Residual β activity of particulate 234Th as a novel proxy for tracking sediment resuspension in the ocean

Sediment resuspension occurs in the global ocean, which greatly affects material exchange between the sediment and the overlying seawater. The behaviours of carbon, nutrients, heavy metals, and other pollutants at the sediment-seawater boundary will further link to climate change, eutrophication, and marine pollution. Residual β activity of particulate 234Th (RAP234) is used as a novel proxy to track sediment resuspension in different marine environments, including the western Arctic Ocean, the South China Sea, and the Southern Ocean. Sediment resuspension identified by high activity of RAP234 is supported by different lines of evidence including seawater turbidity, residence time of total 234Th, Goldschmidt’s classification, and ratio of RAP234 to particulate organic carbon. A conceptual model is proposed to elucidate the mechanism for RAP234 with dominant contributions from 234Th-238U and 212Bi-228Th. The ‘slope assumption’ for RAP234 indicated increasing intensity of sediment resuspension from spring to autumn under the influence of the East Asian monsoon system. RAP234 can shed new light on 234Th-based particle dynamics and should benefit the interpretation of historical 234Th-238U database. RAP234 resembles lithophile elements and has broad implications for investigating particle dynamics in the estuary-shelf-slope-ocean continuum and linkage of the atmosphere-ocean-sediment system.

Mechanism of RA P234 . The potential radionuclides associated with high RA P234 activity and suspended particles in the seawater can be classified into external and internal radionuclides. The external radionuclides associated with suspended particles refer to the surface-bound radionuclides with high particle reactivity. The radionuclides and their activities in natural seawater have previously been compiled 19 and can be classified into low and high particle reactivity (Table 1) according to their particle-seawater distribution coefficient (K d ) 20 . A high K d indicates high particle reactivity. Other artificial radionuclides with short half-lives are not considered due to a lack of nuclear facilities in our sampling region. Otherwise, radionuclides, such as 91 Y, 152 Eu, etc., should be taken into account when nuclear fuel reprocessing facilities are in operation near this sea area 21 .
Although the components of suspended particles, including POC, lithogenic materials, biogenic inorganic materials, and hydrogenous materials, display distinct affinity for radionuclides 22 , the requirements for major external radionuclides are high activity in seawater and high K d . The high activity of 234 Th with high particle reactivity results in the direct measurement of surface-bound particulate 234 Th without additional radiochemical separation. The β counts contributed by 210 Pb/ 210 Bi and other radionuclides with low activity and low energy β particles were shielded by a layer of Mylar film and two layers of aluminium foil (16 mg/cm 2 ) 23,24 , to prevent external contributions to the β counting 21 . Therefore, the dominant external radionuclide associated with the suspended particles is 234 Th after sample collection.
Surface-bound 234 Th with a short half-life (24.1 days) on the suspended particles was unsupported by its parent radionuclides 238 U, which remains in the seawater due to low K d . RA P234 was measured over 120 days after sampling. This surface-bound 234 Th on external particles will decay away. Consequently, the unsupported 234 Th adsorbed on external particles should not contribute to RA P234 .
In our study, only RA P234 derived from the second counting rate of particulate 234 Th was investigated. Low activity and low K d of radium in seawater lead to extremely low activity of radium for adsorption onto external suspended particles. Radium and its progenies, such as 224 Ra 25 , should not contribute to RA P234 via surface adsorption. Additionally, to our knowledge, there is no tectonically active region on the continental shelf of the Chukchi Sea to provide high radium activity 26 . Therefore, the radionuclides with low K d in seawater, such as radium, should not significantly contribute to RA P234 .
The internal radionuclides of suspended particles were more complicate than the external radionuclides. The suspended particles can be classified into terrigenous and biogenic particles in order to analyse the internal radionuclides.
