Using Machine Learning to Identify Hydrologic Signatures With an Encoder–Decoder Framework
Hydrologic signatures are quantitative metrics that describe a streamflow time series. Examples include annual maximum flow, baseflow index and recession shape descriptors. In this paper, we use machine learning (ML) to learn encodings that are optimal ML equivalents of hydrologic signatures, and th...
Saved in:
| Published in: | Water resources research Vol. 59; no. 3 |
|---|---|
| Main Authors: | , |
| Format: | Journal Article |
| Language: | English |
| Published: |
Washington
John Wiley & Sons, Inc
01.03.2023
|
| Subjects: | |
| ISSN: | 0043-1397, 1944-7973 |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Abstract | Hydrologic signatures are quantitative metrics that describe a streamflow time series. Examples include annual maximum flow, baseflow index and recession shape descriptors. In this paper, we use machine learning (ML) to learn encodings that are optimal ML equivalents of hydrologic signatures, and that are derived directly from the data. We compare the learned signatures to classical signatures, interpret their meaning, and use them to build rainfall‐runoff models in otherwise ungauged watersheds. Our model has an encoder–decoder structure. The encoder is a convolutional neural net mapping historical flow and climate data to a low‐dimensional vector encoding, analogous to hydrological signatures. The decoder structure includes stores and fluxes similar to a classical hydrologic model. For each timestep, the decoder uses current climate data, watershed attributes and the encoding to predict coefficients that distribute precipitation between stores and store outflow coefficients. The model is trained end‐to‐end on the U.S. CAMELS watershed data set to minimize streamflow error. We show that learned signatures can extract new information from streamflow series, because using learned signatures as input to the process‐informed model improves prediction accuracy over benchmark configurations that use classical signatures or no signatures. We interpret learned signatures by correlation with classical signatures, and by using sensitivity analysis to assess their impact on modeled store dynamics. Learned signatures are spatially correlated and relate to streamflow dynamics including seasonality, high and low extremes, baseflow and recessions. We conclude that process‐informed ML models and other applications using hydrologic signatures may benefit from replacing expert‐selected signatures with learned signatures.
Key Points
We built an encoder–decoder network to learn an optimal equivalent of hydrologic signatures
Using learned signatures as input to a process‐informed ML model improves prediction accuracy over benchmark configurations
We interpret learned signatures by correlation with classical signatures and by sensitivity analysis of their impact on model store dynamics |
|---|---|
| AbstractList | Hydrologic signatures are quantitative metrics that describe a streamflow time series. Examples include annual maximum flow, baseflow index and recession shape descriptors. In this paper, we use machine learning (ML) to learn encodings that are optimal ML equivalents of hydrologic signatures, and that are derived directly from the data. We compare the learned signatures to classical signatures, interpret their meaning, and use them to build rainfall‐runoff models in otherwise ungauged watersheds. Our model has an encoder–decoder structure. The encoder is a convolutional neural net mapping historical flow and climate data to a low‐dimensional vector encoding, analogous to hydrological signatures. The decoder structure includes stores and fluxes similar to a classical hydrologic model. For each timestep, the decoder uses current climate data, watershed attributes and the encoding to predict coefficients that distribute precipitation between stores and store outflow coefficients. The model is trained end‐to‐end on the U.S. CAMELS watershed data set to minimize streamflow error. We show that learned signatures can extract new information from streamflow series, because using learned signatures as input to the process‐informed model improves prediction accuracy over benchmark configurations that use classical signatures or no signatures. We interpret learned signatures by correlation with classical signatures, and by using sensitivity analysis to assess their impact on modeled store dynamics. Learned signatures are spatially correlated and relate to streamflow dynamics including seasonality, high and low extremes, baseflow and recessions. We conclude that process‐informed ML models and other applications using hydrologic signatures may benefit from replacing expert‐selected signatures with learned signatures.
