Title of dataset: Estimates of ecosystem metabolism for 59 rivers in North America, 2008-2021

This dataset is companion with the github repository https://github.com/nmarzolf91/usgs_metabolism_project that accesses USGS water quality and hydrologic data to estimate ecosystem metabolism in 59 rivers across the US from 2007-2021. Contained in this data citation are information about sites, raw time-series data at sub-daily intervals, prepared time series for modeling metabolism, metabolism outputs in raw form, processed and QAQC'd metabolism estimates, and supplementary information for the associated manuscript(s).

Description of the data and file structure

• 1_site_info.csv Number of variables: 5 Variable list: Site Number: USGS code for gage station Name: USGS name for site Latitude: site latitude, in dd.ddd Longitude: site longitude, in dd.ddd Drainage Areak (km2): drainage area upstream of the gage station, in sq. km

• 2_timeseries_raw.zip In this folder are two sources of the sensor data. Data collected from 2007-2016 are stored in usgs_from_streampulse_old, and data collected from 2017-2021 are stored in one csv file, usgs_data_2017-2021. The type of data are exactly the same, but the difference is due to who and when the data were aggregated and how the data were accessed, described in the manuscript for this dataset.

  • 2_timeseries_raw/usgs_data_2017_2021.csv : high-frequency river data from USGS NWIS sites, collected 2017-2021.

    • Number of variables: 6

    • Variable list:

      • ...1: row number index site: USGS site code

      • DateTime: Timestamp of data collected

      • disch: river discharge, in cfs

      • wtr: water temperature, in Celsius

      • doobs: dissolved oxygen concentration, in mg/L

  • 2_usgs_from_streampulse_old/: folder containing high-frequency river data from USGS NWIS sites, collected 2007 -- 2016

    • Number of files: 63

    • Naming convention: nwis_{site_code}_{startDate}_{endDate}.csv, where each file contains the entire record of data for a given site from 2007-2016

    • Number of variables in each file: 7

    • Variable list:

      • DateTime_UTC: Timestamp of each measurement, in UTC

      • Region: Two character string for region description in StreamPulse portal

      • Site: USGS site code

      • Value: value of variable defined in variable column

      • Variable: one of DO_mgL, Depth_m, Discharge_m3s, Light_PAR, WaterTemp_C, satDO_mgL

      • Flagtype: indicates potentially bad data, identified visually

      • Flagcomment: additional notes describing why a flagtype was added

• 3_usgs_sm_ready_all: this folder contains the merged form of the data in 2_timeseries_raw and modified to be directly input into streamMetabolizer models using the required format

  • Number of files: 59

  • Naming convention: nwis_{site_code}_{startDate}_{endDate}.csv, here each file contains the entire record of data for a given site from 2007-2021

  • Number of variables in each file: 8

    • site: nwis_{sitecode}

    • solar.time: timestamp of measurement, in solar time UTC

    • temp.water: water temperature, in C

    • DO.obs: dissolved oxygen, mg/L

    • DO.sat: dissolved oxygen at saturation, mg/L

    • depth: mean depth of the study reach, in m

    • discharge: discharge, in m3/s

    • light, in umol m-2 s-1

• 4_model_run_bayes: this directory contains the raw model outputs from streamMetabolizer, storing various outputs of the model in separate directories. The estimates, error, confidence intervals, Bayesian measures of model convergence of daily GPP, ER, and K600 are available in daily/, model-wide fit, predicted DO concentrations, gas exchange scaling relationships, and diagnostics of models are available in this directory. Processed and usable forms of the data are presented in data citation 5_usgs_metabolism.csv, and the relevant parameters are described in detail below.

  • daily

    • Number of files: 854

    • Naming convention: nwis_{sitecode}_{startdate}_{enddate}_daily.csv

    • Number of variables in each file: 74

  • datadaily

    • Number of files: 853

    • Naming convention: nwis_{sitecode}_{startdate}_{enddate}_datadaily.csv

    • Number of variables in each file: 7

  • estimates

    • Number of files: 845

    • Naming convention: nwis_{sitecode}_year.csv

    • Number of variables in each file: 26

  • KQ_overall

    • Number of files: 853

    • Naming convention: nwis_{sitecode}_{startdate}_{enddate}*_*KQ_overall.csv

    • Number of variables in each file: 13

  • mod_and_obs_DO

    • Number of files: 845

    • Naming convention: nwis_{sitecode}_{startdate}_{enddate}_mod_and_obs_DO.csv

    • Number of variables in each file: 9

  • overall

    • Number of files: 853

    • Naming convention: nwis_{sitecode}_{startdate}_{enddate}_overall.csv

    • Number of variables: 33

  • specs

    • Number of files: 853

    • Naming convention: nwis_{sitecode}_{startdate}_{enddate}_specs.csv

    • Number of variables: 1

  • diagnostics.csv: For each site year, descriptive statistics that inform how well the model fit.

