Watershed delineation

The conventional method of finding a watershed boundary is to draw it by hand on a paper topographic map, or on a transparent overlay. The watershed area can then be estimated using a planimeter, by overlaying graph paper and counting grid cells, or the result can be digitized for use with mapping software. The same process can be done on a computer, sketching the watershed boundary (with a mouse or stylus) over a digital copy of a topographic map. This is referred to as "heads up digitizing" or "on-screen digitizing."{{Cite conference| publisher = U.S. Department of Agriculture Forest Service, Southern Research Station| pages = 302| last1 = Maceyka| first1 = Andy| last2 = Hansen| first2 = William F| title = Enhancing hydrologic mapping using LIDAR and high resolution aerial photos on the Francis Marion National Forest in coastal South Carolina| book-title = Headwaters to Estuaries: Advances in Watershed Science and Management, Fifth Interagency Conference on Research in the Watersheds, March 2–5, 2015| location = North Charleston, South Carolina| date = 2016|url=https://www.srs.fs.usda.gov/pubs/gtr/gtr_srs211/gtr_srs211_012.pdf}}

File:Watershed nrcs.jpg

For "manual" watershed delination, one must know how to read and interpret a topographic map, for example to identify ridges, valleys, and the direction of steepest slope.{{Cite web |last1=U.S. Geological Survey|author1link=U.S. Geological Survey| title = To delineate a watershed you must identify land surface features from topographic contours|website=YouTube | accessdate = 2023-02-07| date = 2020-07-22| url = https://www.youtube.com/watch?v=kTO3lRfpKVs}} Even in the computer era, manual watershed delineation is still a useful skill, in order to check whether watersheds generated with software are correct.

Instructions for manual watershed delineation can be found in some textbooks in geography or environmental management, in government pamphlets,{{Citation| publisher = US Department of Agriculture, Natural Resources Conservation Service| last = USDA| title = How to Read a Topographic Map and Delineate a Watershed|date=n.d.|accessdate = 2023-02-07|url=https://bwsr.state.mn.us/sites/default/files/2020-05/MN_Watershed_Delineation.pdf}}{{cite web |last1=Illinois Environmental Protection Agency |title=Determining Your Lake's Watershed |url=http://www.epa.state.il.us/water/conservation/lake-notes/determining-watershed.pdf |access-date=20 January 2023}} or in online video tutorials.{{cite web| last1=U.S. Geological Survey|author1link=U.S. Geological Survey|title=Manual watershed delineation is a five-step process|website=YouTube |date=July 22, 2020|url=https://www.youtube.com/watch?v=IediKgLNMXA&t}}

According to the US Geological Survey, there are 5 steps to manual watershed delineation:

1. Find the point of interest along a stream on the map. This is the "watershed outlet" or "pour point."
2. Imagine or draw surface water flow lines that point downhill perpendicular to the topographic contours (this is the steepest direction).
3. Mark the location of topographical high points (peaks) around the stream.
4. Mark the points along contours that divide flows towards or away from the stream (ridges).
5. Connect the dots to delineate the watershed.

General Rules:

• The watershed boundary should be perpendicular to contour lines where it crosses them.
• The watershed boundary must not cross rivers or streams other than at the outlet. (In some cases, a blue line representing a man-made canal or pipeline may traverse your watershed boundary.)
• The watershed boundary should run along ridgelines and connect high points.

One disadvantage to manual watershed delineation is that it is subject to errors and the individual judgment of the analyst. The Illinois Environmental Protection Agency wrote, "bear in mind that delineating a watershed is an inexact science. Any two people, even if both are experts, will come up with slightly different boundaries."

Especially for smaller watersheds and when accurate results are important, field reconnaissance may be needed to find features that are not shown on maps. "Going out into the field allows you to identify human alterations, such as road ditches, storm sewers and culverts that could change the direction of waters flow and thus change the watershed boundaries."

