In the past two decades, a number of studies have been carried out in the Southern Ocean to look at export production using drifting sediment traps and thorium‐234 based measurements, which allows us to reexamine the validity of using the existing relationships between production, export efficiency, and temperature to derive satellite‐based carbon export estimates in this region. Comparisons of in situ export rates with modeled rates indicate a two to fourfold overestimation of export production by existing models. Comprehensive analysis of in situ data indicates two major reasons for this difference: (i) in situ data indicate a trend of decreasing export efficiency with increasing production which is contrary to existing export models and (ii) the export efficiencies appear to be less sensitive to temperature in this region compared to the global estimates used in the existing models. The most important implication of these observations is that the simplest models of export, which predict increase in carbon flux with increasing surface productivity, may require additional parameters, different weighing of existing parameters, or separate algorithms for different oceanic regimes.

Continental shelves are important land–ocean interfaces where natural and anthropogenic processes as well as sediment geochemistry regulate organic matter sequestration. The major objective of this work is to understand spatial variability in reactive iron (rFe) bound organic carbon (OC) during low‐ and high‐flow conditions in the recent sediments of the Mississippi River‐influenced Louisiana shelf in the Northern Gulf of Mexico. Sediment samples were collected from five sites along the river plume during low (August 2016) and high (May 2017) discharge. Reactive iron was significantly higher in May (6.9 ± 1.7 mg g⁻¹_dw) compared to August (4.3 ± 0.9 mg g⁻¹_dw), while OC in August (11.9 ± 3.7 mg g⁻¹_dw) and May (7.8 ± 4.8 mg g⁻¹_dw) was similar. The fraction of OC bound to rFe (Fe‐OC) in the sediment in May (2.4 ± 1.5 mg g⁻¹_dw) was not significantly different from August (3.7 ± 2.7 mg g⁻¹_dw). Molar OC:Fe ratios in the rFe‐associated OC ranged between 0.9 and 6.2, with higher ratios in the deeper sediments (6–20 cm). Our results show that 31.7% ± 6.2% of the total OC in the recent sediments is associated with the rFe phase. Sediment mineralogical composition was characterized using X‐ray diffractometry and Fe‐K edge X‐ray absorption near edge structure spectroscopy. Our findings indicate that most of the iron is present as nonreactive iron in clay mineral smectite and the dominant form of rFe associated with OC is goethite.

Intercomparision of ²³⁴Th measurements in both water and particulate samples was carried out between 15 laboratories worldwide, as a part of GEOTRACES inter‐calibration program. Particulate samples from four different stations namely BATS (both shallow and deep) and shelf station (shallow) in Atlantic and SAFE (both shallow and deep) and Santa Barbara station (shallow) in Pacific were used in the effort. Particulate intercalibration results indicate good agreement between all the participating labs with data from all labs falling within the 95% confidence interval around the mean for most instances. Filter type experiments indicate no significant differences in ²³⁴Th activities between filter types and pore sizes (0.2–0.8 µm). The only exception are the quartz filters, which are associated with 10% to 20% higher ²³⁴Th activities attributed to sorption of dissolved ²³⁴Th. Flow rate experiments showed a trend of decreasing ²³⁴Th activities with increasing flow rates (2‐9 L min⁻¹) for > 51 µm size particles, indicating particle loss during the pumping process. No change in ²³⁴Th activities on small particles was observed with increasing flow‐rates. ²³⁴Th intercalibration results from deep water samples at SAFe station indicate a variability of < 3% amongst labs while dissolved ²³⁴Th data from surface water at Santa Barbara Station show a less robust agreement, possibly due to the loss of ²³⁴Th from decay and large in‐growth corrections as a result of long gap between sample collection and processing.

Although many studies have attempted to reconstruct millennial‐scale hurricane patterns using various proxy‐based methods, it is still unclear what the most effective proxies are to identify storm surge deposits in different environmental settings. This study quantitatively compares the application of grain‐size, loss‐on‐ignition, stable isotopes, X‐ray fluorescence, and palynological proxies in paleotempestology from an organic‐rich environment in the Florida Everglades. The nonparametric tests indicate that only 9 among the 27 parameters (mean diameter, %water, %organic, %carbonate, Ca, Sr, Ca/Ti, Cl/Br, and marine microfossils) exhibited significant differences between storm‐surge and in situ deposits. The principal component analysis shows that five marine indicators (Sr, Ca, Ca/Ti, %Carbonate, and Marine microfossils) have the closest association with the allochthonous samples, while Cl/Br and Mz are the most sensitive proxies in low‐ and high‐energy environments, respectively. Moreover, organic geochemical proxies (e.g., δ¹³C and δ¹⁵N of bulk sedimentary organic matter) are ineffective for identifying storm‐surge deposits in organic‐rich mangrove environments.