Biogeochemistry of Riverine Organic Matter Inputs to the Patagonian Fjords and Implications for Fjord Organic Carbon Budgets

Journal of Geophysical Research Biogeosciences, Journal Year: 2025, Volume and Issue: 130(1)

Published: Jan. 1, 2025

Abstract Fjords are increasingly recognized as hotspots for organic carbon (OC) burial. The OC buried in fjords is of both marine and terrestrial origin, with a predominance worldwide. proportions traditionally calculated using end‐member modeling based on δ 13 C and/or N/C. However, characterizing the remains challenge, authors inconsistently measurements obtained land plants, soils, river sediments. Here, we analyzed TOC, C, N/C composition soil samples, suspended sediments, bulk grain‐size fractions sediments from main rivers discharging into Patagonian (44–48°S), to identify processes that affect biogeochemistry matter reaching via rivers. Radiocarbon indicate contain 0.18% petrogenic variable concentrations biospheric OC. Despite significantly decreasing precipitation, relatively stable around −27‰. In contrast, highly variable, mostly due high contribution nitrogen glacier‐fed Furthermore, varies sediment grain size, making it virtually impossible define fixed value represent end‐member. By comparison, size has limited influence C. Overall, our results support use riverine mixing models, regardless presence glaciers watershed, they suggest fraction fjord may have been underestimated.

Language: Английский

---

Characterizing sedimentary organic carbon in a hydrothermal spreading center, the Escanaba Trough

Chemical Geology, Journal Year: 2025, Volume and Issue: unknown, P. 122679 - 122679

Published: Feb. 1, 2025

Language: Английский

---

A Deep Learning Approach of Artificial Neural Network With Attention Mechanism to Predicting Marine Biogeochemistry Data

Journal of Geophysical Research Biogeosciences, Journal Year: 2025, Volume and Issue: 130(3)

Published: March 1, 2025

Abstract Predicting marine biogeochemical data is an effective method to solve the problem of data‐scarcity and provides support for fundamental research in science. Machine learning techniques are commonly used improve stability accuracy predicting biogeochemistry data. However, current methods based on Random Forest (RF) Artificial Neural network (ANN) often struggle effectively capture intricate features ocean data, resulting suboptimal prediction accuracy. In this study, we develop a novel deep called artificial neural with attention mechanism (ANN‐att) We compare evaluate performance RF, ANN, ANN‐att two widely sets biogeochemistry: GLODAP v2.2022 MOSAIC 2.0. Our results show that higher than other by 6% 30% v.2.0. Additionally, maps surface dissolved oxygen Δ 14 C West Pacific demonstrate has significant advantage stronger nonlinear characteristics.

Language: Английский

---

Sources and Dynamics of Dissolved Organic Carbon in Kongsfjorden: Insights From Radiocarbon Analysis

Geophysical Research Letters, Journal Year: 2025, Volume and Issue: 52(8)

Published: April 24, 2025

Abstract The Arctic is undergoing rapid warming, resulting in accelerating glacier melt and release of nutrients, particles, organic matter into coastal fjord systems. Kongsfjorden, a Svalbard, serves as natural laboratory for investigating climate‐driven shifts high‐latitude ecosystems their broader implications the Arctic. To examine dissolved carbon (DOC) sources cycling we collected samples seawater, floating ice, river water June 2023 analyzed 14 C 13 DOC. Radiocarbon values DOC (as Δ C) seawater ranged from −302‰ to −253‰, while Bayelva River were higher, suggesting contribution younger, terrigenous However, stable isotope results suggested removal riverine after entering fjord. Additionally, differences between inner outer reflected varying contributions surface primary production glacial meltwater inputs.

Language: Английский

---

A review of machine learning applications to geophysical logging inversion of unconventional gas reservoir parameters

Earth-Science Reviews, Journal Year: 2024, Volume and Issue: unknown, P. 104969 - 104969

Published: Oct. 1, 2024

Language: Английский

---