Key Points
We built an encoder–decoder network to learn an optimal equivalent of hydrologic signatures
Using learned signatures as input to a process‐informed ML model improves prediction accuracy over benchmark configurations
We interpret learned signatures by correlation with classical signatures and by sensitivity analysis of their impact on model store dynamics Hydrologic signatures are quantitative metrics that describe a streamflow time series. Examples include annual maximum flow, baseflow index and recession shape descriptors. In this paper, we use machine learning (ML) to learn encodings that are optimal ML equivalents of hydrologic signatures, and that are derived directly from the data. We compare the learned signatures to classical signatures, interpret their meaning, and use them to build rainfall‐runoff models in otherwise ungauged watersheds. Our model has an encoder–decoder structure. The encoder is a convolutional neural net mapping historical flow and climate data to a low‐dimensional vector encoding, analogous to hydrological signatures. The decoder structure includes stores and fluxes similar to a classical hydrologic model. For each timestep, the decoder uses current climate data, watershed attributes and the encoding to predict coefficients that distribute precipitation between stores and store outflow coefficients. The model is trained end‐to‐end on the U.S. CAMELS watershed data set to minimize streamflow error. We show that learned signatures can extract new information from streamflow series, because using learned signatures as input to the process‐informed model improves prediction accuracy over benchmark configurations that use classical signatures or no signatures. We interpret learned signatures by correlation with classical signatures, and by using sensitivity analysis to assess their impact on modeled store dynamics. Learned signatures are spatially correlated and relate to streamflow dynamics including seasonality, high and low extremes, baseflow and recessions. We conclude that process‐informed ML models and other applications using hydrologic signatures may benefit from replacing expert‐selected signatures with learned signatures. We built an encoder–decoder network to learn an optimal equivalent of hydrologic signatures Using learned signatures as input to a process‐informed ML model improves prediction accuracy over benchmark configurations We interpret learned signatures by correlation with classical signatures and by sensitivity analysis of their impact on model store dynamics Hydrologic signatures are quantitative metrics that describe a streamflow time series. Examples include annual maximum flow, baseflow index and recession shape descriptors. In this paper, we use machine learning (ML) to learn encodings that are optimal ML equivalents of hydrologic signatures, and that are derived directly from the data. We compare the learned signatures to classical signatures, interpret their meaning, and use them to build rainfall‐runoff models in otherwise ungauged watersheds. Our model has an encoder–decoder structure. The encoder is a convolutional neural net mapping historical flow and climate data to a low‐dimensional vector encoding, analogous to hydrological signatures. The decoder structure includes stores and fluxes similar to a classical hydrologic model. For each timestep, the decoder uses current climate data, watershed attributes and the encoding to predict coefficients that distribute precipitation between stores and store outflow coefficients. The model is trained end‐to‐end on the U.S. CAMELS watershed data set to minimize streamflow error. We show that learned signatures can extract new information from streamflow series, because using learned signatures as input to the process‐informed model improves prediction accuracy over benchmark configurations that use classical signatures or no signatures. We interpret learned signatures by correlation with classical signatures, and by using sensitivity analysis to assess their impact on modeled store dynamics. Learned signatures are spatially correlated and relate to streamflow dynamics including seasonality, high and low extremes, baseflow and recessions. We conclude that process‐informed ML models and other applications using hydrologic signatures may benefit from replacing expert‐selected signatures with learned signatures. |
| Author | Botterill, Tom E. McMillan, Hilary K. |
| Author_xml | – sequence: 1 givenname: Tom E. orcidid: 0000-0002-2305-2650 surname: Botterill fullname: Botterill, Tom E. organization: Independent Researcher – sequence: 2 givenname: Hilary K. orcidid: 0000-0002-9330-9730 surname: McMillan fullname: McMillan, Hilary K. email: hmcmillan@sdsu.edu organization: San Diego State University |