Using computer software to delineate watersheds can be much faster than manual methods. It may also be more consistent, as it removes analyst's subjectivity. Automatic methods of watershed delineation have been in use since the 1980s, and are now in widespread use in the science and engineering communities. Researchers have even used computer methods to delineate watersheds on Mars.{{Cite journal| doi = 10.1002/hyp.3360050104| issn = 1099-1085| volume = 5| issue = 1| pages = 31–44| last = Jenson| first = Susan K.| title = Applications of hydrologic information automatically extracted from digital elevation models| journal = Hydrological Processes| accessdate = 2023-01-31| date = 1991| bibcode = 1991HyPr....5...31J| url = https://onlinelibrary.wiley.com/doi/abs/10.1002/hyp.3360050104}}

{{Cite journal| doi = 10.1029/2009JE003429| issn = 0148-0227| volume = 115| issue = E9| pages = –09001| last1 = Mest| first1 = Scott C.| last2 = Crown| first2 = David A.| last3 = Harbert| first3 = William| title = Watershed modeling in the Tyrrhena Terra region of Mars| journal = Journal of Geophysical Research| date = 2010-09-01| bibcode = 2010JGRE..115.9001M| s2cid = 140699479| doi-access = free}}

Automated watershed delineation methods use digital data of the earth's elevation, a Digital Elevation Model, or DEM. Typically, algorithms use the method of "steepest slope" to calculate the flow direction from a grid cell (or pixel) to one of its neighbors.{{Cite book| publisher = ESRI, Inc.| isbn = 978-1-879102-80-4| last1 = Maidment| first1 = David R.| last2 = Djokic| first2 = Dean| title = Hydrologic and Hydraulic Modeling Support: With Geographic Information Systems| date = 2000}}

It is possible to use DEMs in different formats for watershed delineation, such as a Triangular Irregular Network (TIN),{{Cite journal| doi = 10.1061/(ASCE)0733-9429(1990)116:10(1232)| issn = 0733-9429| volume = 116| issue = 10| pages = 1232–1251| last1 = Jones| first1 = Norman L.| last2 = Wright| first2 = Stephen G.| last3 = Maidment| first3 = David R.| title = Watershed Delineation with Triangle-Based Terrain Models| journal = Journal of Hydraulic Engineering| accessdate = 2023-02-08| date = 1990-10-01| url = https://ascelibrary.org/doi/10.1061/%28ASCE%290733-9429%281990%29116%3A10%281232%29}} or Hexagonal tiling{{Cite journal| doi = 10.1016/j.envsoft.2020.104702| issn = 1364-8152| volume = 128| pages = 104702| last1 = Liao| first1 = Chang| last2 = Tesfa| first2 = Teklu| last3 = Duan| first3 = Zhuoran| last4 = Leung| first4 = L. Ruby| title = Watershed delineation on a hexagonal mesh grid| journal = Environmental Modelling & Software| date = 2020-06-01| s2cid = 215844895| doi-access = free| bibcode = 2020EnvMS.12804702L}} however most contemporary algorithms make use of a regular rectangular grid.{{cite journal |last1=O'Callaghan |first1=John F. |last2=Mark |first2=David M. |title=The extraction of drainage networks from digital elevation data |journal=Computer Vision, Graphics, and Image Processing |date=December 1984 |volume=28 |issue=3 |pages=323–344 |doi=10.1016/S0734-189X(84)80011-0}} In the 1980s and 1990s, digital elevation models were often obtained by scanning and digitizing the contours on paper topographic maps, which were then converted to a TIN or a gridded DEM.{{Cite conference| publisher = US Geological Survey| pages = 30| last1 = Wiche| first1 = Gregg J.| last2 = Jenson| first2 = S. K.| last3 = Baglio| first3 = J. V.| last4 = Domingue| first4 = Julia O.| title = Application of digital elevation models to delineate drainage areas and compute hydrologic characteristics for sites in the James River Basin, North Dakota| date = 1990|url=https://pubs.usgs.gov/wsp/2383/report.pdf}} More recently, the DEM is obtained by aerial or satellite remote sensing, using stereophotogrammetry, lidar, or radar.{{Cite book| publisher = CRC press| isbn = 978-0-429-20507-1| last1 = Li| first1 = Zhilin| last2 = Zhu| first2 = Christopher| last3 = Gold| first3 = Chris| title = Digital terrain modeling: principles and methodology| date = 2004|page=Chapter 3}}

To use a rectangular grid DEM for watershed delineation, it must first be processed or "conditioned" in order to return realistic results. The result is sometimes referred to as a "hydro-enforced" DEM or a "HydroDEM." Most of the software packages listed below can perform these functions on a "raw" DEM, or analysts can download hydrologically-conditioned DEMs such as the near-global HydroSHEDS, MERIT-Hydro, or EDNA {{Citation| publisher = U.S. Geological Survey| last = Earth Resources Observation And Science (EROS) Center| title = Elevation Derivatives for National Applications (EDNA) Seamless Three-Dimensional Hydrologic Database| accessdate = 2023-02-21| date = 2017| url = https://www.usgs.gov/centers/eros/science/usgs-eros-archive-digital-elevation-elevation-derivatives-national-applications}} for the continental United States. The usual steps for hydrologic conditioning of a DEM are:

1. Fill sinks.
2. "Burn in" the stream channels.
3. Calculate flow direction.
4. Calculate flow accumulation.

File:DEM conditioning example.jpg (NHDPlus).]]

