The Resource Quality Information Services river coverage is a South African surface drainage network for GIS, originally based on a 1994 1:500 000 rivers coverage from the Chief Directorate of National Geo-spatial Information. This version is made available "as is" for the convenience of those who need a medium-scale river coverage for display or modelling: the developers at Resource Quality Information Services would welcome any comments relating to errors or areas for improvement.

Note that substantial improvements to this dataset were made during the following Water Research Commission project: Weepener, H. L., van den Berg, H. M., Metz, M., Hamandawana, H., Jan. 2012. The development of a hydrologically improved digital elevation model and derived products for South Africa based on the SRTM DEM. Tech. Rep. 1908/1/11, Water Research Commission, Pretoria, South Africa. Many additional coastal rivers became available from the 2011 National Freshwater Ecosystem Priority Areas (NFEPA) project. See also Smith-Adao, L. et al. (2018) Origin, development and improvement of the rivers dataset. Chapter 2 In: Van Deventer, H. et al. 2018. South African Inventory of Inland Aquatic Ecosystems(SAIIAE). Version 2, released on 2018-11-06. South African National Biodiversity Institute, Pretoria. Report Number: CSIR/NRE/ECOS/IR/2018/0001/A.

The modifications made by Resource Quality Information Services to the original 1:500 000 river file are:

1. Alignment of the arcs to within 50m of the 1:50 000 topographical map coverages
2. Addition of arcs where required to link with the national monitoring network
3. Completion of arcs split by dams, lakes and bridges
4. Direction of rivers downstream
5. Ordering of streams (Strahler method)
6. Addition of ESRI's Digital Chart of the World arcs in order to complete the cross-border drainages
7. Addition or modification of the following fields:
   • NAME: generally, the name assigned to the reach on a topographical map
   • sometimes the correctness is uncertain.
   • ALIAS: alternate name or names for the reach, separated by \.
   • RIVN_ROOT: For multilingual annotation purposes (prototype): the root of the river name.
   • RIVN_SUFX: The qualifier of the river name, e.g. river, spruit, loop.
   • CLASS: unreliable indicator of seasonality of river flow. The classes and allocation require review:
   • • Dry=dry river, Perennial=perennial river, Non=non-perennial river,
     • Unknown=not present in original coverage or not transferred from source map.
   • PRIMARY: primary drainage region where the reach occurs.
   • • a lower-case letter indicates a pseudo drainage region outside South Africa.
   • ORDER: Strahler order, starting from source=1.
   • END: S=Source, N=iNtermediate, M=Mouth, E=Endorheic.
   • REACHCODE: 10-character reach code made up as follows:
   • • 1 to 4: quaternary drainage region in which MOST of the river falls.
     • (Note that some USA programs like BASINS expect eight numerical digits. The qat4qat8.txt lookup table contains the correct DWA to USA conversions.)
     • 5 to 6: arbitrary sequence number of the reaches within a quaternary drainage region.
     • 7 to 10: four zeroes (0000) in the 1:500 000 coverage, intended for later use at more detailed scale.
   • SORTER: sorting field - always increases downstream in a river network
   • LENGTH_KM: length of reach in kilometres (not to be confused with the meaningless LENGTH field).
   • RivEX fields: (see http://www.rivex.co.uk/)
   • • SHREVE: Shreve order.
     • US_ACCUM: Accumulated upstream metres at the lower node of each arc.
     • CATCHID: Automatically generated ID for each end node (i.e. when END='E' or 'M').
     • DIST2MTH : Distance to the mouth in metres from the upper node of each arc.
     • LINK_TYPE : same idea as the "END" field, above. E=Exterior, I=Interior, O=Outlet.
     • SEGMENT : Channel identifier.
     • CATCHID : drainage region identifier.

History

• [2012 April release: reach A23A02 name = Moreleta; fix B41K04 topology; 100+ new arcs, mainly from NFEPA, some from WRC project K5/1908, in the D, E, G, J, K, M, P, R, T, U, W, X regions, mostly short coastal rivers - read metadata XML with a plain text editor for details.]
• [2010 June release: Fixed errors in D61F07, F30F05, S10B05; E region Ratel given alias Ratels; Fixed uncertain river directions and quaternary drainage edges in D55F, J1, 7 and D54A5, 6.]
• [2009 January release: 4 new headwater streams in A31A. Naming of streams in E: Oudste, Noordhoeks, Agterste, Voorste, Boskloof and Krom. In T: Nenga River.]
• [2006 December release: Minor changes to the river arcs in B. Major change in L: closure of gap in the Sout River (L11A) near the Skilpadkop River, Gamtoos catchment.]
• [2006 August release: Minor changes to the river arcs in B, V, W & X regions. Inclusion of many "trace" .txt files that were previously missing from the web site.]
• [2006 March release: No major changes to the river arcs and existing attributes; many minor corrections detected using the fast RivEX extension to ArcGIS 9. RivEX has also enabled the addition of new fields: SHREVE, US_ACCUM, CATCHID and DIST2MTH (see definitions in centre column of this page).]
• [2005 July release: Lungile Gaulana has carefully worked through the river coverage during the first part of 2005, mainly checking names (named rivers are up from 6206 to 7568 of about 10000 arcs: 97.8% of a random sample of 1% of all SA arcs had correct naming). We have also used the Shuttle Radar Topography Mission 90-metre cell digital elevation model to try to improve river alignment in Lesotho, where no 1:50 000 rivers are available for comparison. Spatial accuracy has improved, especially in remote parts. Only 1.1% of arcs do not match the 1:50 000 coverage and 4.3% are in areas with no 1:50 000 arcs. The greatest error is in the classification of rivers as perennial and non-perennial: only 83.9% are the same as the 1:50 000 classes.]
• [2004 November release: about 12 minor changes to arcs and incorporation of Mozambique data from the Atlas for Data Preparedness and Response in the Limpopo Basin, INGC, University of Eduardo Mondlane, Department of Geography and USAID/ FEWS NET MIND 2003.]