The concentration of dissolved oxygen (DO), an important measure of water quality and river metabolism, varies tremendously in time and space. Riverine DO is commonly perceived as regulated by interacting and competing drivers (light, temperature and flow) that define rivers’ climate. Its continental-scale drivers, however, have remained elusive, partly due to the scarcity and spatio-temporal inconsistency of water quality data. Here we show, via a deep learning model (long short-term memory) trained using data from 580 rivers, that temperature predominantly drives daily DO dynamics in the contiguous United States. Light comes a close second, whereas flow imparts minimal influence. This work showcases the promise of using deep learning models for data filling that enables large-scale systematic analysis of patterns and drivers. Results show fairly accurate prediction of DO by temperature alone, and declining DO in warming rivers, which has important implications for water security and ecosystem health in the future climate.
This is a preview of subscription content, access via your institution
Subscribe to this journal
Receive 12 digital issues and online access to articles
$79.00 per year
only $6.58 per issue
Rent or buy this article
Get just this article for as long as you need it
Prices may be subject to local taxes which are calculated during checkout
Discharge and biogeochemical data were downloaded from the USGS NWIS77 at the website of https://waterdata.usgs.gov/nwis. The meteorological dataset of DAYMET78 is available from the website of https://daymet.ornl.gov. Basin characteristics of topography, hydro, climate, land cover, soil, and geology from the Geospatial Attributes of Gages for Evaluating Streamflow dataset Version II (GAGES-II)56 are archived at https://water.usgs.gov/GIS/metadata/usgswrd/XML/gagesII_Sept2011.xml.
The deep learning LSTM code and instruction are available from the GitHub website at https://github.com/mhpi/hydroDL. The dataRetrieval R package for downloading discharge and biogeochemical data is available at https://github.com/USGS-R/dataRetrieval. Supporting data and script (for example, model inputs, outputs and a downloading sample script) are available at https://github.com/Li-Reactive-Water-Group/NatureWater-US-river-DO-dataset. The scripts for data analysis and plotting were developed in MATLAB (R2018a) and RSTUDIO (v2022.07.1) and are available from the authors upon reasonable request.
Tesoriero, A. J., Terziotti, S. & Abrams, D. B. Predicting redox conditions in groundwater at a regional scale. Environ. Sci. Technol. 49, 9657–9664 (2015).
Briggs, M. A. et al. Exploring local riverbank sediment controls on the occurrence of preferential groundwater discharge points. Water 14, 11 (2022).
O’Donnell, B. & Hotchkiss, E. R. Resistance and resilience of stream metabolism to high flow disturbances. Biogeosciences 19, 1111–1134 (2022).
Rosamond, M. S., Thuss, S. J. & Schiff, S. L. Dependence of riverine nitrous oxide emissions on dissolved oxygen levels. Nat. Geosci. 5, 715–718 (2012).
Sundby, B. et al. The effect of oxygen on release and uptake of cobalt, manganese, iron and phosphate at the sediment-water interface. Geochim. Cosmochim. Acta 50, 1281–1288 (1986).
Wang, S., Jin, X., Bu, Q., Jiao, L. & Wu, F. Effects of dissolved oxygen supply level on phosphorus release from lake sediments. Colloids Surf. A 316, 245–252 (2008).
Jane, S. F. et al. Widespread deoxygenation of temperate lakes. Nature 594, 66–70 (2021).
Oschlies, A., Brandt, P., Stramma, L. & Schmidtko, S. Drivers and mechanisms of ocean deoxygenation. Nat. Geosci. 11, 467–473 (2018).
Girard, J. E. Principles of Environmental Chemistry (Jones & Bartlett Learning, 2013).
Tromans, D. Temperature and pressure dependent solubility of oxygen in water: a thermodynamic analysis. Hydrometallurgy 48, 327–342 (1998).
Ulseth, A. J. et al. Distinct air–water gas exchange regimes in low- and high-energy streams. Nat. Geosci. 12, 259–263 (2019).
Helton, A. M., Poole, G. C., Payn, R. A., Izurieta, C. & Stanford, J. A. Scaling flow path processes to fluvial landscapes: an integrated field and model assessment of temperature and dissolved oxygen dynamics in a river–floodplain–aquifer system. J. Geophys. Res. Biogeosci. https://doi.org/10.1029/2012JG002025 (2012).
Blaszczak, J. R., Delesantro, J. M., Urban, D. L., Doyle, M. W. & Bernhardt, E. S. Scoured or suffocated: urban stream ecosystems oscillate between hydrologic and dissolved oxygen extremes. Limnol. Oceanogr. 64, 877–894 (2019).