| BookMark | eNp90MFOGzEQBmALUYlAe-sDWOLSA0tnbGe9PlYBClKqSgGUS6WV1zsbTDc2tTdCufUdeMM-SZOGQ4UEpxmNvhmN_kO2H2Igxj4inCII81mAEPMZSAkG99gIjVKFNlrusxGAkgVKow_YYc73AKjGpR6xH7fZhwX_Zt2dD8SnZFPYDobIr1oKg-_W_HLdptjHhXf82i-CHVaJMp_74Y7bwM-Diy2lP7-fzuhfxy-SXdJjTD_fs3ed7TN9eK5H7Pbi_GZyWUy_f72afJkWVgGOCxKNKjsyjSVTYUvaIiK1BKiFs63WnaayUcI1QMqVhHZctY0yHXZd5cDII_Zpd_chxV8rykO99NlR39tAcZVrUSEaA6WSG3r8gt7HVQqb77YKtDJawEaJnXIp5pyoq50f7OBjGJL1fY1QbwOv_w98s3TyYukh-aVN69e43PFH39P6TVvPZ5OZKFU1ln8BtmiTFg |
| CitedBy_id | crossref_primary_10_1007_s00477_025_02954_w crossref_primary_10_5194_hess_28_4971_2024 crossref_primary_10_1016_j_jappgeo_2024_105501 crossref_primary_10_1016_j_jhydrol_2024_130771 crossref_primary_10_1016_j_jhydrol_2024_130772 crossref_primary_10_1029_2024WR039008 crossref_primary_10_1016_j_jhydrol_2024_131389 crossref_primary_10_1016_j_jhydrol_2025_133654 crossref_primary_10_1016_j_teadva_2024_200116 crossref_primary_10_1016_j_jhydrol_2025_133033 crossref_primary_10_1016_j_jhydrol_2025_134044 crossref_primary_10_1016_j_jhydrol_2025_134004 crossref_primary_10_1016_j_atmosres_2025_108196 crossref_primary_10_1016_j_rineng_2025_106345 crossref_primary_10_1016_j_geoen_2024_213028 |
| Cites_doi | 10.3115/v1/D14-1179 10.1007/978-3-030-01246-5_18 10.1029/2022WR032404 10.1093/biosci/biv102 10.1016/j.envsoft.2021.104983 10.1111/j.1365-2427.2009.02204.x 10.1029/2019WR026635 10.1080/02626668609491024 10.21105/joss.04050 10.1029/2018WR023254 10.1111/j.1365-2427.2009.02307.x 10.31223/X5BH0P 10.1002/hyp.13632 10.5194/hess-15-2895-2011 10.1080/00401706.1991.10484804 10.1002/2017wr020528 10.1214/aos/1013203451 10.1109/IJCNN.2017.7966039 10.1109/CVPR42600.2020.00259 10.1111/1752-1688.12964 10.5194/hess-19-3951-2015 10.5194/hess-23-5089-2019 10.5194/hess-25-2045-2021 10.1002/hyp.10972 10.1029/2018wr022606 10.5194/hess-19-209-2015 10.2166/nh.2020.100 10.1002/hyp.11300 10.2166/hydro.2020.095 10.1002/2015wr017635 10.5194/hess-21-5293-2017 10.5281/zenodo.6712302 10.1016/j.jhydrol.2020.124631 10.1029/2018WR022643 10.5281/zenodo.6462813 10.1061/(asce)0733-9496(1994)120:4(485) 10.1002/hyp.6989 10.1029/2020WR028091 10.1029/2019WR026065 10.1029/2019wr026793 10.1016/j.advwatres.2007.01.005 10.1109/CVPRW.2018.00060 10.1038/s41467-021-26107-z 10.1016/s0022-1694(02)00171-3 10.1016/j.jhydrol.2011.11.055 10.1002/2017wr020401 10.1046/j.1523-1739.1996.10041163.x 10.1175/jhm-d-16-0284.1 10.1016/j.envsoft.2021.105159 10.1111/j.1749-8198.2007.00039.x 10.1086/694913 10.2307/1313099 10.1002/hyp.11476 10.1002/rra.700 10.1002/hyp.14596 |
| ContentType | Journal Article |
| Copyright | 2023. The Authors. 2023. This article is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License. |
| Copyright_xml | – notice: 2023. The Authors. – notice: 2023. This article is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License. |
| DBID | 24P AAYXX CITATION 7QH 7QL 7T7 7TG 7U9 7UA 8FD C1K F1W FR3 H94 H96 KL. KR7 L.G M7N P64 7S9 L.6 |
| DOI | 10.1029/2022WR033091 |
| DatabaseName | Wiley Online Library Open Access CrossRef Aqualine Bacteriology Abstracts (Microbiology B) Industrial and Applied Microbiology Abstracts (Microbiology A) Meteorological & Geoastrophysical Abstracts Virology and AIDS Abstracts Water Resources Abstracts Technology Research Database Environmental Sciences and Pollution Management ASFA: Aquatic Sciences and Fisheries Abstracts Engineering Research Database AIDS and Cancer Research Abstracts Aquatic Science & Fisheries Abstracts (ASFA) 2: Ocean Technology, Policy & Non-Living Resources Meteorological & Geoastrophysical Abstracts - Academic Civil Engineering Abstracts Aquatic Science & Fisheries Abstracts (ASFA) Professional Algology Mycology and Protozoology Abstracts (Microbiology C) Biotechnology and BioEngineering Abstracts AGRICOLA AGRICOLA - Academic |
| DatabaseTitle | CrossRef Civil Engineering Abstracts Aquatic Science & Fisheries Abstracts (ASFA) Professional Virology and AIDS Abstracts Technology Research Database Aqualine Water Resources Abstracts Biotechnology and BioEngineering Abstracts Environmental Sciences and Pollution Management Meteorological & Geoastrophysical Abstracts Aquatic Science & Fisheries Abstracts (ASFA) 2: Ocean Technology, Policy & Non-Living Resources Bacteriology Abstracts (Microbiology B) Algology Mycology and Protozoology Abstracts (Microbiology C) ASFA: Aquatic Sciences and Fisheries Abstracts AIDS and Cancer Research Abstracts Engineering Research Database Industrial and Applied Microbiology Abstracts (Microbiology A) Meteorological & Geoastrophysical Abstracts - Academic AGRICOLA AGRICOLA - Academic |
| DatabaseTitleList | CrossRef AGRICOLA Civil Engineering Abstracts |
