I. Data Set Descriptors A. Title: Peter Hawman. 2025. Eddy covariance (EC) vertical carbon fluxes from a Georgia tidal salt marsh from 2014 to 2024. Georgia Coastal Ecosystems LTER Data Catalog (data set MSH-GCEM-2508; /data/MSH-GCEM-2508) B. Accession Number: MSH-GCEM-2508 C. Description 1. Originator(s): Name: Peter Hawman Address: Department of Geography Geography-Geology, Rm 319 Athens, Georgia 30602 Country: USA Email: peterhawman@uga.edu 2. Abstract: We present our methodology and data for science ready vertical carbon fluxes from a Spartina alterniflora tidal salt marsh as part of the Georgia Coastal Ecosystems Long Term Ecological Research (GCE-LTER) site on Sapelo Island, Georgia, USA. Vertical carbon fluxes were measured through the eddy covariance (EC) method from 2014 to 2024. The EC flux tower was located on the western side of Sapelo Island bounded by the Duplin River and Barn Creek. The proportional influence of marsh habitats on the flux measurements were 4% tall, 38% short, and 58% medium height form Spartina alterniflora. We present the net ecosystem exchange (NEE), ecosystem respiration (ER), and gross primary production (GPP) at 30-minute fluxes (μmol CO2 m-2s-1), daily averages (μmol CO2 m-2s-1) and totals (g C m-2 day-1), and annual (g C m-2 year-1) quantities. We provide estimated uncertainty for each flux at each integrated timescale as 95% confidence intervals. Providing open access to 10-year carbon flux datasets can facilitate collaboration for advancing regional and global blue carbon synthesis and scale-up studies. 3. Study Type: Monitoring 4. Study Themes: Marsh Ecology, Plant Ecology 5. LTER Core Areas: Primary Production 6. Georeferences: none 7. Submission Date: Aug 15, 2025 D. Keywords: carbon assimilation, carbon cycling, carbon dioxide, carbon fluxes, Duplin River, ecology, ecosystems, GCE, Georgia, Georgia Coastal Ecosystems, LTER, marshes, photosynthesis, Plant Monitoring, Primary Production, Sapelo Island, USA, vegetation II. Research Origin Descriptors A. Overall Project Description 1. Project Title: Georgia Coastal Ecosystems LTER - IV 2. Principal Investigators: Name: Merryl Alber Address: Dept. of Marine Sciences University of Georgia Athens, Georgia 30602-3636 Country: USA Email: malber@uga.edu 3. Funding Period: Feb 01, 2019 to Jan 31, 2025 4. Objectives: The GCE-LTER project has four goals. 1) Track environmental and human drivers that can cause perturbations in our focal ecosystems. This will be accomplished this through continuing long-term measurements of climate, water chemistry, oceanic exchange, and human activities on the landscape. 2) Describe temporal and spatial variability in physical, chemical, geological and biological characteristics of the study system (coastal wetland complexes) and how they respond to external drivers. This will be accomplished through field monitoring in combination with remote sensing and modeling. 3) Characterize the ecological responses of intertidal marshes to disturbance. This will be accomplished by ongoing monitoring and experimental work to evaluate system responses to major perturbations in three key marsh habitats (changes in inundation and predator exclusion in Spartina-dominated salt marshes; increases in salinity in fresh marshes; changes in runoff in high marshes), by implementing standardized experimental disturbances along salinity and elevation gradients, and by tracking responses to natural disturbances. 4) Evaluate ecosystem properties at the landscape level (habitat distribution, net and gross primary production, C budgets) and assess the cumulative effects of disturbance on these properties. The project will also develop relationships between drivers and response variables, which can be used to predict the effects of future changes. This will be accomplished through a combination of data synthesis, remote sensing and modeling. 