Geomorphometric attributes of the global system of rivers at 30-minute spatial resolution
Introduction
The geomorphology of drainage basins and the organization of stream networks has been well-established for several decades (see Jarvis and Woldenburg, 1984). Quantitative tools emerged initially from field analysis of single, small catchments (e.g. Horton, 1945, Schumm, 1956) or of synthetic basins derived from statistical models (e.g. Shreve, 1966, Werner and Smart, 1973). Early regional-scale studies also exist, such as the summary of river network characteristics for the conterminous United States by Leopold et al. (1964). Recent research has focussed on scale-dependent extraction of drainage basin attributes (e.g. LaBarbera, 1989, Lammers and Band, 1990, Helminger et al., 1993, Band and Moore, 1995) as well as assessments of the influence such attributes have on hydrological response (e.g. Beven et al., 1988, Band et al., 1991, Band et al., 1995, Moore and Grayson, 1991, Famiglietti and Wood, 1994, Rodriguez-Iturbe, 1993, Sivapalan, 1993). However, these studies never progressed to the global scale and the generality of the statistics presented still requires testing as a precursor for use in global change studies.
We recently presented (Vörösmarty et al., 2000a) a gridded river networking scheme, global in domain and organized at 30′ spatial resolution and offered details on the construction and verification of this data base, its geographic co-registration to discharge and river chemistry monitoring stations, and an analysis of land-to-ocean linkages. We have applied versions of the STN-30p data set in water budget and river discharge studies at the regional (Vörösmarty et al., 1996a, Vörösmarty et al., 1991), continental (Lammers et al., 2000), and global scales (Fekete et al., 1999). It has also been used to study the impact of large reservoirs on continental runoff distortion and suspended sediment flux (Vörösmarty et al., 1997b, Vörösmarty et al., 1997c). The 1° to 30′ scale is developing as the focal point for continental and global-scale water and constituent transport modeling (e.g. Seitzinger and Kroeze, 1998, Oki and Sud, 1998, Ludwig et al., 1996, Ludwig and Probst, 1998, Vörösmarty et al., 1997a, Vörösmarty et al., 1997b, Vörösmarty et al., 1997c, Oki et al., 1995, Esser and Kolhmaier, 1991), which will require simulated river networks like STN-30p. The 30′ spatial resolution appears to be a sensible compromise between the necessary level of topological detail and computational requirements of finer-scale global data sets (e.g. Graham et al., 1999, USGS-EDC, 1998). Ongoing work is aimed at developing tools to create and analyze the nature of aggregated river networks using finer-scale data sets.
Section snippets
Methods
The steps and algorithms used in constructing the digital river network data set are summarized in Fig. 1. We developed the Simulated Topological Network for potential flow pathways (STN-30p) by spatially aggregating to 30′ (longitude×latitude) the ETOPO5 five to ten-minute digital elevation model (DEM) (Edwards, 1989), which was the best global data set available to us at the time we initiated this study. Because the data set is in geographic coordinates, individual cell areas change with
Results and discussion
In the following sections we present a set of geomorphometric attributes describing the STN-30p river networks and their associated drainage basins. We provide summaries for each of six continents and the globe, and use several individual river systems to highlight our major findings. The STN-30p river networks are shown in Fig. 3and the drainage basins in Fig. 4.
Conclusions
We have analyzed the spatial organization of the global land mass using a simulated topological network (STN-30p) representing potential flow pathways across the entire non-glacierized surface of the Earth at 30′ (longitude×latitude) spatial resolution. We derived from STN-30p a set of geomorphometric statistics on river segments defining sub-basins, complete drainage basins, individual continents, ocean basins, and the globe.
From both our study of individual stream segments reported here and
Acknowledgements
We wish to thank colleagues who assisted in the verification of STN-30p digital products (S. Kempe, University of Darmstadt, GERMANY; N. Fleming, CSIRO Division of Water Resources, Canberra AUSTRALIA; R. Wasson, Australian National University, Canberra AUSTRALIA). We also thank two anonymous reviewers and W. Ludwig for helpful reviews. We recognize important assistance on data base development and production of graphics, which were provided by S. Glidden and B. Tucker. Financial support came
References (65)
Grid mapping of river discharge
J. Hydrol.
(1995)- et al.
Forest ecosystem processes at the watershed scale: basis for distributed simulation
Ecol. Model.
(1991) - et al.
On hydrological heterogeneity — catchment morphology and catchment response
J. Hydrol.
(1988) - et al.
Automating object descriptions of drainage basins
Comput. Geosci.
(1990) - et al.
A model of river runoff for use in coupled atmosphere–ocean models
J. Hydrol.
(1994) - et al.
Potential evaporation functions compared on U.S. watersheds: implications for global-scale water balance and terrestrial ecosystem modeling
J. Hydrol.
(1998) The sediment delivery problem
J. Hydrol.
(1983)Hydrology and freshwater ecology
- et al.
Scale: landscape attributes and geographical information systems
Hydrol. Process.
(1995) - et al.
The effect of different terrain representations and resolution on simulated watershed processes
Z. Geoporph. N.F. Suppl.-Bd. (Berlin/Stuttgart)
(1995)