In the bottom nepheloid layer, terrigenous particles resuspended from marine sediment could reach 70% 27 . The concentration of suspended particles in the bottom layer of seawater reached values of up to 9.5 mg/L on the continental shelf of the western Arctic Ocean during the 5th Chinese National Arctic Research Expedition (CHINARE-5), which was significantly higher than that of the upper seawater and indicated the occurrence of sediment resuspension. High concentrations of suspended particle material were also observed for bottom seawater in the western Arctic Ocean 10 . The activity, emitting particle type with energy, and the yield for radionuclides in the marine sediment are presented in Table 2. The activities of some radionuclides are consistent with the limiting direct measurement in the Chukchi Sea 28 . Some α -particle-emitting radionuclides are also presented, because the daughter radionuclides supported by these α -particle-emitting radionuclides could contribute to the β count, such as 226 Ra and its daughter radionuclides. Therefore, an exhaustive overview of the radionuclides in biogenic and terrigenous particles will benefit comprehensive understanding.
The requirements of the major contributors to internal radionuclides include high activity, high energy of β particles, and high yield. 234m Pa, the daughter radionuclide of 234 Th, emits β particles with a maximum energy of 2.28 MeV ( Table 2). The small-volume technique via β counting of 234 Th is on the basis of 234m Pa measurement. The lower energy β particles from other radionuclides were significantly shielded during source preparation with a layer of Mylar film and two layers of aluminium foil 23 . 234 Th, the parent radionuclide of 234m Pa, is supported by the primordial radionuclide 238 U in the minerals derived from the crust. 238 U in marine sediments was reported to have a mean activity of 50 Bq/kg in the Chukchi Sea 28 . Terrigenous particles from marine sediment via sediment resuspension can reach 70% in the bottom nepheloid layer 27 . It has also been reported that particulate 238 U can even reach 95% of total 238 U due to sediment resuspension 29 . Therefore, 234 Th supported by 238 U is probably the dominant contribution to internal radionuclides, especially on the shallow continental shelf with active hydrodynamics.
Although the β energy of 90 Y was the same as 234m Pa (Table 2), the activity of particulate 90 Y was very low in seawater. The anthropogenic radionuclide of 90 Sr, the parent radionuclide of 90 Y, was mainly from global fallout in the Chukchi Sea. Although direct measurement for 90 Sr had not, to our knowledge, been reported, a fingerprint activity ratio of 90 Sr to 137 Cs had a value of 0.63 for global fallout 30 . Combining the reported activity of 137 Cs with an average value of 2.0 Bq/kg for surface sediment in the Chukchi Sea 28 , the activity of 90 Sr was about 1.3 Bq/kg, which was less than 5% of that of 234m Pa supported by 234 Th/ 238 U. Therefore, the β counting rate contributed by 90 Y should be minor. 212 Bi, the progeny of 228 Th-232 Th, has the β energy of 2.25 MeV, with a yield of 48.4%. The β energy of 212 Bi is also close to that of 234m Pa with similar detector efficiency. The activity of 232 Th was generally comparable with www.nature.com/scientificreports/ that of 238 U in the marine sediment derived from the crust 31,32 . High activity of 228 Th in bottom layer seawater had previously been directly measured as a result of sediment resuspension in the Baltic Sea 31 . Therefore, 212 Bi supported by 228 Th-232 Th should be considered when sediment resuspension occurs due to its high activity, β energy, and yield. Although the existence of 40 K, 226 Ra, and 210 Pb had been confirmed by γ spectrometry for bottom seawater in the northeast Atlantic Ocean 33 , these radionuclides and their progenies have lower β particle energies. Their contribution to β counting rate should be minimal due to the shielding effect of aluminium foil 24 .
Biogenic particles make a major contribution to suspended particles in the upper ocean when photosynthesis occurs. The biotas make preferential use of low atomic number elements, such as C, N, P, S and others. Many radionuclides with high atomic numbers are not essential elements for these biotas. Typical radionuclides found in marine biotas are shown in Table 3. The dominant radionuclide amongst the marine biotas is 40 K, the activity of which is two orders of magnitude greater than that of the other radionuclides. The shield effect of aluminium foil during source preparation of particulate 234 Th limits the contribution by 40 K to RA P234 due to low β energy. Therefore, the activity of RA P234 was low in the euphotic layer due to a major fraction of suspended particles from photosynthesis.