Arroita, M., Elosegi, A. & Hall, R. O. Jr Twenty years of daily metabolism show riverine recovery following sewage abatement. Limnol. Oceanogr. 64, S77–S92 (2019).
Diamond, J. S. et al. Light and hydrologic connectivity drive dissolved oxygen synchrony in stream networks. Limnol. Oceanogr. https://aslopubs.onlinelibrary.wiley.com/doi/abs/10.1002/lno.12271 (2022).
Amon, R. M. & Benner, R. Photochemical and microbial consumption of dissolved organic carbon and dissolved oxygen in the Amazon River system. Geochim. Cosmochim. Acta 60, 1783–1792 (1996).
Bernhardt, E. S. et al. The metabolic regimes of flowing waters. Limnol. Oceanogr. 63, S99–S118 (2018).
Utz, R. M., Bookout, B. J. & Kaushal, S. S. Influence of temperature, precipitation, and cloud cover on diel dissolved oxygen ranges among headwater streams with variable watershed size and land use attributes. Aquat. Sci. 82, 82 (2020).
Caraco, N. F. et al. Dissolved oxygen declines in the hudson river associated with the invasion of the zebra mussel (Dreissena polymorpha). Environ. Sci. Technol. 34, 1204–1210 (2020).
Palmer, M. J., Chételat, J., Jamieson, H. E., Richardson, M. & Amyot, M. Hydrologic control on winter dissolved oxygen mediates arsenic cycling in a small subarctic lake. Limnol. Oceanogr. https://doi.org/10.1002/lno.11556 (2021).
Canadell, M. B., Gómez‐Gener, L., Clémençon, M., Lane, S. N. & Battin, T. J. Daily entropy of dissolved oxygen reveals different energetic regimes and drivers among high-mountain stream types. Limnol. Oceanogr. 66, 1594–1610 (2021).
Guo, D. et al. Key factors affecting temporal variability in stream water quality. Water Resour. Res. 55, 112–129 (2019).
Burns, D. A. et al. Monitoring the riverine pulse: applying high-frequency nitrate data to advance integrative understanding of biogeochemical and hydrological processes. WIREs Water 6, e1348 (2019).
Zhi, W. et al. From hydrometeorology to river water quality: can a deep learning model predict dissolved oxygen at the continental scale. Environ. Sci. Technol. 55, 2357–2368 (2021).
Shen, C. A transdisciplinary review of deep learning research and its relevance for water resources scientists. Water Resour. Res. 54, 8558–8593 (2018).
Moatar, F., Abbott, B. W., Minaudo, C., Curie, F. & Pinay, G. Elemental properties, hydrology, and biology interact to shape concentration-discharge curves for carbon, nutrients, sediment, and major ions. Water Resour. Res. 53, 1270–1287 (2017).
Li, L. et al. Climate controls on river chemistry. Earths Future 10, e2021EF002603 (2022).
Moriasi, D. N., Gitau, M. W., Pai, N. & Daggupati, P. Hydrologic and water quality models: performance measures and evaluation criteria. Trans. ASABE 58, 1763–1785 (2015).
Feng, D., Fang, K. & Shen, C. Enhancing streamflow forecast and extracting insights using long-short term memory networks with data integration at continental scales. Water Resour. Res. https://doi.org/10.1029/2019WR026793 (2020).
Kratzert, F. et al. Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large-sample datasets. Hydrol. Earth Syst. Sci. 23, 5089–5110 (2019).
Fang, K., Kifer, D., Lawson, K., Feng, D. & Shen, C. The data synergy effects of time-series deep learning models in hydrology. Water Resour. Res. 58, e2021WR029583 (2022).
Doran, P. M. in Bioprocess Engineering Principles 2nd edition (ed. Doran, P. M.) 379–444 (Academic Press, 2013).
Townsend, S. A., Webster, I. T. & Schult, J. H. Metabolism in a groundwater-fed river system in the Australian wet/dry tropics: tight coupling of photosynthesis and respiration. J. North Am. Benthol. Soc. 30, 603–620 (2011).
Davison, I. R. Environmental effects on algal photosynthesis: temperature. J. Phycol. 27, 2–8 (1991).
Hancke, K. & Glud, R. N. Temperature effects on respiration and photosynthesis in three diatom-dominated benthic communities. Aquat. Microb. Ecol. 37, 265–281 (2004).
Zhi, W. & Li, L. The shallow and deep hypothesis: subsurface vertical chemical contrasts shape nitrate export patterns from different land uses. Environ. Sci. Technol. 54, 11915–11928 (2020).