5. Abstract: The Georgia Coastal Ecosystems (GCE) Long Term Ecological Research (LTER) program, based at the University of Georgia Marine Institute on Sapelo Island, Georgia, was established in 2000 to study long-term change in coastal ecosystems. Estuaries (places where salt water from the ocean mixes with fresh water from the land) and their adjacent marshes provide food and refuge for fish, shellfish and birds; protect the shoreline from storms; help to keep the water clean; and store carbon. The GCE LTER researchers study marshes and estuaries to understand how these ecosystems function, to track how they change over time, and to predict how they might be affected by future changes in climate and human activities. They accomplish this by tracking the major factors that can cause long-term change in coastal areas (e.g. sea level, rainfall, upstream development), and measuring the effects of these factors on the study site. They also conduct focused studies to assess how key marsh habitats will respond to major changes expected in the future, including large-scale experiments to evaluate the effects of a) increases in the salinity of the water that floods freshwater marshes (mimicking drought and/or sea level rise), b) changes in water runoff from land into the upland marsh border (mimicking drought or upland development), and c) exclusion of larger organisms in the salt marsh (mimicking long-term declines in predators). During this award they will initiate additional studies to systematically evaluate how coastal wetlands respond to disturbances. Disturbances, or disruptions in the environment, are particularly important to understand in the context of long-term background changes such as increasing sea level, and GCE researchers are working to assess the cumulative effects of multiple disturbances on the landscape. The GCE education and outreach program works to share an understanding of coastal ecosystems with teachers and students, coastal managers, citizen scientist and the general public. 6. Funding Source: NSF OCE 1832178 B. Sub-project Description 1. Site Description a. Geographic Location: GCE_Flux -- GCE Flux Tower Marsh, Sapelo Island, Georgia Flux_Tower -- GCE Flux Tower Coordinates: GCE_Flux -- NW: 081 17 19.06 W, 31 27 23.72 N NE: 081 16 21.23 W, 31 27 23.72 N SE: 081 16 21.23 W, 31 26 12.12 N SW: 081 17 19.06 W, 31 26 12.12 N Flux_Tower -- 81 17 00.5 W, 31 26 38.7 N b. Physiographic Region: GCE_Flux -- Lower coastal plain Flux_Tower -- Lower coastal plain c. Landform Components: GCE_Flux -- Intertidal salt marsh bordering maritime forest Flux_Tower -- Intertidal salt marsh bordering maritime forest d. Hydrographic Characteristics: GCE_Flux -- Site is along the Duplin River, and bounded by Barn Creek on the south and east, and is subject to 2-3m semi-diurnal tides Flux_Tower -- Site is along the Duplin River, and bounded by Barn Creek on the south and east, and is subject to 2-3m semi-diurnal tides e. Topographic Attributes: GCE_Flux -- Flat, with elevations ranging from 0-3m above mean low tide Flux_Tower -- Flat, with elevations ranging from 0-3m above mean low tide f. Geology, Lithology and Soils: GCE_Flux -- unspecified Flux_Tower -- unspecified g. Vegetation Communities: GCE_Flux -- Dominated by Spartina alterniflora Flux_Tower -- Dominated by Spartina alterniflora h. History of Land Use and Disturbance: none recorded i. Climate: Climate summary for Sapelo Island, Georgia, based on NWS data from 1980-2010: Daily-aggregated Values: Mean (sample standard deviation) mean air temperature: 20.09°C (7.28°C) minimum air temperature: 15.02°C (7.96°C) maximum air temperature: 24.82°C (6.98°C) total precipitation: 3.26mm (10.3mm) Yearly-aggregated Daily Values: Mean (sample standard deviation) total precipitation (1980-2010): 1124mm (266mm) 2. Experimental or Sampling Design a. Design Characteristics: Long-term measurements were collected from a tower at a fixed location in a Spartina alterniflora marsh in coastal Georgia. b. Permanent Plots: not specified c. Data Collection Duration and Frequency: Measurements were automatically logged at 10 Hz intervals using a Campbell Scientific Instruments CR3000 data logger. Beginning of Observations: Jan 01, 2014 End of Observations: Dec 31, 2024 3. Research Methods a. Field and Laboratory Methods: Method 1: Eddy Covariance Raw 10 Hz data collection -- Raw turbulence data were collected at 10 Hz using an Eddy Covariance (EC) system consisting of a LiCOR LI-7200 infrared gas analyzer and CSAT3 Sonic Anemometer. The 10 Hz data were processed to 30-min fluxes using EddyPro 7 (LI-COR Biosciences, Lincoln, NE, USA). We adjusted for changing vertical separation between the EC system and the marsh surface from daily tidal flooding based on RTK-GPS-corrected water levels and seasonal vegetation growth, which would directly affect roughness layer thickness, by using dynamic metadata in EddyPro 7. Method 2: Eddy Covariance Raw 10 Hz turbulance processing to 30 minute CO2 fluxes -- We performed double rotation and linear detrending on sonic anemometer data using the method described in Wilczak et al. (2001) and the Webb-Pearman-Leuning (WPL) correction to correct densities of CO2 due to the presence of heat and H2O fluxes using the method from Webb et al. (1980). We followed the quality control checks suggested by Mauder and Foken (2006), performed statistical analysis following Vickers and Mahrt (1997), and estimated random uncertainty as suggested by Finkelstein and Sims (2001). Spectral analyses were completed by filtering co-spectra following Vickers and Mahrt (1997) and quality tests following Mauder and Foken (2006). We finally performed low frequency (Moncrieff et al., 2005) and high frequency (Moncrieff et al., 1997) spectral corrections. we added the storage terms to CO2 fluxes to derive net ecosystem exchange (NEE). Then we removed data during maintenance. We set any data that was flagged with a QC value of “2” (Mauder and Foken 2006) to “NA”. We also performed a visual inspection of data, removing obvious error periods, and periods when the CO2 mixing ratio was implausible. Next, we set any nighttime negative NEE values to NA, as these values were questionable. We then filtered extreme values, removing NEE measurements that were larger than 15 μmol CO2 m-2 s-1 or smaller than -20 μmol CO2 m-2 s-1. Finally, using the R package REddyProc (Wutzler et al. 2018), we performed friction velocity (u*) filtering by first determining seasonal thresholds using bootstrapping and then removing data where u* values were too low. Method 3: Eddy Covariance Dual Sensor Integration -- During the years 2014 to 2017, there were two identical sets of IRGA and anemometers deployed together on the tower to reduce data gaps. We used data from each sensor set when the other was missing. When both sensors provided quality measurements, we averaged the two flux values only if their standard deviation was less than 5 μmol CO2 m-2 s-1. Method 4: Eddy Covariance Gap-filling Net Ecosystem Exchange -- 30‐min NEE were gap‐filled using extreme gradient boosting (XGBoost) (Chen and Guestrin 2016), a machine learning algorithm. NEE data were filtered to highest quality fluxes (qc = 0) (Mauder and Foken 2006) and split into training data (n = 37,877), for model fitting and testing, and holdout (n = 12,628) data, for final model assessment. For model predictors, please see Hawman et al. (2024). We