The dominant radionuclides contributing to the high activity of RA P234 in bottom seawater are likely to be 234 Th-238 U and 212 Bi-228 Th in the particles resuspended from marine sediments. RA P234 was re-measured three times 120 days after the sampling date to check its stability. It indicated that RA P234 is constant due to the long half-lives of 238 U and 228 Th-232 Th and the relatively enclosed environment of the crystal lattice in the mineral derived from marine sediments that constrain any deficit or ingrowth process of the daughter-parent radionuclides.
However, the exact percentage of 234 Th-238 U and 212 Bi-228 Th for RA P234 was not obtained in our study due to limitations on the volume of seawater available. Only 4-8 L of seawater was sampled on the continental shelf for 234 Th analysis. The phenomenon, abnormally high activity of RA P234 , was found during data analysis. Although the radionuclides of 234 Th-238 U and 212 Bi-228 Th could be checked by α -spectrometry and γ -spectrometry, a large amount of seawater (> 100 L) should be sampled due to the lower detector efficiency of γ -spectrometry (< 1%) and α -spectrometry (10%~20%) relative to that of β -counter in this study (47 ± 2%). Chemical recovery should also be considered via α -spectrometry. RA P234 , which combines the dominant β signal from 234 Th-238 U and 212 Bi-228 Th, is sensitive enough to indicate sediment resuspension with small volume of seawater via β -counter with high detector efficiency.
The relative contributions of 234 Th-238 U and 212 Bi-228 Th should vary with distinct sea areas. The specific character of sediment and intensity of sediment resuspension will determine their relative contributions as well as activity of RA P234 .

Radionuclide
Typical activity (Bq/kg-d.w.) Emitting particle Energy (MeV) Yield (%) 40 15,34 . These observations indicate that there was an active interaction between the sediment and the seawater on the shallow continental shelf, especially during the ice-free summer season. This provides an environmental benefit for generating and sustaining sediment resuspension. An exceptionally strong summer cyclone was reported in early August, 2012 in the Chukchi Sea 35 . The strong mixing and upwelling caused by the cyclone resulted in a relatively well-mixed, vertically homogeneous water column on the continental shelf. Thus, the summer cyclone is another factor that is favourable to the generation of sediment resuspension. Therefore, it is reasonable to infer that sediment resuspension probably took place and redistributed RA P234 on the continental shelf.

Residence time of total 234
Th to indicate sediment resuspension. The residence time of total 234 Th was calculated and represented using an irreversible steady-state model (Appendix Table A1) 36 . Our results were consistent with other studies of this region [37][38][39] . The residence time of total 234 Th on the continental shelf was significantly shorter than that in the open Arctic Ocean. The high nutrients waters supplied from the North Pacific Ocean can support high photosynthesis and scavenge 234 Th on the continental shelf relating to nutrient depletion and low photosynthesis in the open ocean 40,41 .
The residence time of total 234 Th for bottom seawater was shorter than that for upper-layer seawater, which had been attributed to sediment resuspension to enhance the scavenging of 234 Th from seawater 42 . Therefore, the short residence time of total 234 Th for the bottom layer also provided another clue to sediment resuspension on this shallow, but hydrodynamically active, continental shelf. RA P234 and POC to indicate sediment resuspension. The relationship between RA P234 and POC was investigated in the western Arctic Ocean (Fig. 3). The slope of linear regression line between RA P234 and POC was about 0.160 Bq/mmol C for suspended particles. As for the end-member of sediment in the Chukchi Sea, the activity of 238 U was about 50 Bq/kg 28 , while the activity of 232 Th was generally comparable with that of 238 U in the marine sediment 32 . The average concentration of POC was about 1% with a range of 0.5% to 2% in the marine sediment 43 . Thus, the sediment fingerprint is characterized by its ratio of 238 U-234 Th and 232 Th-228 Th to POC of 0.09 Bq/mmolC with a range of 0.045 to 0.18, which was consistent with the ratio of 0.16 for RA P234 to POC (Fig. 3). The linear regression between RA P234 and POC also gave a clue to sediment resuspension.