    • Number of variables: 13

      • site

      • year

      • n_days: number of days with estimates

      • f_days: n_days/# of days in the year

      • resolution: sampling frequency of the input sensor data

      • K600_daily_sigma_Rhat: Gelman-Rubin convergence statistics for K600

      • err_obs_iid_sigma_Rhat: observation error convergence statistic

      • err_proc_iid_sigma_Rhat: process error convergence statistics

      • K_median: median daily gas exchange

      • K_range: range of gas exchange estimated for the given site year

      • neg_GPP: % days with negative GPP estimates (<-0.5 g O2 m-2 d-1)

      • pos_ER: % days with positive ER estimates (>0.5 g O2 m-2 d-1)

      • ER_K_r2: linear R2 between daily ER and K600, where high correlation between variables suggest poor model convergence

• 5_usgs_metabolism.csv

  • Number of variables: 26

  • Variable list:

    • site: usgs sitecode

    • resolution: resolution of sensor data

    • date: date of estimate

    • GPP: median GPP estimate

    • GPP.lower: 2.5%-ile GPP estimate

    • GPP.upper: 97.5%-ile GPP estimate

    • GPP.n_eff: number of effective samples to estimate GPP

    • GPP.Rhat: Gelman-Rubin convergence statistic for GPP

    • ER: median ER estimate

    • ER.lower: 2.5%-ile ER estimate

    • ER.upper: 97.5%-ile GPP estimate

    • ER.n_eff: number of effective samples to estimate ER

    • ER.Rhat: Gelman-Rubin convergence statistic for ER

    • K600: median: K600 estimate

    • K600.lower: 2.5%-ile K600 estimate

    • K600.upper: 97.5%-ile K600 estimate

    • K600.n_eff: number of effective samples to estimate K600

    • K600.Rhat: Gelman-Rubin convergence statistic for ER

    • DO.obs: mean daily DO

    • DO.sat: mean daily DO at saturation

    • DO.amp: daily amplitude of DO for the

    • DO.psat: mean daily %DO saturation

    • depth: mean daily depth

    • temp.water: mean daily temperature

    • discharge: mean daily discharge

    • shortwave: total daily incident shortwave radiation

• 6_fluxnet_metabolism.csv

  • Number of variables: 12

  • Variable list:

    • site: Fluxnet site name

    • biome: Fluxnet listed biome

    • date: date of estimate

    • year: year

    • DOY: Julian Day

    • GPP: GPP estimate (g C m-2 d-1)

    • ER: ER estimate (g C m-2 d-1)

    • Net: GPP - ER (g C m-2 d-1)

    • Temp: temperature, C

    • Precip: precipitation, mm

    • VPD: vapor pressure deficit, hPa

    • SW: shortwave radiation, W m-2

• 7_river_light.csv

  • Number of variables: 5

  • Variable list

    • site: nwis_{sitecode}

    • date

    • daily_SW_inc: total daily incoming shortwave radiation

    • daily_LAI: daily leaf area index

    • daily_PAR: total daily photosynthetically active radiation incident at the river surface

• ms_usgs_annual_metab: these files are supplementary tables for the submitted manuscript

  • table_s1_site_info: identical to 1_site_info.csv

  • table_s2_trend_stats.csv

    • Number of variables: 5

    • Variable list

      • site: nwis_{sitecode}

      • Sen's Slope: monotonic slope of annual metabolism by site

      • Sen's pvalue: pvalue associated with Sen's Slope

      • Mann-Kendall S: test statistic for Mann-Kendall test

      • Mann-Kendall p-value: p-value associated with Mann-Kendall test

  • table_s3_random_effects.csv

    • Number of variables: 5

    • Variable list:

      • site: nwis_{sitecode}

      • Interaction: fixed effect value

      • Light: fixed effect for light

      • Temperature: fixed effect for temperature

      • CVQ: fixed effect for flow variability

  • table_s4_driver_trends.csv

    • Number of variables: 5

    • Variable list

      • site: nwis_{sitecode}

      • ann_GPP_trend: character, is annual GPP increasing or decreasing

      • mat_mk_p: Mann-Kendall p-value for mean annual temperature

      • cv_q_p: Mann-Kendall p-value for flow variability

      • light_mk_p: Mann-Kendall p-value for total incident light

Sharing/Access information

Code and some data available at https://github.com/nmarzolf91/usgs_metabolism_project Data available on dryad at: https://doi.org/10.5061/dryad.bcc2fqzj2

Code/Software

R version 4.2.3 with various packages, including streamMetabolizer v0.12.0 and dataRetrieval v2.7.12