Dear Reviewer 1,

Thanks for the detailed comments, which really helped to improve the MOREPOC dataset and the manuscript. Please check our reply to the comments below:

1. Dataset

We have made several modifications to MOREPOC v1.0, and now version 1.1 is available on Zenodo. Each data entry is an individually collected sample unless several size fractions were reported in the original study.

To summarize:

MOREPOC V1.1

• Renamed field header. The names of some fields were changed to make them more intuitive, e.g., "riv_na" was changed to "bas-id" to refer to the sampled drainage basin and "bas_na" was changed to "riv_id" to refer the name of the sampled river. “RCA” was changed to “age_14C”, and “cn_mar” and “as_mar” were changed to “cn_ratio” and “alsi_ratio”, respectively.
• Added extra fields. The “country” column was created to separate it from “riv_id”; “perc_poc_1sd”, ”d13C_1sd”, and ” D14C_1sd” columns were added to provide analytical uncertainties for POC content, δ¹³C and Δ¹⁴C data, respectively.
• Corrections made. There were some wrong entries about the country where samples were collected in the “country” column, which has been corrected. The content of the “time_m” column was changed from strings to numeric. In addition, we removed brackets, points, commas or empty spaces in the data fields. Coordinates were carefully checked through, and wrong entries were corrected. The reversed latitude and longitude of the one sample (Amazon River at Obidos, 2005) were corrected. Only the average is now presented in Figure 1 if several samples exist from a given location.
• Added data. We incorporated data from some recently published or missed articles, e.g., Menges et al., 2020 and Holmes et al., 2022. Besides, we looked through all references from MOREPOC v1.1 to add analytical uncertainties if available. However, most papers do not report uncertainties. We also added more sampling date information, such as those from Bouchez et al., (2014). Now there are only 3 samples without associated dates. However, some data entries only have the sampling date information as a period (year) because there is no specific date reported for each data entry, such as for a large amount of data from Taiwan rivers (Hilton et al., 2010).

MOREPOC V1.1_RM

• Reorganized references. References not cited in MOREPOC v1.1 were removed, and newly incorporated references were added.
• Added extra fields. The “para_m” and “para_c” columns were added to clarify the sources of parameters in MOREPOC v1.1. Data entries of some parameters were directly taken from the cited references, while some were obtained through later calculations and conversions.
• Completed information. Some missing information in “filter” and “Note” were added.

Besides, a Readme.txt was provided with the MOREPOC v1.1 dataset, also available on Zenodo.

1. Manuscript text

Line 7-8: Rephrased.

Line 31: Corrected.

Line 45: We agree that SOC in permafrost has a dynamical turnover time subjected to varying climate conditions, especially the current elevation of ambient temperature will accelerate the thawing of permafrost. In the manuscript, we now explain the term "turnover time" following Eglinton et al. (2021), which is the ratio of soil carbon stock to input flux, to make it more proper to discuss the relative long SOC turnover time in permafrost regions.

Line 47: Carvalhais et al. (2014) was removed from the Introduction.

Line 53: Replaced.

Line 66: Rephrased.

Line 81: We agree that “carbonate removal” is more appropriate and is now used to replace “decarbonization” in the manuscript. “acid adopted” means the type of acid used for removing carbonate - this is clarified now.

Lines 134-136: Time is in units of hours and temperature is in units of Celsius degrees. We added this information in the manuscript to explain the parameter in MOREPOC v1.1_RM.

Line 142: Rephrased.

Line 146: Corrected.

Lines 142-154: We agree that the definition of F¹⁴C should be in included in the manuscript - the equation was added.

Line 170: Corrected.

Line 259: Good suggestion, we rephrased the expression to “MOREPOC will benefit the scientific community carrying out research on riverine POC sources, transport, and fate, furthermore, helping inform and validate Earth system models to improve the ability to model and understand the global carbon cycle.”

Dear Reviewer 2,

Thanks for the detailed comments and recommendations. We hope MOREPOC can benefit the Earth Science community, in particular, researchers interested in understanding OC-related Earth surface processes and carbon cycling at different time- and spatial scales. Now the compilation includes more data entries with 265 more radiocarbon activity (Δ¹⁴C) values, providing exciting observations.

We provide our reply to your comments below.

Data Comments

• A readme file with information on variables and units was added to the database and is now available on Zenodo too. Equations for parameter definition and conversion were included.
• These variable names were changed to a more intuitive term ("riv_na" to "bas_id", "bas_na" to "riv_id").
• Uncertainties were added for elemental and isotopic carbon values if available. “perc_poc_1sd”, ”d13C_1sd”, and ” D14C_1sd” columns were added to provide analytical uncertainties for POC content, δ¹³C and Δ¹⁴C data, respectively.
• We added extra fields to clarify which variable was given in which paper. The columns “para_m” and “para_c” were added to clarify the sources of parameters in MOREPOC v1.1: data entries of some parameters were directly taken from the cited references, while some were obtained through later calculations and conversions.
• As explained above, we added two columns to clarify the sources of the compiled data.