Conceptual model of RA P234 . The conceptual model of RA P234 is illustrated in Fig. 4. Biogenic and terrigenous particles make the dominant contributions to suspended particles in the upper ocean and bottom nepheloid layer, respectively. Both kinds of particles can adsorb high particle-reactive radionuclides onto particle surfaces. In seawater, the dominant surface-bound radionuclide is 234 Th. The external and unsupported 234 Th adsorbed on biogenic and terrigenous particles decays away after 120 days and should not contribute to RA P234 . Radionuclides with high atomic number are seldom taken up by biotas as essential elements. Thus, biogenic particles play a minor role in RA P234 . The internal radionuclides of terrigenous particles, dominated by 234 Th supported by 238 U and 212 Bi supported by 228 Th, still exist and contribute to the second β counting rate of particulate 234 Th after 120 days due to the long half-lives of 238 U (4.47 × 10 9 y) and 228 Th-232 Th (1.91y and 1.4 × 10 10 y) in the minerals. Both 228 Th and 238 U have been categorised as lithophile elements according to Goldschmidt's classification 44 Table 3. Typical activities of radionuclides in marine biotas (in Bq/kg-f.w.) 59,60 . a "ND" denotes no data.
which is analogous to aluminium, titanium and other lithophile elements to trace the terrigenous fraction 45 . Although 228 Th and 238 U were not measured directly in our study due to the limitation of seawater volume, both 228 Th and 238 U in resuspended particles had been directly measured and attributed to sediment resuspension in other studies 29,31 . On the continental shelf, low activity of RA P234 in the upper layer and high value in the deep layer at SR3 can be interpreted as dominant photosynthesis and sediment resuspension, respectively. In comparison, low activity of RA P234 remained stable considering of its activity uncertainty at SR15 in the open ocean (Fig. 1a), while a peak value of POC was observed in the subsurface layer at a depth of 47 m (Appendix Table A1, 3.57 mmolC/m 3 at SR15). Subsurface chlorophyll maximum had been widely observed in the Arctic Ocean due to the supplement of nutrients in the subsurface layer 46 . Although POC was variable due to heterogeneous photosynthesis, RA P234 was vertically uniform as a result of small contributions to RA P234 from biogenic particles.
Consequently, RA P234 refers to the terrigenous particles resuspended from marine sediment, which is probable to trace sediment resuspension with sufficient sensitivity via β counter. RA P234 could be a nice addition to the 234 Th/ 238 U disequilibrium and seawater turbidity methods to distinguish particle processes related to photosynthesis and sediment resuspension.
Advantages of RA P234 . The relationship between RA P234 and POC was utilized to distinguish particle processes, including photosynthesis and sediment resuspension, in the western Arctic Ocean (Fig. 3). Biogenic  Scientific RepoRts | 6:27069 | DOI: 10.1038/srep27069 particle were characterized by low RA P234 in addition to variable concentrations of POC, which depended on intensity of photosynthesis. In comparison, sediment resuspension can elevate RA P234 . Therefore, sediment resuspension and photosynthesis could be distinguished with distinct RA P234 , while seawater turbidity and 234 Th/ 238 U disequilibrium method could not differentiate sediment resuspension from photosynthesis. Additionally, the slope of linear regression between RA P234 and POC, 'slope assumption' , has the potential to indicate the intensity of sediment resuspension (Fig. 3).