Additionally, some methods allow for "fencing ridgelines" and burning in flow pathways through lakes.{{Cite book| publisher = ESRI, Inc.| isbn = 978-1-879102-80-4| last1 = Maidment| first1 = David R.| last2 = Djokic| first2 = Dean| title = Hydrologic and Hydraulic Modeling Support: With Geographic Information Systems| year= 2000}} Some methods also enforce a small slope onto flat areas so that flow will continue to move toward the outlet.{{Cite journal| doi = 10.1016/0098-3004(92)90007-E| issn = 0098-3004| volume = 18| issue = 6| pages = 747–761| last1 = Martz| first1 = L.W.| last2 = Garbrecht| first2 = J.| title = Numerical definition of drainage network and subcatchment areas from Digital Elevation Models| journal = Computers & Geosciences| accessdate = 2022-06-17| date = July 1992| bibcode = 1992CG.....18..747M| url = https://linkinghub.elsevier.com/retrieve/pii/009830049290007E}} The step of "burning in" stream channels involves artificially deepening the channel, by subtracting a large elevation value from pixels that represent the channel. This ensures that once flow has entered the channel, it will stay there rather than jumping out and flowing overland or into another channel. Some algorithms infer the location of channels automatically from the DEM. Better results are usually obtained by burning in mapped stream channels, or channels derived from satellite or aerial imagery.{{Cite journal| doi = 10.1002/esp.3888| issn = 1096-9837| volume = 41| issue = 5| pages = 658–668| last = Lindsay| first = John B.| title = The practice of DEM stream burning revisited| journal = Earth Surface Processes and Landforms| accessdate = 2023-02-06| date = 2016| bibcode = 2016ESPL...41..658L| s2cid = 130330511| url = https://onlinelibrary.wiley.com/doi/abs/10.1002/esp.3888}}

There are several different algorithms available for calculating flow direction from a DEM. The first method, introduced by Australian geographers O'Callaghan and Mark in 1984, is referred to as D8. Water flows from a pixel to one of 8 possible directions to a neighboring cell (including diagonally), based on the direction of steepest slope. There are disadvantages to this method as water flow is limited to 8 directions, separated by 45°, which may result in unrealistic flow patterns. Also, because all of the flow is routed in one direction, the D8 method is unable to model situations where the flow diverges, such as on convex hillsides, in a river delta, or in branched or braided rivers. Alternative algorithms have been proposed and implemented to overcome this limitation, such as D∞.{{cite journal |last1=Tarboton |first1=David G. |title=A new method for the determination of flow directions and upslope areas in grid digital elevation models |journal=Water Resources Research |date=February 1997 |volume=33 |issue=2 |pages=309–319 |doi=10.1029/96WR03137|bibcode=1997WRR....33..309T |s2cid=6142858 |doi-access=free }} Nevertheless, the D8 algorithm remains in widespread use, and has been used to create important datasets such as HydroBasins {{Cite journal| doi = 10.1029/2008EO100001| issn = 2324-9250| volume = 89| issue = 10| pages = 93–94| last1 = Lehner| first1 = Bernhard| last2 = Verdin| first2 = Kristine| last3 = Jarvis| first3 = Andy| title = New Global Hydrography Derived From Spaceborne Elevation Data| journal = Eos, Transactions American Geophysical Union| date = 2008| bibcode = 2008EOSTr..89...93L| doi-access = free}} and MERIT-Basins.