For references only reporting POC wt.% and concentration, we did use the two parameters to derive SPM concentration. Parameters obtained from this conversion were clarified.

• Format problems: We did not change “modern” to “0”, since some high ¹⁴C activity data are because of the influence of the bomb carbon effect. As a consequence, we would like to have users decide how to use these values themselves. However, we changed the field name as suggested. For ‘values’ in data fields, we converted strings to a numeric format, and eliminated the use of brackets, points, commas, or empty spaces.
• Country information is now included in a new column “country”.

Article Comments

L21: We added a sentence on the role of alteration and degradation on fluvial POC goes during transport in terrestrial environment before entering the coastal environment and being buried in the ocean.

L24: We clarified this statement about the composition of biospheric OC. Riverine authigenic POC is now mentioned to partly explain the depleted ¹³C signals in riverine POC (section 3.1).

L32: We now refer to the global POC_bio burial in the ocean, as well as to the oxidation of POC_petro. The terrestrial POC_bio burial can be up to around 70 MtC/yr considering an average burial efficiency of 30% to an input of ~110-230 MtC/yr (Blair and Aller, 2012; Burdige, 2005; Galy et al, 2015), while the oxidation of POC_petro in sedimentary rocks can contribute ~40-100 MtC/yr to atmospheric CO₂ (Petsch, 2014; Hilton and West, 2020).

L35: Rephrased.

L47-48: We highlighted that the input of aged biospheric OC from thawing permafrost is the major reason (Wild et al., 2019; Hilton et al., 2015). Besides, we extended the discussion on permafrost-derived fluvial POC in section 3 (in particular section 3.2) and figures 6 and 7.

L50-51: We agree that sediment dynamics and oxygen availability in marine environments are important factors (Blair and Aller, 2012). However, in this paragraph, we want to discuss how POC can be altered in terrestrial settings before entering the marine environment. We do emphasize the importance of sediment dynamics in the terrestrial environment due to the different tectonic settings.

L54: Reference added. Besides, we also added some clarification on how human activities influence erosion and on the resulting changes in the fluxes of sediments and associated POC.

L61-62: We added a sentence on the improvement of water quality datasets and increasingly sophisticated models of riverine carbon cycling as you suggested.

L65: We added more explanation for D14C and Fm14C in section 2.7.

L79: This is based on non-numeric string detection in value fields, categorical summarization, and extreme numbers detection. We were wrong here to use the term – statistical examinations, so we deleted this expression.

L155: We added explanations on why Al/Si ratio is an important parameter that is included in MOREPOC as follows:

"Lastly, if available, the aluminum-to-silicon mass ratio (Al/Si) is also provided in MOREPOC v1.1. This elemental ratio is an efficient proxy for the particle size of riverine sediment, allowing to characterize the particle size effect of sediments on POC loading in fluvial delivery (Galy et al., 2008b; Bouchez et al., 2011; Hilton et al., 2015). The mineralogy and particle size of sediments are generally not totally independent of each other, coarse particles tend to be quartz-rich (low Al/Si ratios) and fine particles tend to be clay-rich (high Al/Si ratios) (Galy et al., 2008b). POC contents are usually positively correlated with proportions of fine-grained fractions (Mayer, 1994; Galy et al., 2008b; Bouchez et al., 2014)."

L158: We added this information “The MOREPOC database also indicates the lack of study on POC in fluvial systems in high-latitude regions such as the Antarctic and Greenland, as well as in arid regions including Australia or vast areas spanning from northern Africa to middle east Asia (Figure 1).”

L178-179: We added this information: “Around the Qinghai-Tibet Plateau, where most large river systems in eastern and southern Asia share similar high-elevation headwaters, POC is usually characterized by relatively depleted ¹⁴C signals due to strong erosion of sedimentary rocks, such as the Ganges-Brahmaputra (Galy et al., 2007) or the Changjiang (Wang et al., 2012; Wang et al., 2019), and to the erosion of soil, pre-aged OC, e.g. the Huanghe (Tao et al., 2015).”

 L197: Net primary production does not show any clear relation with ¹⁴C in fluvial POC nor with the relative abundance of biospheric OC. It is rather the erosion rate of the catchment which controls the flux of biospheric OC (Galy et al., 2015). We added additional explanations for this trend: “This might reflect the existence of major POC components: 1) for rivers dominated by POC_bio, the combined effects of increasing coverage of C4 plants towards tropical regions and the input of pre-aged OC_bio from C3-derived OC from degrading permafrost at high latitudes (Cerling et al., 1997; Still et al., 2003); 2) for rivers dominated by POC_petro, for example in mountainous regions, strong erosion of ¹³C-enriched petrogenic OC (Hilton et al., 2010; Galy et al., 2007). In addition, in-river ("authigenic") POC production can be an important mechanism contributing ¹³C-depleted and ¹⁴C-enriched POC (Longworth et al., 2007; Marwick et al., 2015; Wu et al., 2018).”