tuned XGBoost models using nested resampling (20 resamples) and K‐fold cross‐validation (K = 5). When splitting data into training and holdout data sets and during K‐fold cross‐validation, data were proportionally sampled based on K‐Mean clusters using fuzzy variables representing time of day (morning, afternoon, evening, and night) and season (spring, summer, autumn, and winter) (Papale and Valentini 2003; Knox et al. 2015). We assessed model performance by predicting on the holdout data and using the root mean squared error (RMSE) and normalized RMSE (NRMSE) from the predicted and observed NEE. We also fit ordinary least squares regressions using the coefficient of determination (R2), slope, and intercept to assess model goodness of fit and bias. The resulting 20 XGBoost models were predicted on the entire timeseries. We used the median of the 20 model predictions to fill NEE gaps. The median of predictions from all gap-filling models on the holdout data was: RMSE = 1.72 ± 0.01 μmol CO2 m-2 s-1; NRMSE = 5 ± 0.03 %; R2 = 0.85 ± 0.002; slope = 0.84 ± 0.004. Method 5: Eddy Covariance Carbon Flux Partitioning -- After gap-filling NEE, we partitioned NEE into ecosystem respiration (ER) and gross primary production (GPP) (-NEE = ER - GPP). We used a modified nighttime approach (Reichstein et al. 2005) but fit XGBoost models to nighttime NEE data, representing ER, and then applied those models to daytime periods to predict daytime respiration. We used the same modeling framework as our gap-filling technique. The median of predictions from all partitioning models on the holdout data was: RMSE = 0.82 ± 0.003 μmol CO2 m-2 s-1; NRMSE = 11 ± 0.04%; R2 = 0.69 ± 0.003; slope = 0.68 ± 0.009. We then predicted ER for the entire timeseries and then subtracted NEE from the ER predictions during daytime periods to estimate GPP. We used the directional convention for NEE where negative values indicated a net CO2 uptake by the ecosystem and positive values a net source of CO2 to the atmosphere. GPP and ER are shown as both positive values i.e., no signage for direction). Finally, NEE, GPP, and ER were integrated to daily (g C m-2 day-1) and annual summations (g C m-2 year-1). Method 6: Eddy Covariance Uncertainty Estimation -- We quantified uncertainty in flux measurements. For NEE, there are two sources of uncertainty: random measurement uncertainty and gap-filling uncertainty. To estimate random uncertainty, we used the residuals from the 20 gap-filling models, shown to represent random uncertainty (Moffat et al. 2007), to parameterize Laplace double exponential distributions for binned fluxes (5 μmol CO2 m-2 s-1 intervals). We then ran 500 Monte Carlo simulations drawing samples from the distribution (Anderson et al. 2016) and measured the standard deviations over the integration periods (Knox et al. 2015). To estimate gap-filling uncertainty, we used the standard deviations from the 20 gap-filling predictions, again over the integration periods. We then added in quadrature the standard deviations representing the random measurement uncertainty and gap-filling uncertainty over the integration periods (i.e., 30-minute, daily, and yearly) to acquire 95% confidence intervals (Richardson and Hollinger 2007). Uncertainties associated with ER include the uncertainties associated with NEE (random measurement uncertainty and gap-filling uncertainty) and the uncertainty from the ER models trained on nighttime NEE to predict daytime ER. We measured the uncertainty associated with the ER model using the same method as the gap-filling uncertainty, by calculating the standard deviations of the 20 ER model predictions. Finally, added in quadrature random measurement uncertainty, gap-filling uncertainty, and ER model uncertainty to estimate ER