Computerized watershed delineation is not always correct. Some errors stem from incorrectly placing the watershed outlet on the digital river network, or "snapping the pour point."{{Cite journal| volume = 54| issue = 11| pages = 1593–1600| last1 = Jenson| first1 = Susan K.| last2 = Domingue| first2 = Julia O.| title = Extracting topographic structure from digital elevation data for geographic information system analysis| journal = Photogrammetric Engineering and Remote Sensing| date = 1988|url=https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=06a20725ae38b4dce81951bbb230b197dd346daa}} Another class of errors stems from inaccuracies in the digital terrain data, or where its resolution is too coarse to capture flow pathways. In general, DEMs with higher spatial resolution can more realistically describe topography of the land surface and flow direction. However, there is a tradeoff, as a finer grid with more pixels increases computing time.{{Cite journal| doi = 10.5194/hess-13-2241-2009| issn = 1607-7938| volume = 13| issue = 11| pages = 2241–2251| last1 = Yamazaki| first1 = D.| last2 = Oki| first2 = T.| last3 = Kanae| first3 = S.| title = Deriving a global river network map and its sub-grid topographic characteristics from a fine-resolution flow direction map| journal = Hydrology and Earth System Sciences| accessdate = 2022-06-22| date = 2009-11-26| bibcode = 2009HESS...13.2241Y| url = https://hess.copernicus.org/articles/13/2241/2009/| doi-access = free}} Nevertheless, even high-resolution data may not adequately capture flow pathways in complex environments like cities and suburbs, where flow is directed by curbs, culverts, and storm drains.{{Cite journal| doi = 10.1002/hyp.13229| issn = 1099-1085| volume = 32| issue = 18| pages = 2845–2855| last1 = Kayembe| first1 = Aimé| last2 = Mitchell| first2 = Carl P. J.| title = Determination of subcatchment and watershed boundaries in a complex and highly urbanized landscape| journal = Hydrological Processes| accessdate = 2022-10-31| date = 2018| bibcode = 2018HyPr...32.2845K| s2cid = 133665805| url = http://onlinelibrary.wiley.com/doi/abs/10.1002/hyp.13229}} Finally, some errors can result from the algorithm or the choice of parameters.{{Cite journal| doi = 10.2166/nh.2022.126| issn = 0029-1277| volume = 53| issue = 8| pages = 1047–1062| last1 = Datta| first1 = Srijon| last2 = Karmakar| first2 = Shyamal| last3 = Mezbahuddin| first3 = Symon| last4 = Hossain| first4 = Mohammad Mozaffar| last5 = Chaudhary| first5 = B. S.| last6 = Hoque| first6 = Md. Enamul| last7 = Abdullah Al Mamun| first7 = M. M.| last8 = Baul| first8 = Tarit Kumar| title = The limits of watershed delineation: implications of different DEMs, DEM resolutions, and area threshold values| journal = Hydrology Research| date = 2022-07-15| s2cid = 250590129| doi-access = free}}

Because errors are common, some authorities insist that the results of automated delineation must be carefully checked. The US Geological Survey's standards for the US Watershed Boundary Dataset allow the use of software "to generate intermediate or “draft” boundary lines," which then must be verified by the analyst by overlaying them on a computer display over basemaps (scanned topographic maps, aerial photographs) to verify their accuracy.{{cite book |title=Federal guidelines, requirements, and procedures for the national Watershed Boundary Dataset |date=2009 |publisher=U.S. Geological Survey |url=https://pubs.usgs.gov/tm/11/a3/pdf/tm11-a3_1ed.pdf |ref=Techniques and Methods 11–A3}}

Some of the first watershed delineation software was written in FORTRAN, such as CATCH {{cite journal| doi = 10.1016/0098-3004(88)90018-0| issn = 0098-3004| volume = 14| issue = 5| pages = 627–640| last1 = Martz| first1 = Lawrence W.| last2 = Jong| first2 = Eeltje de| title = CATCH: A FORTRAN program for measuring catchment area from digital elevation models| journal = Computers & Geosciences| accessdate = 2022-06-05| date = January 1988| bibcode = 1988CG.....14..627M| url = https://linkinghub.elsevier.com/retrieve/pii/0098300488900180}} and DEDNM.

Watershed delineation tools are a part of several Geographic Information System software packages such as ArcGIS, QGIS, and GRASS GIS. There are standalone programs for watershed delineation such as TauDEM. Watershed delineation tools are also incorporated into some hydrologic modeling software packages.

Software developers have also published libraries or modules in several languages (see list below). Many of these packages are free and open source, which means they can be expanded or adapted by those willing and able to write or modify code. Finally, there are web applications for delineating watersheds. Some of these web apps have extra features for science and engineering like calculating flow statistics or watershed land cover types (e.g.: StreamStats, Model My Watershed).