The linear regression between activity ratio of 234 Th to 238 U and POC (Fig. 5) was compared with that of RA P234 and POC (Fig. 3). The correlation coefficient of RA P234 and POC (0.815) is greater than that of 234 Th/ 238 U and POC (0.44). Both sediment resuspension and photosynthesis can enhance the scavenging of 234 Th. It is difficult to distinguish these two processes via 234 Th/ 238 U method. However, RA P234 is directly related to the terrigenous fraction from sediment resuspension based on the conceptual model (Fig. 4). Additionally, the 234 Th/ 238 U disequilibrium method has a memory effect that records the integrated particle dynamics during the past several months 47 . Both RA P234 and POC are instantaneous parameters relative to the parameters of 234 Th/ 238 U disequilibrium method with memory effects. Therefore, a better regression result for RA P234 and POC was obtained compared with the 234 Th/ 238 U disequilibrium method. RA P234 and its implications for export flux of 234 Th. The 234 Th/ 238 U disequilibrium method reflects the integrated particle dynamics, including sediment resuspension and photosynthesis, on the shallow continental shelf. Sediment resuspension can enhance the scavenging of 234 Th and deficit of 234 Th to 238 U, overestimating export flux of 234 Th. From the conceptual model of RA P234 , high activity of RA P234 was directly related to sediment resuspension. Sediment resuspension can be qualitatively identified on the basis of RA P234 to screen out the layer in which sediment resuspension occurred when export flux of 234 Th was integrated into the shallow water column. However, export flux of 234 Th may be underestimated following screening when photosynthesis occurs in conjunction with sediment resuspension.
Two endmembers, biogenic particles and resuspended particles, are assumed to exist in bottom seawater. The surface-bound concentrations of 234 Th on biogenic and resuspended particles were assumed to be f 1 and f 2 ,  respectively, in order to estimate the export fluxes of 234 Th from these two kinds of particles. The exact values of f 1 and f 2 were determined by two factors: particle concentration and the adsorbing ability of the particles. Most of the time, the particle concentration could be quantified by chemical proxies with distinct values for biogenic and resuspended particles, such as δ 13 C, Al, Ti and others. Biogenic and resuspended particles have low and high activity of RA P234 , respectively. Therefore, there is a potential to quantify the concentrations of biogenic and resuspended particles by mean of RA P234 .
The adsorbing capacity of distinct particle compositions can be quantified by different values of K d for thorium [48][49][50] . The particle compositions include lithogenic particles, opal, carbonate carbon, organic carbon, etc. If the K d for thorium can be obtained for biogenic and resuspended particles, f 1 and f 2 can be calculated (Eqs 1 and 2).
where a and b represent the fraction of biogenic and resuspended particles derived from chemical proxies. TSP is the total suspended particles in the seawater (mg/L). K d−bio. and K d−res. are the particle-seawater distribution coefficients for biogenic and resuspended particles (L/kg), and A D234 is the dissolved activity of 234 Th in seawater (Bq/m 3 ). If f 1 and f 2 (Bq/m 3 ) can be calculated, the export flux of 234 Th (F 234−bio. ) derived from the biogenic process can be obtained:  Therefore, the export flux of 234 Th (F 234−bio. ) derived from biogenic process can be determined from the fraction of biogenic particles and K d . In natural seawater, the fraction of biogenic particles, along with its uncertainty, can be quantified by chemical proxy. Large uncertainties in the fraction of particles occur, because the chemical proxies for endmember are generally difficult to identify. Although the K d of thorium for distinct particle compositions had been derived under the laboratory conditions 48-50 , K d−Bio. and K d−Res. are difficult to obtain in natural seawater, especially when complex particle compositions co-occur in biogenic and resuspended particles. The accuracy of the particle fraction and K d will constrain the exact estimation for export fluxes of 234 Th derived from biogenic particles. RA P234 : a linkage of the atmosphere-ocean-sediment system. To validate the 'slope assumption' , the A transect was revisited to investigate RA P234 and POC in the South China Sea during spring and autumn. The slope of linear regression between RA P234 and POC in Fig. 6 was also greater in autumn (0.30) than in spring (0.19), which may be attributed to sediment resuspension. Sediment resuspension could increase the terrigenous fraction and elevate RA P234 . The intensity of sediment resuspension had been indicated to be high in the same sea region in autumn relative to spring via the ratio of particulate 234 Th to dissolved 234 Th under the influence of the East Asian monsoon system 8 . Therefore, the assumption of the slope of linear regression between RA P234 and POC is confirmed in the South China Sea. RA P234 will shed new light on 234 Th-based particle dynamics to investigate the linkage of the atmosphere-ocean-sediment system, such as the typhoons and their impacts on sediment.