uncertainty. For GPP, an uncertainty exists associated with the subtraction of ER from NEE, the partitioning uncertainty. To quantify this uncertainty, we subtracted each of the 20 ER model predictions from NEE, giving us 20 estimates of GPP. We then measured the standard deviation of these GPP estimates over the given integration period. Finally, we summed in quadrature the random measurement uncertainty, gap-filling uncertainty, ER model uncertainty, and partition uncertainty to estimate GPP uncertainty. b. Protocols: Method 1: none Method 2: Sonic Anemometer Tilt Correction Algorithms Organization: Boundary-Layer Meteorology Description: Wilczak, J. M., Oncley, S. P., and Stage, S. A. (2001). Sonic Anemometer Tilt Correction Algorithms. Boundary-Layer Meteorology, 99(1), 127–150. URL: https://doi.org/10.1023/A:1018966204465Averaging Detrending and Filtering of Eddy Covariance Time Series Organization: Handbook of Micrometeorology Description: Moncrieff, J., Clement, R., Finnigan, J., and Meyers, T. (2005). Averaging, Detrending, and Filtering of Eddy Covariance Time Series. In X. Lee, W. Massman, and B. Law (Eds.), Handbook of Micrometeorology (Vol. 29, pp. 7–31). Dordrecht: Kluwer Academic Publishers. URL: https://doi.org/10.1007/1-4020-2265-4_2Quality Control and Flux Sampling Problems for Tower and Aircraft Data Organization: Journal of Atmospheric and Oceanic Technology Description: Vickers, D., and Mahrt, L. (1997). Quality Control and Flux Sampling Problems for Tower and Aircraft Data. Journal of Atmospheric and Oceanic Technology, 14(3), 512–526. URL: https://doi.org/10.1175/1520-0426(1997)014<0512:QCAFSP>2.0.CO;2Protocol for calculating sampling error in CO2 flux measurements Organization: Journal of Geophysical Research Description: Finkelstein, P. L., and Sims, P. F. (2001). Sampling error in eddy correlation flux measurements. Journal of Geophysical Research: Atmospheres, 106(D4), 3503–3509. URL: https://doi.org/10.1029/2000JD900731A system to measure surface fluxes of momentum, sensible heat, water vapour and carbon dioxide Organization: Journal of Hydrology Description: Moncrieff, J. B., Massheder, J. M., de Bruin, H., Elbers, J., Friborg, T., Heusinkveld, B., et al. (1997). A system to measure surface fluxes of momentum, sensible heat, water vapour and carbon dioxide. Journal of Hydrology, 188–189, 589–611. URL: https://doi.org/10.1016/S0022-1694(96)03194-0Impact of post-field data processing on eddy covariance flux estimates and energy balance closure Organization: Meteorologische Zeitschrift Description: Mauder, M., and Foken, T. (2006). Impact of post-field data processing on eddy covariance flux estimates and energy balance closure. Meteorologische Zeitschrift, 15(6), 597–609. URL: https://doi.org/10.1127/0941-2948/2006/0167Correction of flux measurements for density effects due to heat and water vapour transfer Organization: Quarterly Journal of the Royal Meteorological Society Description: Webb, E. K., Pearman, G. I., and Leuning, R. (1980). Correction of flux measurements for density effects due to heat and water vapour transfer. Quarterly Journal of the Royal Meteorological Society, 106(447), 85–100. URL: https://doi.org/10.1002/qj.49710644707 Method 3: none Method 4: none Method 5: none Method 6: none c. Instrumentation: Method 1: Closed Path CO2/H2O Analyzer Manufacturer: LI-COR (Model: LI-7200) Parameter: CO2 concentration (Accuracy: 1%, Range: 0-3000 ppm) Parameter: H2O concentration (Accuracy: 2%, Range: 0 to 60 ppt) CSAT3 Sonic Anemometer Manufacturer: Campbell Scientific Instruments (Model: CSAT3) Parameter: ux (Accuracy: 2-6%, Readability: 1 mm/s, Range: 0-60 m/s) Parameter: uy (Accuracy: 2-6%, Readability: 1 mm/s, Range: 0-60 m/s) Parameter: uz (Accuracy: 2-6%, Readability: 0.5 