=== Standalone watershed delineation software ===

• TauDEM,{{Citation| publisher = Utah State University, Hydrology Research Group| last = David Tarbotton| title = Terrain Analysis Using Digital Elevation Models (TauDEM)| accessdate = 2023-02-21| date = 2016| url = https://hydrology.usu.edu/taudem/taudem5/}} Toolbox for ArcGIS, or command line executable for Windows.
• TOPAZ,{{Citation| publisher = United States Department of Agriculture| last = USDA| title = TOPAZ| accessdate = 2023-02-21| date = 1999| url = https://www.ars.usda.gov/ARSUserFiles/30700510/TOPAZ_Table-of-Contents3.pdf}} from the US Department of Agriculture, Windows executable.

• WMS (hydrology software)
• SWAT model
• BASINS{{Citation| publisher = United States Environmental Protection Agency, Office of Research and Development| last = US EPA| title = Better Assessment Science Integrating Point and Non-point Sources (BASINS)| accessdate = 2023-02-21| date = 2015-07-23| url = https://www.epa.gov/ceam/better-assessment-science-integrating-point-and-non-point-sources-basins}}
• BasinMaker,{{Citation| publisher = University of Waterloo| last1 = M. Han| last2 = H. Shen| last3 = B. A. Tolson| last4 = J. R. Craig| last5 = J. Mai| last6 = S. Lin| last7 = N. B. Basu| last8 = F. Awol| title = BasinMaker 3.0: a GIS toolbox for distributed watershed delineation of complex lake-river routing networks| journal = Environmental Modelling & Software| accessdate = 2023-02-21| date = 2023| volume = 164| doi = 10.1016/j.envsoft.2023.105688| bibcode = 2023EnvMS.16405688H| url = http://hydrology.uwaterloo.ca/basinmaker/index.html| doi-access = free}} for use with the RAVEN software suite, for parts of US and Canada only
• WEAP Model

• ESRI ArcGIS or the ArcGIS online web application
• GRASS GIS, modules r.water, r.watershed, and r.stream
• QGIS
• SAGA GIS
• Whitebox Geospatial Analysis Tools
• ILWIS
• TerrSet (formerly IDRISI)
• TNTmips
• MapWindow GIS
• Manifold System
• {{section link|Blue Marble Geographics#Global Mapper}}

• Global Watersheds,{{Cite web| last = Heberger| first = Matthew| title = Global Watersheds App| work = mghydro.com| accessdate = 2023-02-21| date = 2023| url = https://mghydro.com/watersheds}} web app and API
• Stream Stats,{{Cite web| title = StreamStats| work = U.S. Geological Survey| accessdate = 2023-02-21| date = 2021| url = https://streamstats.usgs.gov/ss/}} from the US Geological Survey, allows you to delineate watersheds in the US only
• Model My Watershed,{{Cite web| last = Stroud Water Research Center| title = Model My Watershed| work = WikiWatershed| accessdate = 2023-02-21| date = 2023| url = https://modelmywatershed.org/draw}} by the Stroud Water Research Center, US only, can delineate watersheds based on an outlet point, and perform analyses related to water quality
• Ontario Watershed Information Tool,{{Cite web| title = Ontario Watershed Information Tool (OWIT) {{!}} ontario.ca| work = Government of Ontario| accessdate = 2023-02-21| date = 2022| url = http://www.ontario.ca/page/ontario-watershed-information-tool-owit}} for the province of Ontario in Canada

There are a number of vector datasets representing watersheds as polygons that can be displayed and analyzed with GIS or other software. In these datasets, the entire land surface is divided into "subwatersheds" or "unit catchments." Individual unit watersheds can be combined or merged to find larger watersheds. The unit catchments have linked hydrological code data or similar metadata to create a flow network, so flow pathways and connections can be determined via network analysis.{{Citation| publisher = McGill University| last = Lehner| first = Bernhard| title = HydroBASINS Technical Documentation Version 1.c| accessdate = 2023-02-16| year = 2014| url = https://data.hydrosheds.org/file/technical-documentation/HydroBASINS_TechDoc_v1c.pdf}}

File:HUC subregions.png

This list is non-exhaustive, as many organizations and territories have produced their own watershed map data and have published via the web. Notable datasets include:

• United States Watershed Boundary Dataset,{{Cite web| title = Watershed Boundary Dataset| work = U.S. Geological Survey| accessdate = 2023-02-19| url = https://www.usgs.gov/national-hydrography/watershed-boundary-dataset}} [https://www.usgs.gov/national-hydrography/watershed-boundary-dataset website] (continually updated)
• Canadian National Hydrographic Network Watershed Boundaries,{{Cite web| last = Natural Resources Canada| title = National Hydrographic Network| accessdate = 2023-02-19| date = 2018-09-10| url = https://natural-resources.canada.ca/science-and-data/science-and-research/earth-sciences/geography/topographic-information/geobase-surface-water-program-geeau/national-hydrographic-network/21361}} [https://www.nrcan.gc.ca/science-and-data/science-and-research/earth-sciences/geography/topographic-information/geobase-surface-water-program-geeau/watershed-boundaries/20973 website]
• HydroBasins {{Cite journal| doi = 10.1002/2017GL072874| volume = 27| issue = 15| pages = 2171–2186| last1 = Lehner| first1 = Bernhard| last2 = Grill| first2 = Günther| title = Global river hydrography and network routing: baseline data and new approaches to study the world's large river systems| journal = Hydrological Processes| date = 2013| s2cid = 134761792| hdl = 1983/5567093b-7d5e-4ee9-988d-b0bf242c16f3| hdl-access = free}} [https://www.hydrosheds.org/products/hydrobasins website] (global, 2013)
• Hydrologic Derivatives for Modeling and Applications (HDMA),{{Cite conference| publisher = U.S. Geological Survey| pages = 24| last = Verdin| first = Kristine L.| title = Hydrologic Derivatives for Modeling and Analysis—A new global high-resolution database| location = Reston, VA| series = Data Series| accessdate = 2022-06-15| date = 2017| url = http://pubs.er.usgs.gov/publication/ds1053}}{{Citation| publisher = U.S. Geological Survey| last = Verdin| first = Kristine L.| title = Hydrologic Derivatives for Modeling and Applications (HDMA) Data| accessdate = 2023-03-02| date = 2017| url = https://www.sciencebase.gov/catalog/item/5910def6e4b0e541a03ac98c}} (global, 2017)
• MERIT-Basins,{{Cite journal| doi = 10.1029/2019WR025287| volume = 55| issue = 8| pages = 6499–6516| last1 = Lin| first1 = Peirong| last2 = Pan| first2 = Ming| last3 = Beck| first3 = Hylke E.| last4 = Yang| first4 = Yuan| last5 = Yamazaki| first5 = Dai| last6 = Frasson| first6 = Renato| last7 = David| first7 = Cédric H.| last8 = Durand| first8 = Michael| last9 = Pavelsky| first9 = Tamlin M.| last10 = Allen| first10 = George H.| last11 = Gleason| first11 = Colin J.| last12 = Wood| first12 = Eric F.| title = Global Reconstruction of Naturalized River Flows at 2.94 Million Reaches| journal = Water Resources Research| date = 2019| pmid = 31762499| pmc = 6853258| bibcode = 2019WRR....55.6499L}}{{Cite journal| doi = 10.1029/2019WR024873| volume = 55| issue = 6| pages = 5053–5073| last1 = Yamazaki| first1 = Dai| last2 = Ikeshima| first2 = Daiki| last3 = Sosa| first3 = Jeison| last4 = Bates| first4 = Paul D.| last5 = Allen| first5 = George H.| last6 = Pavelsky| first6 = Tamlin M.| title = MERIT Hydro: A High-Resolution Global Hydrography Map Based on Latest Topography Dataset| journal = Water Resources Research| date = 201| bibcode = 2019WRR....55.5053Y| s2cid = 189960352| doi-access = free| hdl = 1983/74b93ef1-da84-49db-b178-1cc1df6cbc0f| hdl-access = free}} [https://www.reachhydro.org/home/params/merit-basins website] (global, 2021)
• Hydrography90m,{{Cite journal| doi = 10.5194/essd-14-4525-2022| issn = 1866-3508| volume = 14| issue = 10| pages = 4525–4550| last1 = Amatulli| first1 = Giuseppe| last2 = Garcia Marquez| first2 = Jaime| last3 = Sethi| first3 = Tushar| last4 = Kiesel| first4 = Jens| last5 = Grigoropoulou| first5 = Afroditi| last6 = Üblacker| first6 = Maria M.| last7 = Shen| first7 = Longzhu Q.| last8 = Domisch| first8 = Sami| title = Hydrography90m: a new high-resolution global hydrographic dataset| journal = Earth System Science Data| accessdate = 2023-03-08| date = 2022-10-17| url = https://essd.copernicus.org/articles/14/4525/2022/| doi-access = free| bibcode = 2022ESSD...14.4525A}} [https://fred.igb-berlin.de/data/package/762 website] (2022, global, shows smaller headwater streams)

{{reflist}}

Category:Drainage basins

Category:Hydrology