| Database_xml | – sequence: 1 dbid: 24P name: Wiley Online Library Open Access url: https://authorservices.wiley.com/open-science/open-access/browse-journals.html sourceTypes: Publisher |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Geography Economics |
| EISSN | 1944-7973 |
| EndPage | n/a |
| ExternalDocumentID | 10_1029_2022WR033091 WRCR26485 |
| Genre | researchArticle |
| GrantInformation_xml | – fundername: California Department of Water Resources funderid: 4600013361; 4600014294 – fundername: NSF Division of Earth Sciences funderid: 2124923 – fundername: U.S. Army Corps of Engineers Engineer Research and Development Center funderid: USACE W912HZ‐15‐2‐0019 |
| GroupedDBID | -~X ..I .DC 05W 0R~ 123 1OB 1OC 24P 31~ 33P 50Y 5VS 6TJ 7WY 7XC 8-1 8CJ 8FE 8FG 8FH 8FL 8G5 8R4 8R5 8WZ A6W AAESR AAHBH AAIHA AAIKC AAMMB AAMNW AANHP AANLZ AASGY AAXRX AAYCA AAYJJ AAYOK AAZKR ABCUV ABJCF ABJNI ABPPZ ABUWG ACAHQ ACBWZ ACCMX ACCZN ACGFO ACGFS ACIWK ACKIV ACNCT ACPOU ACPRK ACRPL ACXBN ACXQS ACYXJ ADBBV ADEOM ADKYN ADMGS ADNMO ADOZA ADXAS ADXHL ADZMN AEFGJ AEIGN AENEX AETEA AEUYN AEUYR AFBPY AFGKR AFKRA AFRAH AFWVQ AFZJQ AGQPQ AGXDD AIDBO AIDQK AIDYY AIURR ALMA_UNASSIGNED_HOLDINGS ALUQN ALXUD AMYDB ASPBG ATCPS AVWKF AZFZN AZQEC AZVAB BDRZF BENPR BEZIV BFHJK BGLVJ BHPHI BKSAR BMXJE BPHCQ BRXPI CCPQU CS3 D0L D1J DCZOG DDYGU DPXWK DRFUL DRSTM DU5 DWQXO EBS EJD F5P FEDTE FRNLG G-S GNUQQ GODZA GROUPED_DOAJ GUQSH HCIFZ HVGLF HZ~ K60 K6~ L6V LATKE LEEKS LITHE LK5 LOXES LUTES LYRES M0C M2O M7R M7S MEWTI MSFUL MSSTM MVM MW2 MXFUL MXSTM MY~ O9- OHT OK1 P-X P2P P2W PALCI PATMY PCBAR PHGZM PHGZT PQBIZ PQBZA PQQKQ PROAC PTHSS PYCSY Q2X R.K RIWAO RJQFR ROL SAMSI SUPJJ TAE TN5 TWZ UQL VJK VOH WBKPD WXSBR XOL XSW YHZ YV5 ZCG ZY4 ZZTAW ~02 ~KM ~OA ~~A AAYXX AFFHD AIQQE CITATION PQGLB WIN 7QH 7QL 7T7 7TG 7U9 7UA 8FD C1K F1W FR3 H94 H96 KL. KR7 L.G M7N P64 7S9 L.6 |
| ID | FETCH-LOGICAL-a4015-e2b46fe9bae981de7a111ede0172cad77f7e6b42cb0e4c6e1a58db49f1ff8c093 |
| IEDL.DBID | 24P |
| ISICitedReferencesCount | 23 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000949520600001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0043-1397 |
| IngestDate | Fri Sep 05 17:23:12 EDT 2025 Wed Aug 13 03:01:24 EDT 2025 Tue Nov 18 21:00:07 EST 2025 Sat Nov 29 01:36:51 EST 2025 Sun Jul 06 04:45:27 EDT 2025 |
| IsDoiOpenAccess | true |
| IsOpenAccess | true |
| IsPeerReviewed | true |
| IsScholarly | true |
| Issue | 3 |
| Language | English |
| License | Attribution |
| LinkModel | DirectLink |
| MergedId | FETCHMERGED-LOGICAL-a4015-e2b46fe9bae981de7a111ede0172cad77f7e6b42cb0e4c6e1a58db49f1ff8c093 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 content type line 23 |
| ORCID | 0000-0002-9330-9730 0000-0002-2305-2650 |
| OpenAccessLink | https://onlinelibrary.wiley.com/doi/abs/10.1029%2F2022WR033091 |
| PQID | 2810749720 |
| PQPubID | 105507 |
| PageCount | 19 |
| ParticipantIDs | proquest_miscellaneous_2811990643 proquest_journals_2810749720 crossref_citationtrail_10_1029_2022WR033091 crossref_primary_10_1029_2022WR033091 wiley_primary_10_1029_2022WR033091_WRCR26485 |
| PublicationCentury | 2000 |
| PublicationDate | March 2023 2023-03-00 20230301 |
| PublicationDateYYYYMMDD | 2023-03-01 |
| PublicationDate_xml | – month: 03 year: 2023 text: March 2023 |
| PublicationDecade | 2020 |
| PublicationPlace | Washington |
| PublicationPlace_xml | – name: Washington |
| PublicationTitle | Water resources research |
| PublicationYear | 2023 |
| Publisher | John Wiley & Sons, Inc |
| Publisher_xml | – name: John Wiley & Sons, Inc |
| References | 2021; 25 2010; 55 1986; 31 2019; 55 1997; 47 2016; 31 2016; 30 2020; 56 2003; 19 2011; 15 2007; 30 2001 2017; 36 2020; 51 2002; 268 2019; 23 2022; 36 2008; 22 2007; 1 2018; 32 2021; 8 2015; 19 2012; 420 2012 2020; 583 1991; 33 2017; 21 2021b 2016; 52 2021; 144 1996; 10 2021a; 138 2021; 57 2017; 53 2015; 28 2006; 307 2021; 12 2022 2021 2020 2022; 7 1994; 120 2015; 65 2018 2017 2016 2017; 18 2014 2020; 22 2018; 54 e_1_2_8_28_1 e_1_2_8_24_1 Lees T. (e_1_2_8_30_1) 2021 e_1_2_8_47_1 McMillan H. K. (e_1_2_8_34_1) 2021 e_1_2_8_26_1 e_1_2_8_49_1 e_1_2_8_3_1 e_1_2_8_5_1 Mathevet T. (e_1_2_8_32_1) 2006 e_1_2_8_7_1 e_1_2_8_9_1 e_1_2_8_43_1 e_1_2_8_22_1 e_1_2_8_45_1 e_1_2_8_64_1 e_1_2_8_62_1 e_1_2_8_41_1 Thornton P. E. (e_1_2_8_55_1) 2012 e_1_2_8_60_1 e_1_2_8_17_1 e_1_2_8_19_1 e_1_2_8_13_1 e_1_2_8_36_1 e_1_2_8_59_1 e_1_2_8_15_1 e_1_2_8_38_1 e_1_2_8_57_1 Hoedt P.‐J. (e_1_2_8_20_1) 2021 e_1_2_8_11_1 e_1_2_8_53_1 e_1_2_8_51_1 e_1_2_8_29_1 Goodfellow I. (e_1_2_8_18_1) 2016 e_1_2_8_25_1 e_1_2_8_46_1 e_1_2_8_27_1 e_1_2_8_48_1 e_1_2_8_2_1 e_1_2_8_4_1 e_1_2_8_6_1 e_1_2_8_8_1 e_1_2_8_42_1 e_1_2_8_23_1 e_1_2_8_44_1 e_1_2_8_65_1 e_1_2_8_63_1 e_1_2_8_40_1 e_1_2_8_61_1 e_1_2_8_39_1 e_1_2_8_14_1 e_1_2_8_35_1 e_1_2_8_16_1 e_1_2_8_37_1 e_1_2_8_58_1 Jaderberg M. (e_1_2_8_21_1) 2015; 28 e_1_2_8_10_1 e_1_2_8_31_1 e_1_2_8_56_1 e_1_2_8_12_1 e_1_2_8_33_1 e_1_2_8_54_1 e_1_2_8_52_1 e_1_2_8_50_1 |