mm/s, Range: 0-60 m/s) Data Logger Manufacturer: Campbell Scientific Instruments (Model: CR3000) Parameter: Voltage (Accuracy: 0.04%, ) Net Radiometer Manufacturer: Hukseflux (Model: NR01) Parameter: Infrared radiation (Accuracy: 2.5% (10% daily sums), Range: 0 to 1000 W/m^2) Net Radiometer Manufacturer: Kipp & Zonen (Model: CNR2) Parameter: Infrared radiation (Accuracy: 5%, Range: 0-20 uV/W/m^2) Parameter: Short-wave radiation (Accuracy: 5%, Range: 0-20 uV/W/m^2) Net Radiometer Manufacturer: Hukseflux (Model: NR01) Parameter: Solar short-wave radiation (Accuracy: 2.5% (10% daily sums), Range: 0 to 2000 W/m^2) Parameter: Temperature (Accuracy: 1%, Range: -40 to 80ºC) PAR Sensor Manufacturer: LI-COR (Model: LI190R) Parameter: PAR (Accuracy: 5% NIST, Range: 0-10,000 umol/s/m^2) Pressure Transducer Manufacturer: Campbell Scientific Instruments (Model: CS455) Parameter: Water pressure (Accuracy: 0.05%, Range: 0 to 50 kPa) Pressure Transducer Manufacturer: Campbell Scientific Instruments (Model: CS456) Parameter: Water pressure (Accuracy: 0.05%, Range: 0 to 50 kPa) Parameter: Water temperature (Accuracy: 0.2ºC, Readability: 0.1ºC, Range: -10 to 80ºC) Pressure Transducer Manufacturer: Campbell Scientific Instruments (Model: CS455) Parameter: Water temperature (Accuracy: 0.2ºC, Readability: 0.1ºC, Range: -10 to 80ºC) Rain Gauge Manufacturer: Texas Electronics (Model: TE525WS) Parameter: Precipitation (Accuracy: 1%, Readability: 0.254 mm, Range: 0 to 50 mm/hr) Soil Thermocouple Probe Manufacturer: Campbell Scientific Instruments (Model: TCAV) Parameter: Soil Temperature Temperature Probe Manufacturer: Campbell Scientific Instruments (Model: 107) Parameter: Temperature (Readability: 0.2ºC, Range: -35 to 50ºC) Temperature/Relative Humidity Probe Manufacturer: Campbell Scientific Instruments (Model: HMP45C) Parameter: Air temperature (Accuracy: 0.2-0.5ºC, Readability: 0.1ºC, Range: -39.2 to 60ºC) Parameter: Relative humidity (Accuracy: 1-3%, Readability: 0.1%, Range: 0.8 to 100% RH) Total Solar Pyranometer Manufacturer: LI-COR (Model: LI200R) Parameter: Solar Radiation (Accuracy: 5% NIST, Range: 3000 W/m^2) Method 2: none Method 3: none Method 4: none Method 5: none Method 6: none d. Taxonomy and Systematics: Method 1: not applicable Method 2: not applicable Method 3: not applicable Method 4: not applicable Method 5: not applicable Method 6: not applicable e. Speclies List: f. Permit History: Method 1: not applicable Method 2: not applicable Method 3: not applicable Method 4: not applicable Method 5: not applicable Method 6: not applicable 4. Project Personnel a. Personnel: 1: Peter Hawman 2: Deepak Mishra 3: Wade M. Sheldon, Jr. b. Affiliations: 1: University of Georgia, Athens, Georgia 2: University of Georgia, Athens, Georgia 3: University of Georgia, Athens, Georgia III. Data Set Status and Accessibility A. Status 1. Latest Update: 06-Oct-2025 2. Latest Archive Date: 06-Oct-2025 3. Latest Metadata Update: 06-Oct-2025 4. Data Verification Status: Reviewed by IM B. Accessibility 1. Storage Location and Medium: Stored at GCE-LTER Data Management Office Dept. of Marine Sciences Univ. of Georgia Athens, GA 30602-3636 USA on media: electronic data download (WWW) or compact disk 2. Contact Person: Name: Adam Sapp Address: Department of Marine Sciences University of Georgia Athens, Georgia 30602 Country: USA Email: asapp@uga.edu 3. Copyright Restrictions: not copyrighted 4. Restrictions: This information is licensed under a Creative Commons Attribution 4.0 International License (see: https://creativecommons.org/licenses/by/4.0/). The consumer of these data ("Data User" herein) has an ethical obligation to cite it appropriately in any publication that results from its use. The Data User should realize that these data may be actively