Figure 2: We did not add different potential endmembers because we want users to interpret possible sources for fluvial POC.

L211: Potential explanations for this observation were added.

L216: This occurrence was not about carbon loading, we are sorry about the confusion caused. We were talking about the POC flux that is loaded in a specific SPM concentration. It is now clarified in the text.

L219: We added a figure as suggested, providing POC concentration vertical variation in the water column as obtained from depth profiles in large rivers. Selected depth profiles are from the Yukon (Holmes et al., 2022), Mackenzie including Peel and Arctic Red (Hilton et al., 2015), Amazon (Bouchez et al., 2014), and Ganges-Brahmaputra-Meghna systems (Galy et al., 2007, 2008b).

L221: We found this sentence was not correct under certain circumstances, so we reworded it to “Small SPM concentrations (less than 10 mg/L) are generally found in rivers in frozen seasons or rivers draining either high-latitude or tropical areas characterized by low-relief settings, in which POC content is relatively high (Gao et al., 2007; Holmes et al., 2022).”

L235: Added.

L244: Good point, we now refer to petrogenic OC mobilization. However, this paragraph is focused on the erosion of sediments and aims to explain the depleted ¹⁴C nature of fluvial POC observed in some conditions. As a consequence, the role of oxidation of petrogenic OC is not mentioned in the text.

Dear Reviewer 3,

Thanks for your comments, which helped to improve the dataset. MOREPOC v1.1 is now available on Zenodo. This refined version of the database now includes 3,546 SPM data entries, among which 3,053 with POC content, 3,402 with stable carbon isotope (δ¹³C) values, 2,283 with radiocarbon activity (Δ¹⁴C) values, 1,936 with total nitrogen content. This represents a significant update compared to MOREPOC v1.0.

Our replies to your comments are as follows:

Line 26, Line 27: We rephrased this sentence in more appropriate terms: "Land plants, soils, aquatic organisms, and microbes can all contribute radiocarbon-active POC_bio to riverine POC, with ages ranging from modern to multi-millennial (Galy et al 2007; Blair et al., 2010; Hilton et al., 2011)".

Line 35: The sentence was rephrased: "The terrestrial POC_bio burial can be up to around 70 MtC/yr considering an average burial efficiency of 30% to an input of ~110-230 MtC/yr (Blair and Aller, 2012; Burdige, 2005; Galy et al, 2015), while the oxidation of POC_petro in sedimentary rocks can contribute ~40-100 MtC/yr to atmospheric CO₂ (Petsch, 2014; Hilton and West, 2020). These fluxes are comparable to those induced by silicate weathering, carbonate weathering by oxidation of sulfides and volcanism, demonstrating that POC could play an important role in the Earth’s long term carbon cycle (Berner, 2003; Hilton et al., 2014; Petsch, 2014; Galy et al., 2007; Galy and Eglinton, 2011; Hilton and West, 2020)."

Introduction: we added the definition of POC to the text: "POC is the fraction of total organic carbon that remains on a filter of a given mesh size".

Line 84: We were referring to the coordinate system used when digitalizing MOREPOC data entries in a shapefile layer in ArcGIS 10.3, We rephrased this sentence: “The location of samples was digitalized if available, and an associate ArcGIS data layer in shapefile format (see MOREPOC_v1.1.rar) is provided with all points projected in a Geographic Coordinate System using the World Geodetic System 1984 (WGS1984).”

Section 2.4: This is definitely a good point that will need more work in the future. In such a wide compilation, sampling locations correspond to catchments of different scales, i.e., main channels vs. tributaries as well as sub-tributaries. However, it is challenging to summarize the information on whether samples were collected at a river mouth, along the flowing routing or the headwater. Nevertheless, we believe that by looking into the map of MOREPOC sampling points (such as the ArcGIS product provided with MOREPOC) the reader will be able to find that most sampling locations represent integrated signals of biogeochemical processes over a whole catchment.

Line 121: We changed to use mesh size.

Line 172: C3 plants make up over 95% of the global biomass, such that it would be hard to find significant C4 signals in global POC patterns. However we now explain in section 3.1: “However, it can also be observed that POC-rich riverine SPM can be relatively enriched in ¹³C, with δ¹³C values larger than -20‰ (Figure 2 and Figure 3). This pattern might indicate the presence of an additional pool of ¹⁴C- and ¹³C-rich POC in the terrestrial environment (Cerling et al., 1997), consisting of modern C4-plants in catchments dominated by grasslands or savannah (e.g., Marwick et al., 2015).”