| References_xml | – volume: 54 start-page: 8558 issue: 11 year: 2018 end-page: 8593 article-title: A transdisciplinary review of deep learning research and its relevance for water Resources Scientists publication-title: Water Resources Research – volume: 15 start-page: 2895 issue: 9 year: 2011 end-page: 2911 article-title: Catchment classification: Empirical analysis of hydrologic similarity based on catchment function in the eastern USA publication-title: Hydrology and Earth System Sciences – volume: 7 issue: 71 year: 2022 article-title: NeuralHydrology — A Python library for deep learning research in hydrology publication-title: Journal of Open Source Software – volume: 23 start-page: 5089 issue: 12 year: 2019 end-page: 5110 article-title: Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large‐sample datasets publication-title: Hydrology and Earth System Sciences – volume: 138 year: 2021a article-title: TOSSH: A toolbox for streamflow signatures in hydrology publication-title: Environmental Modelling & Software – volume: 31 start-page: 4757 issue: 26 year: 2016 end-page: 4761 article-title: Five guidelines for selecting hydrological signatures publication-title: Hydrological Processes – year: 2021b article-title: TOSSHtoolbox/TOSSH: First release – Minor update (1.0.1) publication-title: Zenodo – volume: 420 start-page: 171 year: 2012 end-page: 182 article-title: A review of efficiency criteria suitable for evaluating low‐flow simulations publication-title: Journal of Hydrology – start-page: 284 year: 2018 end-page: 299 – start-page: 2514 year: 2020 end-page: 2523 – volume: 53 start-page: 8020 issue: 9 year: 2017 end-page: 8040 article-title: Towards seamless large‐domain parameter estimation for hydrologic models publication-title: Water Resources Research – volume: 21 start-page: 5293 issue: 10 year: 2017 end-page: 5313 article-title: The CAMELS data set: Catchment attributes and meteorology for large‐sample studies publication-title: Hydrology and Earth System Sciences – volume: 144 year: 2021 – volume: 57 issue: 3 year: 2021 article-title: What role does hydrological science play in the age of machine learning? publication-title: Water Resources Research – volume: 47 start-page: 769 issue: 11 year: 1997 end-page: 784 article-title: The natural flow regime publication-title: BioScience – year: 2014 – volume: 33 start-page: 161 issue: 2 year: 1991 end-page: 174 article-title: Factorial sampling plans for preliminary computational experiments publication-title: Technometrics – volume: 120 start-page: 485 issue: 4 year: 1994 end-page: 504 article-title: Flow‐duration curves. I: New interpretation and confidence intervals publication-title: Journal of Water Resources Planning and Management – volume: 55 start-page: 171 issue: 1 year: 2010 end-page: 193 article-title: Classification of natural flow regimes in Australia to support environmental flow management publication-title: Freshwater Biology – volume: 56 issue: 9 year: 2020 article-title: Enhancing streamflow forecast and extracting insights using long‐short term memory networks with data integration at continental scales publication-title: Water Resources Research – volume: 55 start-page: 11344 issue: 12 year: 2019 end-page: 11354 article-title: Toward improved predictions in ungauged basins: Exploiting the power of machine learning publication-title: Water Resour. Res. – year: 2022 – volume: 32 start-page: 1120 issue: 8 year: 2018 end-page: 1125 article-title: Upper and lower benchmarks in hydrological modelling publication-title: Hydrological Processes – volume: 10 start-page: 1163 issue: 4 year: 1996 end-page: 1174 article-title: A method for assessing hydrologic alteration within ecosystems publication-title: Conservation Biology – volume: 30 start-page: 4019 issue: 22 year: 2016 end-page: 4035 article-title: Snow hydrology signatures for model identification within a limits‐of‐acceptability approach publication-title: Hydrological Processes – volume: 268 start-page: 244 issue: 1–4 year: 2002 end-page: 258 article-title: The use of indices of flow variability in assessing the hydrological and instream habitat impacts of upland afforestation and drainage publication-title: Journal of Hydrology – volume: 25 start-page: 2045 issue: 4 year: 2021 end-page: 2062 article-title: Rainfall–runoff prediction at multiple timescales with a single Long Short‐Term Memory network publication-title: Hydrology and Earth System Sciences – start-page: 1189 year: 2001 end-page: 1232 article-title: Greedy function approximation: A gradient boosting machine publication-title: Annals of Statistics – volume: 57 