used by others for ongoing research and that coordination may be necessary to prevent duplicate publication. The Data User is urged to contact the authors of these data if any questions about methodology or results occur. Where appropriate, the Data User is encouraged to consider collaboration or co-authorship with the authors. The Data User should realize that misinterpretation of data may occur if used out of context of the original study. While substantial efforts are made to ensure the accuracy of data and associated documentation, complete accuracy of data sets cannot be guaranteed. All data are made available "as is." The Data User should be aware, however, that data are updated periodically and it is the responsibility of the Data User to check for new versions of the data. The data authors and the repository where these data were obtained shall not be liable for damages resulting from any use or misinterpretation of the data. a. Release Date: Affiliates: Aug 15, 2025, Public: Aug 15, 2025 b. Citation: Data provided by the Georgia Coastal Ecosystems Long Term Ecological Research Project, supported by funds from NSF OCE 1832178 (data set MSH-GCEM-2508) c. Disclaimer: The user assumes all responsibility for errors in judgement based on interpretation of data and analyses presented in this data set. 5. Costs: free electronic data download via WWW, distribution on CD may be subject to nominal processing and handling fee IV. Data Structural Descriptors A. Data Set File 1. File Name: MSH-GCEM-2508_daily_1_2.CSV 2. Size: 4018 records 3. File Format: ASCII text (comma-separated value format) 3a. Delimiters: single comma 4. Header Information: 5 lines of ASCII text 5. Alphanumeric Attributes: 6. Quality Control Flag Codes: Q = questionable value, I = invalid value, E = estimated value 7. Authentication Procedures: 8. Calculations: 9. Processing History: Software version: GCE Data Toolbox Version 3.9.10 (23-May-2022) Data structure version: GCE Data Structure 1.1 (29-Mar-2001) Original data file processed: MSH-GCEM-2508_daily.txt (4018 records) Data processing history: 02-Sep-2025: new GCE Data Structure 1.1 created ('newstruct') 02-Sep-2025: 4018 rows imported from ASCII data file 'MSH-GCEM-2508_daily.txt' ('imp_ascii') 02-Sep-2025: 13 metadata fields in file header parsed ('parse_header') 02-Sep-2025: data structure validated ('gce_valid') 02-Sep-2025: updated 1 metadata fields in the Dataset section(s) ('addmeta') 02-Sep-2025: imported Dataset, Project, Site, Study, Status, Supplement metadata descriptors from the GCE Metabase ('imp_gcemetadata') 02-Sep-2025: updated 57 metadata fields in the Dataset, Project, Site, Status, Study, Supplement section(s) ('addmeta') 06-Oct-2025: updated 1 metadata fields in the Dataset section(s) ('addmeta') 06-Oct-2025: imported Dataset, Project, Site, Study, Status, Supplement metadata descriptors from the GCE Metabase ('imp_gcemetadata') 06-Oct-2025: updated 57 metadata fields in the Dataset, Project, Site, Status, Study, Supplement section(s) ('addmeta') 06-Oct-2025: updated 6 metadata fields in the Data section(s) ('addmeta') 06-Oct-2025: updated 15 metadata fields in the Status, Data sections to reflect attribute metadata ('updatecols') 06-Oct-2025: parsed and formatted metadata ('listmeta') B. Variable Information 1. Variable Name: column 1. Date column 2. proportion_gapfilled column 3. NEE_daily_mean column 4. NEE_daily_std column 5. NEE_daily_total column 6. NEE_daily_CI_95 column 7. ER_daily_mean column 8. ER_daily_std column 9. ER_daily_total column 10. ER_daily_CI_95 column 11. GPP_daily_mean column 12. GPP_daily_std column 13. GPP_daily_total column 14. GPP_daily_CI_95 2. Variable Definition: column 1. Calendar date of observation column 2. Proportion of 30-minute flux measurements that were gap-filled column 3. Daily mean of 30-minute net ecosystem exchange (carbon dioxide) column 4. Daily standard deviation of 30-minute net ecosystem exchange (carbon dioxide) column 5. Daily total net ecosystem exchange (carbon) column 6. 