A novel approach of RA P234 is proposed for the first time to trace sediment resuspension from low-to high-latitude oceans. High activity of RA P234 was widely observed on the continental shelf in relation to sediment resuspension (Fig. 1). Sediment resuspension was also corroborated by seawater turbidity, residence time of total 234 Th, Goldschmidt's classification, and fingerprint ratio of RA P234 to POC from the sediment endmember in the western Arctic Ocean. The mechanism and conceptual model of RA P234 was investigated and illustrated (Fig. 4.). RA P234 is sufficiently sensitive to identify sediment resuspension via β counter with high detector efficiency. The advantage of RA P234 is that it is a supplementary parameter to the 234 Th/ 238 U disequilibrium method and does not require any additional sampling and measurement to distinguish sediment resuspension from photosynthesis, while both the 234 Th/ 238 U disequilibrium and seawater turbidity methods cannot differentiate biogenic particles from terrigenous particles. RA P234 is a potential proxy to trace sediment resuspension without a memory effect. RA P234 could also be used to screen out the layer to bias integration of 234 Th and POC fluxes. The slope of the linear regression between RA P234 and POC was used to indicate the higher intensity of sediment resuspension in the South China Sea during autumn. Similar to the definition of gross β , RA P234 may stimulate some debate but is also meaningful to identify and indicate the intensity of sediment resuspension. From the mechanism proposed, RA P234 refers to the terrigenous fraction and has potentially broad implications for investigating the dynamics of suspended particles in the estuary-shelf-slope-ocean continuum and the linkage of the atmosphere-ocean-sediment system.

Methods
Sampling. Seawater samples were collected for 234 Th analysis from low-to high-latitude ocean in the western Arctic Ocean, the South China Sea, and the Southern Ocean (Fig. 7). Seven stations (SR1, SR3, SR5, SR7, SR9, SR12, SR15) were sampled in the western Arctic Ocean during the 5th Chinese National Arctic Research Expedition (CHINARE-5) in September, 2012 (Fig. 7b.). The sea ice extent during the sampling period was the lowest since the first satellite measurement taken in 1979 51 . Seawater turbidity was measured and was indicated by red stars on the continental shelf (SR1, SR3, SR5, SR7, SR9) and the open ocean (R10) (Fig. 7b).
A Analysis of 234 Th. The 234 Th/ 238 U disequilibrium method has been widely applied in the global ocean with a huge database to quantify the marine biological carbon pump 52 , which modulates glacial/interglacial atmospheric carbon dioxide and climate change 53 . The international calibration of 234 Th was conducted under the framework of GEOTRACES 54 . The small-volume technique via β counting of 234 Th has been extensively studied due to its high sampling resolution 13,23 . The radiochemical analysis of 234 Th had been described 8,37 .