start-page: 885 issue: 6 year: 2021 end-page: 905 article-title: Post‐processing the national water model with long short‐term memory networks for streamflow predictions and model diagnostics publication-title: JAWRA Journal of the American Water Resources Association – volume: 54 start-page: 4059 issue: 6 year: 2018 end-page: 4083 article-title: Signature‐domain calibration of hydrological models using approximate Bayesian computation: Theory and comparison to existing applications publication-title: Water Resources Research – start-page: 224 year: 2018 end-page: 236 – volume: 22 start-page: 3802 issue: 18 year: 2008 end-page: 3813 article-title: Reconciling theory with observations: Elements of a diagnostic approach to model evaluation publication-title: Hydrological Processes: An International Journal – volume: 583 year: 2020 article-title: Exploring a long short‐term memory based encoder‐decoder framework for multi‐step‐ahead flood forecasting publication-title: Journal of Hydrology – volume: 65 start-page: 963 issue: 10 year: 2015 end-page: 972 article-title: Functional flows in modified riverscapes: Hydrographs, habitats and opportunities publication-title: BioScience – volume: 18 start-page: 2215 issue: 8 year: 2017 end-page: 2225 article-title: Benchmarking of a physically based hydrologic model publication-title: Journal of Hydrometeorology – start-page: 1578 year: 2017 end-page: 1585 – volume: 51 start-page: 1136 issue: 5 year: 2020 end-page: 1149 article-title: The exploration of a temporal convolutional network combined with encoder‐decoder framework for runoff forecasting publication-title: Hydrology Research – volume: 55 start-page: 4364 issue: 5 year: 2019 end-page: 4392 article-title: Flow prediction in ungauged catchments using probabilistic random forests regionalization and new statistical adequacy tests publication-title: Water Resources Research – volume: 307 start-page: 211 year: 2006 – volume: 55 start-page: 147 issue: 1 year: 2010 end-page: 170 article-title: The ecological limits of hydrologic alteration (ELOHA): A new framework for developing regional environmental flow standards publication-title: Freshwater Biology – volume: 52 start-page: 1847 issue: 3 year: 2016 end-page: 1865 article-title: Uncertainty in hydrological signatures for gauged and ungauged catchments publication-title: Water Resources Research – year: 2020 article-title: Signatures of hydrologic function across the critical zone observatory network publication-title: Water Resources Research – volume: 54 start-page: 8792 issue: 11 year: 2018 end-page: 8812 article-title: A ranking of hydrological signatures based on their predictability in space publication-title: Water Resources Research – year: 2016 – year: 2012 article-title: Daymet: Daily surface weather on a 1 km grid for North America, 1980‐2008. Oak ridge National Laboratory. ORNL Distribution Act publication-title: Archive Centre for Biogeochemical Dynamics. DAAC – volume: 36 start-page: 927 issue: 4 year: 2017 end-page: 940 article-title: Biological relevance of streamflow metrics: Regional and national perspectives publication-title: Freshwater Science – year: 2020 article-title: Linking hydrologic signatures to hydrologic processes: A review publication-title: Hydrological Processes – start-page: 1 year: 2021 end-page: 37 article-title: Hydrological concept formation inside long short‐term memory (LSTM) networks publication-title: Hydrology and Earth System Sciences Discussions – volume: 8 year: 2021 – volume: 31 start-page: 13 issue: 1 year: 1986 end-page: 24 article-title: Operational testing of hydrological simulation models publication-title: Hydrological Sciences Journal – volume: 22 start-page: 541 issue: 3 year: 2020 end-page: 561 article-title: Deep learning convolutional neural network in rainfall–runoff modelling publication-title: Journal of Hydroinformatics – volume: 36 issue: 6 year: 2022 article-title: Coevolution of machine learning and process‐based modelling to revolutionize Earth and environmental sciences: A perspective publication-title: Hydrological Processes – volume: 12 issue: 1 year: 2021 article-title: From calibration to parameter learning: Harnessing the scaling effects of big data in geoscientific modeling publication-title: Nature Communications – start-page: 4275 year: 2021 end-page: 4286 – volume: 19 start-page: 101 issue: 2 year: 2003 end-page: 121 article-title: Redundancy and the choice of hydrologic indices for characterizing streamflow regimes publication-title: River Research and Applications – year: 2022 article-title: mcmillanhk/HydroML: V1.0 Initial Release publication-title: Zenodo – volume: 19 start-page: 3951 issue: 9 year: 2015 end-page: 3968 article-title: Uncertainty in hydrological