Stewart, B. et al. Streams as mirrors: reading subsurface water chemistry from stream chemistry. Water Resour. Res. 58, e2021WR029931 (2022).
Hirsch, R. M., Moyer, D. L. & Archfield, S. A. Weighted regressions on time, discharge, and season (wrtds), with an application to chesapeake bay river inputs. J. Am. Water Resour. Assoc. 46, 857–880 (2010).
Rodell, M. et al. The global land data assimilation system. Bull. Am. Meteorol. Soc. 85, 381–394 (2004).
Berg, P., Almén, F. & Bozhinova, D. HydroGFD3.0 (Hydrological Global Forcing Data): a 25 km global precipitation and temperature data set updated in near-real time. Earth Syst. Sci. Data 13, 1531–1545 (2021).
Hersbach, H. et al. The ERA5 global reanalysis. Q. J. R. Meteorolog. Soc. 146, 1999–2049 (2020).
Do, H. X., Gudmundsson, L., Leonard, M. & Westra, S. The global streamflow indices and metadata archive (GSIM)—part 1: the production of a daily streamflow archive and metadata. Earth Syst. Sci. Data 10, 765–785 (2018).
Virro, H., Amatulli, G., Kmoch, A., Shen, L. & Uuemaa, E. GRQA: global river water quality archive. Earth Syst. Sci. Data https://doi.org/10.5194/essd-2021-51 (2021).
Bernhardt, E. S. et al. Light and flow regimes regulate the metabolism of rivers. Proc. Natl. Acad. Sci. U.S.A. 119, e2121976119 (2022).
Troch, P. A., Carrillo, G., Sivapalan, M., Wagener, T. & Sawicz, K. Climate–vegetation–soil interactions and long-term hydrologic partitioning: signatures of catchment co-evolution. Hydrol. Earth Syst. Sci. 17, 2209–2217 (2013).
Breitburg, D. et al. Declining oxygen in the global ocean and coastal waters. Science 359, eaam7240 (2018).
Ni, W., Li, M., Ross, A. C. & Najjar, R. G. Large projected decline in dissolved oxygen in a eutrophic estuary due to climate change. J. Geophys. Res. Oceans 124, 8271–8289 (2019).
Schmidtko, S., Stramma, L. & Visbeck, M. Decline in global oceanic oxygen content during the past five decades. Nature 542, 335–339 (2017).
IPCC Climate Change 2021: The Physical Science Basis (eds. Masson-Delmotte, V. et al) (Cambridge Univ. Press, 2021).
Danladi Bello, A.-A., Hashim, N. & Mohd Haniffah, M. Predicting impact of climate change on water temperature and dissolved oxygen in tropical rivers. Climate 5, 58 (2017).
Ice, G. & Sugden, B. Summer dissolved oxygen concentrations in forested streams of northern Louisiana. South. J. Appl. For. 27, 92–99 (2003).
Perkins-Kirkpatrick, S. E. & Lewis, S. C. Increasing trends in regional heatwaves. Nat. Commun. 11, 3357 (2020).
Blaszczak, J. R. et al. Extent, patterns, and drivers of hypoxia in the world’s streams and rivers. Limnol. Oceanogr. Lett. https://doi.org/10.1002/lol2.10297 (2022).
Carter, A. M., Blaszczak, J. R., Heffernan, J. B. & Bernhardt, E. S. Hypoxia dynamics and spatial distribution in a low gradient river. Limnol. Oceanogr. 66, 2251–2265 (2021).
Gorelick, N. et al. Google Earth Engine: planetary-scale geospatial analysis for everyone. Remote Sens. Environ. 202, 18–27 (2017).
Falcone, J. A. GAGES-II: Geospatial Attributes of GAGES for Evaluating Streamflow (US Geological Survey, 2011); https://water.usgs.gov/GIS/metadata/usgswrd/XML/gagesII_Sept2011.xml
Hirsch, R. M. & De Cicco, L. A. User Guide to Exploration and Graphics for RivEr Trends (EGRET) and dataRetrieval: R Packages for Hydrologic Data Report No. 2328-7055 (US Geological Survey, 2015).
Sterle, G. et al. CAMELS-Chem: augmenting CAMELS (Catchment Attributes and Meteorology for Large-sample Studies) with atmospheric and stream water chemistry data. Hydrol. Earth Syst. Sci. Discuss. 2022, 1–23 (2022).