95% confidence interval of daily total net ecosystem exchange (carbon) column 7. Daily mean of 30-minute ecosystem respiration (carbon dioxide) column 8. Daily standard deviation of 30-minute ecosystem respiration (carbon dioxide) column 9. Daily total ecosystem respiration (carbon) column 10. 95% confidence interval of daily total ecosystem respiration (carbon) column 11. Daily mean of 30-minute gross primary production (carbon dioxide) column 12. Daily standard deviation of 30-minute gross primary production (carbon dioxide) column 13. Daily total gross primary production (carbon) column 14. 95% confidence interval of daily total gross primary production (carbon) 3. Units of Measurement: column 1. yyyy-mm-dd column 2. none column 3. umol/m^2/s column 4. umol/m^2/s column 5. g/m^2/day column 6. g/m^2/day column 7. umol/m^2/s column 8. umol/m^2/s column 9. g/m^2/day column 10. g/m^2/day column 11. umol/m^2/s column 12. umol/m^2/s column 13. g/m^2/day column 14. g/m^2/day 4. Data Type a. Storage Type: column 1. string column 2. floating-point column 3. floating-point column 4. floating-point column 5. floating-point column 6. floating-point column 7. floating-point column 8. floating-point column 9. floating-point column 10. floating-point column 11. floating-point column 12. floating-point column 13. floating-point column 14. floating-point b. Variable Codes: c. Numeric Range: column 1. (none) column 2. 0 to 1 column 3. -3.7973 to 2.4806 column 4. 0.28654 to 8.3294 column 5. -3.9403 to 2.5741 column 6. 0.40423 to 0.98149 column 7. 0.2737 to 5.3513 column 8. 0.051853 to 1.9607 column 9. 0.28401 to 5.5529 column 10. 0.40616 to 0.98662 column 11. -0.429 to 7.1929 column 12. 0.29265 to 8.4578 column 13. -0.44516 to 7.4639 column 14. 0.40808 to 0.99173 d. Missing Value Code: 5. Data Format a. Column Type: column 1. text column 2. numerical column 3. numerical column 4. numerical column 5. numerical column 6. numerical column 7. numerical column 8. numerical column 9. numerical column 10. numerical column 11. numerical column 12. numerical column 13. numerical column 14. numerical b. Number of Columns: 14 c. Decimal Places: column 1. 0 column 2. 7 column 3. 7 column 4. 7 column 5. 7 column 6. 7 column 7. 7 column 8. 7 column 9. 7 column 10. 7 column 11. 7 column 12. 7 column 13. 7 column 14. 7 6. Logical Variable Type: column 1. calculation (none) column 2. calculation (continuous) column 3. calculation (continuous) column 4. calculation (continuous) column 5. calculation (continuous) column 6. calculation (continuous) column 7. calculation (continuous) column 8. calculation (continuous) column 9. calculation (continuous) column 10. calculation (continuous) column 11. calculation (continuous) column 12. calculation (continuous) column 13. calculation (continuous) column 14. calculation (continuous) 7. Flagging Criteria: column 1. none column 2. none column 3. none column 4. none column 5. none column 6. none column 7. none column 8. none column 9. none column 10. none column 11. none column 12. none column 13. none column 14. none C. Data Anomalies: V. Supplemental Descriptors A. Data Acquisition 1. Data Forms: 2. Form Location: 3. Data Entry Validation: B. Quality Assurance/Quality Control Procedures: C. Supplemental Materials: D. Computer Programs: E. Archival Practices: F. Publications: not specified G. History of Data Set Usage 1. Data Request History: not specified 2. Data Set Update History: none 3. Review History: none 4. Questions and Comments from Users: none