signatures publication-title: Hydrology and Earth System Sciences – volume: 28 year: 2015 article-title: Spatial transformer networks publication-title: Advances in Neural Information Processing Systems – volume: 30 start-page: 1756 issue: 8 year: 2007 end-page: 1774 article-title: Regionalization of constraints on expected watershed response behavior for improved predictions in ungauged basins publication-title: Advances in Water Resources – volume: 1 start-page: 901 issue: 4 year: 2007 end-page: 931 article-title: Catchment classification and hydrologic similarity publication-title: Geography Compass – volume: 19 start-page: 209 issue: 1 year: 2015 end-page: 223 article-title: Development of a large‐sample watershed‐scale hydrometeorological data set for the contiguous USA: Data set characteristics and assessment of regional variability in hydrologic model performance publication-title: Hydrology and Earth System Sciences – ident: e_1_2_8_7_1 doi: 10.3115/v1/D14-1179 – ident: e_1_2_8_47_1 doi: 10.1007/978-3-030-01246-5_18 – ident: e_1_2_8_11_1 doi: 10.1029/2022WR032404 – ident: e_1_2_8_65_1 doi: 10.1093/biosci/biv102 – ident: e_1_2_8_16_1 doi: 10.1016/j.envsoft.2021.104983 – ident: e_1_2_8_44_1 doi: 10.1111/j.1365-2427.2009.02204.x – ident: e_1_2_8_63_1 doi: 10.1029/2019WR026635 – start-page: 4275 volume-title: International Conference on Machine Learning year: 2021 ident: e_1_2_8_20_1 – ident: e_1_2_8_26_1 doi: 10.1080/02626668609491024 – ident: e_1_2_8_27_1 doi: 10.21105/joss.04050 – ident: e_1_2_8_45_1 doi: 10.1029/2018WR023254 – ident: e_1_2_8_24_1 doi: 10.1111/j.1365-2427.2009.02307.x – year: 2012 ident: e_1_2_8_55_1 article-title: Daymet: Daily surface weather on a 1 km grid for North America, 1980‐2008. Oak ridge National Laboratory. ORNL Distribution Act publication-title: Archive Centre for Biogeochemical Dynamics. DAAC – ident: e_1_2_8_13_1 doi: 10.31223/X5BH0P – ident: e_1_2_8_33_1 doi: 10.1002/hyp.13632 – ident: e_1_2_8_51_1 doi: 10.5194/hess-15-2895-2011 – ident: e_1_2_8_37_1 doi: 10.1080/00401706.1991.10484804 – ident: e_1_2_8_23_1 doi: 10.1002/2017wr020528 – ident: e_1_2_8_14_1 doi: 10.1214/aos/1013203451 – ident: e_1_2_8_60_1 doi: 10.1109/IJCNN.2017.7966039 – volume: 28 year: 2015 ident: e_1_2_8_21_1 article-title: Spatial transformer networks publication-title: Advances in Neural Information Processing Systems – ident: e_1_2_8_8_1 doi: 10.1109/CVPR42600.2020.00259 – ident: e_1_2_8_12_1 doi: 10.1111/1752-1688.12964 – ident: e_1_2_8_61_1 doi: 10.5194/hess-19-3951-2015 – ident: e_1_2_8_29_1 doi: 10.5194/hess-23-5089-2019 – ident: e_1_2_8_15_1 doi: 10.5194/hess-25-2045-2021 – ident: e_1_2_8_52_1 doi: 10.1002/hyp.10972 – ident: e_1_2_8_2_1 doi: 10.1029/2018wr022606 – ident: e_1_2_8_39_1 doi: 10.5194/hess-19-209-2015 – ident: e_1_2_8_31_1 doi: 10.2166/nh.2020.100 – ident: e_1_2_8_35_1 doi: 10.1002/hyp.11300 – ident: e_1_2_8_42_1 – start-page: e1499 volume-title: A review of hydrologic signatures and their applications year: 2021 ident: e_1_2_8_34_1 – ident: e_1_2_8_57_1 doi: 10.2166/hydro.2020.095 – ident: e_1_2_8_62_1 doi: 10.1002/2015wr017635 – ident: e_1_2_8_3_1 doi: 10.5194/hess-21-5293-2017 – ident: e_1_2_8_5_1 doi: 10.5281/zenodo.6712302 – ident: e_1_2_8_22_1 doi: 10.1016/j.jhydrol.2020.124631 – ident: e_1_2_8_54_1 doi: 10.1029/2018WR022643 – ident: e_1_2_8_17_1 doi: 10.5281/zenodo.6462813 – ident: e_1_2_8_58_1 doi: 10.1061/(asce)0733-9496(1994)120:4(485) – ident: e_1_2_8_25_1 – ident: e_1_2_8_19_1 doi: 10.1002/hyp.6989 – ident: e_1_2_8_38_1 doi: 10.1029/2020WR028091 – ident: e_1_2_8_28_1 doi: 10.1029/2019WR026065 – ident: e_1_2_8_10_1 doi: 10.1029/2019wr026793 – start-page: 1 year: 2021 ident: e_1_2_8_30_1 article-title: Hydrological concept formation inside long short‐term memory (LSTM) networks publication-title: Hydrology and Earth System Sciences Discussions – ident: e_1_2_8_64_1 doi: 10.1016/j.advwatres.2007.01.005 – ident: e_1_2_8_9_1 doi: 10.1109/CVPRW.2018.00060 – ident: e_1_2_8_56_1 doi: 10.1038/s41467-021-26107-z – ident: e_1_2_8_4_1 doi: 10.1016/s0022-1694(02)00171-3 – ident: e_1_2_8_46_1 doi: 10.1016/j.jhydrol.2011.11.055 – ident: e_1_2_8_36_1 doi: 10.1002/2017wr020401 – ident: e_1_2_8_50_1 doi: 10.1046/j.1523-1739.1996.10041163.x – ident: e_1_2_8_40_1 doi: 10.1175/jhm-d-16-0284.1 – ident: e_1_2_8_48_1 doi: 10.1016/j.envsoft.2021.105159 – ident: e_1_2_8_59_1 doi: 10.1111/j.1749-8198.2007.00039.x – ident: e_1_2_8_6_1 doi: 10.1086/694913 – ident: e_1_2_8_43_1 doi: 10.2307/1313099 – start-page: 211 volume-title: A bounded version of the Nash‐Sutcliffe criterion for better model assessment on large sets of basins year: 2006 ident: e_1_2_8_32_1 – ident: e_1_2_8_53_1 doi: 10.1002/hyp.11476 – ident: e_1_2_8_41_1 doi: 10.1002/rra.700 – ident: e_1_2_8_49_1 doi: 10.1002/hyp.14596 – volume-title: Deep learning year: 2016 ident: e_1_2_8_18_1 |