Spahr, N. E., Dubrovsky, N. M., Gronberg, J. M., Franke, O. & Wolock, D. M. Nitrate Loads and Concentrations in Surface-Water Base Flow and Shallow Groundwater for Selected Basins in the United States, Water Years 1990–2006 (US Geological Survey, 2010).
Greff, K., Srivastava, R. K., Koutník, J., Steunebrink, B. R. & Schmidhuber, J. LSTM: a search space odyssey. IEEE Trans Neural Netw. Learn. Syst. 28, 2222–2232 (2016).
Hochreiter, S. & Schmidhuber, J. Long short-term memory. Neural Comput. 9, 1735–1780 (1997).
Rahmani, F. et al. Exploring the exceptional performance of a deep learning stream temperature model and the value of streamflow data. Environ. Res. Lett. https://doi.org/10.1088/1748-9326/abd501 (2020).
Ma, K. et al. Transferring hydrologic data across continents-leveraging data-rich regions to improve hydrologic prediction in data-sparse regions. Water Resour. Res. https://doi.org/10.1029/2020wr028600 (2021).
Feng, D., Lawson, K. & Shen, C. Mitigating prediction error of deep learning streamflow models in large data‐sparse regions with ensemble modeling and soft data. Geophys. Res. Lett. https://doi.org/10.1029/2021gl092999 (2021).
Fang, K., Pan, M. & Shen, C. P. The value of SMAP for long-term soil moisture estimation with the help of deep learning. IEEE Trans. Geosci. Remote 57, 2221–2233 (2019).
Peterson, R. A. Finding optimal normalizing transformations via bestNormalize. R J. 13, 294–313 (2021).
Kratzert, F., Klotz, D., Brenner, C., Schulz, K. & Herrnegger, M. Rainfall–runoff modelling using long short-term memory (LSTM) networks. Hydrol. Earth Syst. Sci. 22, 6005–6022 (2018).
Gal, Y. & Ghahramani, Z. A theoretically grounded application of dropout in recurrent neural networks. Adv. Neural Inf. Process. Syst. 29, 1019–1027 (2016).
Fang, K., Shen, C., Kifer, D. & Yang, X. Prolongation of SMAP to spatiotemporally seamless coverage of continental U.S. using a deep learning neural network. Geophys. Res. Lett. 44, 030–011,039 (2017).
Zhao, N., Fan, Z. & Zhao, M. A new approach for estimating dissolved oxygen based on a high-accuracy surface modeling method. Sensors 21, 3954 (2021).
Stajkowski, S., Zeynoddin, M., Farghaly, H., Gharabaghi, B. & Bonakdari, H. A methodology for forecasting dissolved oxygen in urban streams. Water 12, 2568 (2020).
Stefan, H. G. & Fang, X. Dissolved oxygen model for regional lake analysis. Ecol. Modell. 71, 37–68 (1994).
Zhu, S. & Heddam, S. Prediction of dissolved oxygen in urban rivers at the Three Gorges Reservoir, China: extreme learning machines (ELM) versus artificial neural network (ANN). Water Qual. Res. J. 55, 106–118 (2020).
Yu, X., Shen, J. & Du, J. A machine-learning-based model for water quality in coastal waters, taking dissolved oxygen and hypoxia in Chesapeake Bay as an example. Water Resour. Res. https://doi.org/10.1029/2020wr027227 (2020).
Fang, K., Kifer, D., Lawson, K., Feng, D. & Shen, C. The data synergy effects of time‐series deep learning models in hydrology. Water Resour. Res. https://doi.org/10.1029/2021wr029583 (2022).
Sivapalan, M. Prediction in ungauged basins: a grand challenge for theoretical hydrology. Hydrol. Processes 17, 3163–3170 (2003).
National Water Information System data available on the World Wide Web (USGS, 2016); http://waterdata.usgs.gov/nwis/
Thornton, M. M. et al. Daymet: daily surface weather data on a 1-km grid for North America, version 4. ORNL DAAC https://doi.org/10.3334/ORNLDAAC/1840 (2020).
The project was supported by the Barry and Shirley Isett professorship to W.Z. and L.L. from the Department of Civil and Environmental Engineering at Penn State University. W.O. and C.S. were supported by the Office of Biological and Environmental Research of the United States Department of Energy under contract DE-SC0016605.
The authors declare no competing interests.
Peer review information
Nature Water thanks Alice Carter, Ryan Utz, Danlu Guo and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Figs. 1–8 and Table 1.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Zhi, W., Ouyang, W., Shen, C. et al. Temperature outweighs light and flow as the predominant driver of dissolved oxygen in US rivers. Nat Water 1, 249–260 (2023). https://doi.org/10.1038/s44221-023-00038-z