14C models of our samples show negative ages which is likely due to the ubiquitous pedogenic nature of subsurface carbonate in the system with non-zero 14C activity. There are additional complications, including both C3 and C4 plants in the system and still undetermined spatial recharge patterns. We used a constant initial 14C value of 0.85 PMC, however it is possible that the Ao range can be between 0.66 and 0.95PMC based on preliminary modeling. For example, there are samples with tritium less than 0.4TU and nearly zero terrigenic 4He and have 14C activities greater than 0.9 PMC. Importantly, while the 14C ages need further work, we don’t anticipate changes to the basic shape of the TTD or substantially change the mean residence time.
Submitted by Roland Purtschert on Wednesday 24th June 16:44
Thanks for the excellent and interesting presentation. Could you explain a bit in more detail how you extraxt from 3H/3He data for the young water and 14C data for old water a complete TTD over a large range of timescales (as you show on the slide "overview of points...."). You assume a 2 (or 3) parameter model?
Hello Roland and thank you for the question. The slide “Overview of Points Method” show raw transit time distributions without modeling. This distribution is created by computing the fraction of discharge of a single age measurement to the total groundwater discharging into the system and plotting this in cumulative fashion. Unfortunately, I did not have time in my presentation to show initial modeling results, but I have included those here [to be shown in Q/A session]. Three different models are used in these distributions: (1) the Gamma model, (2) the Exponential Model (EM), and (3) The Exponential Piston-flow Model (EPM). The gamma model has two parameters, alpha and beta of which the product is the mean transit time (MTT). The exponential model assumes a simple aquifer of uniform thickness and areal recharge and can be thought of as the gamma model with alpha equal to one. The EPM model assumes uniform aquifer thickness, porosity, and areal recharge with flow into a confined portion of the aquifer; all model parameters can be variable. A continuity problem emerges when modeling at such large spatial and temporal scales. As seen in the “raw” TTD, there is a significant gap between waters of roughly 100 years old and waters around 1000 years old. This apparent bimodal distribution could be an artifact of merging 3H/3He with 14C age-dating methods. Further refinement of the 14C ages in necessary to be confident in the 14C ages as previously stated in the above question. Another possible solution to this apparent age gap is to incorporate another age-dating technique such as 39Ar, which could nicely target waters outside the range of the applied 3H/3He and 14C methods. To my knowledge, argon-39 age-dating has not been applied in the Sandhills, but would be a welcome addition to the ongoing project.
Thanks for the excellent and interesting presentation. Could you explain a bit in more detail how you extraxt from 3H/3He data for the young water and 14C data for old water a complete TTD over a large range of timescales (as you show on the slide "overview of points...."). You assume a 2 (or 3) parameter model?
Hello Roland and thank you for the question. The slide “Overview of Points Method” show raw transit time distributions without modeling. This distribution is created by computing the fraction of discharge of a single age measurement to the total groundwater discharging into the system and plotting this in cumulative fashion. Unfortunately, I did not have time in my presentation to show initial modeling results, but I have included those here [to be shown in Q/A session]. Three different models are used in these distributions: (1) the Gamma model, (2) the Exponential Model (EM), and (3) The Exponential Piston-flow Model (EPM). The gamma model has two parameters, alpha and beta of which the product is the mean transit time (MTT). The exponential model assumes a simple aquifer of uniform thickness and areal recharge and can be thought of as the gamma model with alpha equal to one. The EPM model assumes uniform aquifer thickness, porosity, and areal recharge with flow into a confined portion of the aquifer; all model parameters can be variable. A continuity problem emerges when modeling at such large spatial and temporal scales. As seen in the “raw” TTD, there is a significant gap between waters of roughly 100 years old and waters around 1000 years old. This apparent bimodal distribution could be an artifact of merging 3H/3He with 14C age-dating methods. Further refinement of the 14C ages in necessary to be confident in the 14C ages as previously stated in the above question. Another possible solution to this apparent age gap is to incorporate another age-dating technique such as 39Ar, which could nicely target waters outside the range of the applied 3H/3He and 14C methods. To my knowledge, argon-39 age-dating has not been applied in the Sandhills, but would be a welcome addition to the ongoing project.
Sign in to ask a question.