| SSID | ssj0014567 |
| Score | 2.5123506 |
| Snippet | Hydrologic signatures are quantitative metrics that describe a streamflow time series. Examples include annual maximum flow, baseflow index and recession shape... |
| SourceID | proquest crossref wiley |
| SourceType | Aggregation Database Enrichment Source Index Database Publisher |
| SubjectTerms | Base flow Climate Climatic data Coders Coefficients data collection deep learning Dynamics encoder–decoder flow indices Flow mapping flow metric Hydrologic models hydrologic signature Hydrology Information processing Learning algorithms Machine learning Maximum flow meteorological data Model accuracy Modelling Neural networks Outflow Precipitation prediction Rainfall-runoff relationships Runoff Runoff models Seasonal variations Seasonality Sensitivity analysis Signatures Stream discharge Stream flow time series analysis Watersheds |
| Title | Using Machine Learning to Identify Hydrologic Signatures With an Encoder–Decoder Framework |
| URI | https://onlinelibrary.wiley.com/doi/abs/10.1029%2F2022WR033091 https://www.proquest.com/docview/2810749720 https://www.proquest.com/docview/2811990643 |
| Volume | 59 |
| WOSCitedRecordID | wos000949520600001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVWIB databaseName: Wiley Online Library - Journals customDbUrl: eissn: 1944-7973 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014567 issn: 0043-1397 databaseCode: DRFUL dateStart: 19970101 isFulltext: true titleUrlDefault: https://onlinelibrary.wiley.com providerName: Wiley-Blackwell – providerCode: PRVWIB databaseName: Wiley Online Library Free Content customDbUrl: eissn: 1944-7973 dateEnd: 20231209 omitProxy: false ssIdentifier: ssj0014567 issn: 0043-1397 databaseCode: WIN dateStart: 19970101 isFulltext: true titleUrlDefault: https://onlinelibrary.wiley.com providerName: Wiley-Blackwell |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1fS9xAEB9EBX2xWhXPXmUL7ZMNXjabbPJYTg8FPeRaOR8KYbOZ6IHk5HIK99bv0G_oJ3F2sznPhwrFt4TMJmF359_OzG8AvoaBDCIyIzxSpsoTgtwdxfPQ41xHItQSk6xuNiH7_fj6Orl0B26mFqbGh5gfuBnOsPLaMLjKKgc2YDAyyWvnw0GH_HFTvL7i-4E0u5qLy3kUgYwD2USYjaXjEt9p_NHi6Ncq6cXOXLRWrbrpfXjvj27ChjM02Y96Z2zBEpYfYa2pQ67o2vU_v51tw2-bOcAubGYlMge6esOmY1ZX8hYzdjrLJ7WkZD9HNzUgaMWGo-ktUyU7KU1x_OTpz99jtFes16R97cBV7-RX99RzfRc8Rd5W6CHPRFTQKilMyJxFqUggYo7GXdQql7KQGGWC66yDQkfoqzDOM5EUflHEupMEu7BcjkvcAxZjLLEINPnZ9G5BIk0EsTSg-AGJSpG34LCZ-lQ7UHLTG-MutcFxnqSLs9eCb3Pq-xqM4x907WYVU8eSVcpjk3uaSN5pwZf5Y2ImEyFRJY4fLI1P6pmstBZ8t2v65nfS4aA7MCmC4f7_kX-CddO2vs5la8PydPKAn2FVP05H1eTAbuIDWDke9K7O6W541n8GwYjxUg |
| linkProvider | Wiley-Blackwell |
| linkToHtml | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1LT-MwEB6tykpw2QcPURZ2jbScIKJ1nDg5ImjVFW2FuqByQIocZ9JWQumqD6Te-A_8w_0lO06cUg4gob1ZysSJbM_MN54XwE_Pla5PMMIhZaocIcjcUTzxHM61LzwtMYyLZhOy2w1ub8Mr2-fU5MIU9SGWF26GM3J5bRjcXEjbagOmSCaZ7bzfq5FBbrLX1wSdJK8Caxe95k176UggfCBLJ7MBOzb2nWY4XX3_pVZ6hpqrgDXXOM3P__2vX-CTBZvsrDgdX-EDZpuwXuYiT2lse6APF1twl0cPsE4eXYnMFl4dsNmYFdm86YK1FsmkkJbs92hQFAWdsv5oNmQqY43MJMhP_j4-XWA-Ys0y9GsbbpqN6_OWY3svOIosLs9BHgs_pZ1SGBKkRalIKGKCxmTUKpEylejHguu4hkL7WFdekMQiTOtpGuha6O5AJRtnuAsswEBi6mqytWluQWJNuIE0hfFdEpciqcJxufaRtoXJTX-M-yh3kPMwWl29Khwtqf8UBTleodsvtzGybDmNeGDiT0PJa1U4XD4mhjJeEpXheJ7T1ElFE1Krwkm-qW9-J-r3znsmTNDbex_5D1hvXXfaUftX9_IbbJg29kVs2z5UZpM5HsBH_TAbTSff7Zn-B1hj9CA |
| linkToPdf | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV3NattAEB5CGtpc0vQnxImbbqA9tSL2aqWVjsU_OLQxxm2xDwWxWo1sQ5CD7RR8yzvkDfMknV2tXPfQQOlNoF1J7O7MfKOZ-QbgXeBLPyQY4ZExVZ4Q5O4ongUe5zoUgZYYp2WzCdnvR-NxPHB9Tk0tTMkPsfnhZiTD6msj4HiT5Y5twJBkktvOR8MGOeSmev2JCEjNGmpnMdiEEQgdyCrEbKCOy3yn-Rfbs_-0Sb-B5jZctfam-_y_v_QQDhzUZJ_Ks_ECdrB4Cc-qSuQlXbsO6NP1K_hhcwfYlc2tROZoVydsNWdlLW--Zr11tih1Jfs6m5SUoEs2mq2mTBWsU5jy-MXD3X0b7RXrVolfr-F7t_Ot1fNc5wVPkb8VeMhTEea0TwpjArQoFalEzNA4jFplUuYSw1RwnTZQ6BCbKoiyVMR5M88j3Yj9I9gt5gUeA4swkpj7mjxterYgpSb8SBpafJ-Upchq8KFa-0Q7WnLTHeM6seFxHifbq1eD95vRNyUdx1_G1attTJxQLhMemezTWPJGDc43t0mcTIxEFTi_tWOaZKAJp9Xgo93UR9-TjIatoUkSDE7-bfhbeDpod5Mvl_3Pp7BvetiXiW112F0tbvEN7Omfq9lycWYP9C-PUPIJ |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Using+Machine+Learning+to+Identify+Hydrologic+Signatures+With+an+Encoder%E2%80%93Decoder+Framework&rft.jtitle=Water+resources+research&rft.au=Botterill%2C+Tom+E.&rft.au=McMillan%2C+Hilary+K.&rft.date=2023-03-01&rft.issn=0043-1397&rft.eissn=1944-7973&rft.volume=59&rft.issue=3&rft.epage=n%2Fa&rft_id=info:doi/10.1029%2F2022WR033091&rft.externalDBID=10.1029%252F2022WR033091&rft.externalDocID=WRCR26485 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0043-1397&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0043-1397&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0043-1397&client=summon |