Following filtration of seawater with 25-mm diameter Quartz Microfiber (QMA, nominal pore size 1.0 μ m), the direct measurement of particulate 234 Th without radiochemical separation was obtained from the difference in values between the first β counting after sampling and the second β counting after 120 days as a result of high activity of 234 Th in seawater 19 . After treating with MnO 2 co-precipitation, the activity of total 234 Th was also calculated from the difference between the first and second β counting rates of total 234 Th. The activity of 234 Th and its associated uncertainty were calculated according to Eqs 4-9.   The dimensions and definitions of the parameters are given in Table 4. Equations 4-9 had been deduced in detail with the similar principle 55 . Definition and calculation of RA P234 . The second counting rate of particulate 234 Th (n P2 ) was usually overlooked, because only the difference between the first and second counting rates (n P1 -n P2 ) was used to calculate particulate 234 Th using Eq. 4. In the open ocean, the second counting rate of particulate 234 Th (n P2 ) was relatively stable with a value of 0.3~0.4 cpm, which also depends on the instrumental background with a normal value of 0.15~0.2 cpm via gas-flow proportional low-level RISØ β -counter (Model GM-25-5, RISØ National Laboratory, Denmark) [12][13][14] . In this study, the abnormally high second counting rate of particulate 234 Th was observed for bottom seawater on the continental shelf in the western Arctic Ocean. This phenomenon was further confirmed in the South China Sea and the Southern Ocean. RA P234 derived from the second counting rate of particulate 234 Th and instrumental background was proposed for the first time to investigate this abnormal value of particulate 234 Th. Equations 10 and 11 were used to calculate RA P234 and its uncertainty: All the parameters in Eqs 10 and 11 are defined in Table 4. The detector efficiency of RA P234 is equal to that of particulate 234 Th because of the similar energy of β particles being emitting by radionuclide candidates. The second counting rate of particulate 234 Th (0.54 ± 0.02 cpm) was very constant after 136 days, 304 days, and 495 days from the sampling date, which indicates that it was mainly radionuclides with long half-lives that contributed to RA P234 . The stability of the second counting rate of particulate 234 Th has been demonstrated 24 .
Notice that RA P234 was not a signal from a certain radionuclide. In fact, RA P234 was the residual β activity for particulate 234 Th after more than 120 days, which was usually recognized to be the stable methodological background for particulate 234 Th and was therefore neglected. This residual β activity may include several radionuclides with long half-lives. It should be treated as a supplementary parameter for total 234 Th and particulate 234 Th and has advantage of being able to trace sediment resuspension without any additional sampling and analysis based on the small-volume technique for 234 Th.
The definition of RA P234 is similar to that of gross β in drinking water. Most of the time, the exact radionuclides and their contributions to gross β cannot be identified 11 . However, gross β is an essential parameter for screening the level of radiological pollution, especially during nuclear emergency. The detector efficiencies of 90 Sr and 137 Cs are artificially chosen to calculate that of gross β for drinking water, although a spread of energies of β particles from distinct radionuclides ( 40 K) with different detector efficiencies is very common 11 . Analogously, the definition of RA P234 is proposed and is convenient for tracing sediment resuspension without any additional sampling and measurement.
Although the abnormally high second counting rate of total 234 Th was also observed for the bottom seawater on the continental shelf as that of particulate 234 Th, the second counting rate of total 234 Th was not discussed in this study. The radionuclides contributing to the second counting rate of total 234 Th are more complex than that of particulate 234 Th due to the additional MnO 2 co-precipitation. The radiochemical treatment of MnO 2 co-precipitation for total 234 Th can scavenge other radionuclides of low K d with variable chemical recovery, such as radium and its progenies, onto the MnO 2 -particle surface 13,56 , especially for the tectonically active sea region with 224 Ra diffusion into the overlying seawater 25  Particulate organic carbon. Following the second counting of particulate 234 Th, the POC was measured with an Elemental Analyzer (Elementar vario EL III) after removing the carbonate fraction by fuming with concentrated hydrochloric acid 57 . The blank of the method was subtracted. The analytical precision was always better than 10%.
Seawater turbidity. The seawater turbidity was measured using a turbidity sensor (Rinko-profiler) during the 6th CHINARE from 27 July to 7 August 2014. The turbidity sensor works on the basis of backscattering principle and has a range of 0~1 FTU. The reference material was Formazin. A few abnormal values over 1 FTU arising from the present of bubbles were discarded. The precision of the turbidity sensor was 0.03 FTU.