General 3.1: δ¹³C values can be affected by the degradation of POC during fluvial transport, as shown by Mayorga et al. (2005), with the preferential degradation of young, labile biospheric OC resulting in an increase of δ¹³C values. However, such effect typically does not result in δ¹³C values outside the range of C3 plant OC. We added an explanation for the very ¹³C-depleted signals in fluvial POC in Section 3.1. Actually, these samples are mostly from permafrost-draining rivers; and secondly, aquatic authigenic OC production can be an important mechanism contributing ¹³C-depleted and ¹⁴C-enriched POC (Longworth et al., 2007; Marwick et al., 2015; Wu et al., 2018).

Line 257: Al/Si was not introduced properly in the previous version of the manuscript. We now explain in section 2.7 why MOREPOC features Al/Si data: “Lastly, if available, the aluminum-to-silicon mass ratio (Al/Si) is also provided in MOREPOC v1.1. This elemental ratio is an efficient proxy for the particle size of riverine sediment, allowing to characterize the grain size effect of sediments on POC loading in fluvial load (Galy et al., 2008b; Bouchez et al., 2011; Hilton et al., 2015). The mineralogy and particle size of sediments are generally related, with coarse particles being quartz-rich (low Al/Si ratios) and fine particles being clay-rich (high Al/Si ratios) (Galy et al., 2008b). POC contents are usually positively related to the fraction of fine grains in the sediment (Mayer, 1994; Galy et al., 2008b; Bouchez et al., 2014).

Dear Dr. Jin Wang,

Thanks for the comments. We have revised the manuscript accordingly.

MOREPOC v1.1 is an updated version of MOREPOC v1.0, both available on Zenodo. The number of data entries increases to an amount of 3,546 during revision, among which 3,053 with POC content, 3,402 with stable carbon isotope (δ¹³C) values, 2,283 with radiocarbon activity (Δ¹⁴C) values, 1,936 with total nitrogen content.

Our replies to your comments are as follows:

General comment:

1. We realized that there was only a poor introduction to the interest of the Al/Si ratio in the context of POC studies, a parameter included in MOREPOC v1.1. We added to section 2.7: “Lastly, if available, the aluminum-to-silicon mass ratio (Al/Si) is also provided in MOREPOC v1.0 1. This elemental ratio is an efficient proxy for the particle size of riverine sediment, allowing to characterize the particle size effect of sediments on POC loading in fluvial delivery (Galy et al., 2008b; Bouchez et al., 2011; Hilton et al., 2015). The mineralogy and particle size of sediments are generally related, with coarse particles being quartz-rich (low Al/Si ratios) and fine particles being clay-rich (high Al/Si ratios) (Galy et al., 2008b). POC contents are usually positively related to the fraction of fine grains in the sediment (Mayer, 1994; Galy et al., 2008b; Bouchez et al., 2014).”
2. This confusion has been resolved in the revised manuscript: in Line 210, we changed the term "POC concentration" to "POC content".
3. We added a color bar (for POC wt.%) in Figure. 4 to indicate the dilution of organic carbon by inorganic materials. The constant-POC contour lines drawn in the figure also provide the same information.

Specific comments in the text:

1. Line 47: We changed to “riverine POC_bio” in the text.
2. Line 78: In some references, data tables were indeed images, such that we had to use automatic conversion tools to transfer image data into tables. However, this may result in some numbers being converted into letters, which we had to carefully check. We should have not taken any data from figures that need to use Digitalizer to take the approximate number, we sent emails to corresponding authors to inquire data.
3. Line 129: Reference to this carbonate removal method was added to the manuscript, although it seems more common in soil OC studies than riverine POC studies. We also added the data from Menges et al. (2020) to MOREPOC v1.1 (section 2.6).
4. Line 201-205: We provided extra explanations to strengthen the statement: “This might reflect the major POC components: 1) dominated by POC_bio, the combined effects of increasing coverage of C4 plants in tropical regions and the input of pre-aged OC_bio of C3 plants from degrading permafrost at high latitude (Cerling et al., 1997; Still et al., 2003); 2) dominated by POC_petro, rivers in mountainous regions tend to erode ¹³C-enrich petrogenic OC (Hilton et al., 2010; Galy et al., 2007).”
5. Line 220: We added the suggested references.
6. Line 222: We realized that there was a problem with this sentence, which we reworded to “Small SPM concentrations (less than 10 mg/L) are generally found in rivers during cold seasons, or in rivers draining either high-latitude or tropical areas characterized by low-relief settings, in which POC content is relatively high (Gao et al., 2007; Holmes et al., 2022).”
7. Missing references were added to the Reference section.