Endnotes

1 Address for correspondence: Dept. of Geosciences, Oregon State University, 104 Wilkinson Hall, Corvallis, OR 97331-5506, USA. Tel: +1-541-737-2722; Fax: +1-737-1200; e-mail: wolfa@geo.orst.edu.

2 The literature on individual watershed studies is too extensive to cite here.See Beach et al. (2000) for more information.

3 Work on this study began when Wolf was with the University of Alabama’s Department of Geography.

4 Much of the Database is available on the online Transboundary Freshwater Dispute Database. The annotated bibliography grew from the work of Beach et al. (2000).

5 Similarly, the 1997 UN Convention on Non-Navigational Uses of International Watercourses defines a “watercourse” as “a system of surface and underground waters constituting by virtue of their physical relationship a unitary whole and flowing into a common terminus. ”An “international watercourse” is a watercourse, parts of which are situated in different States [nations].

6 Perennial streams flow year-round, as opposed to intermittent streams, which have periods of no flow.

7 This definition, which we feel is the best available, does allow for one occasional inconsistency: If a basin is shared by only two nations, and all tributaries which cross the boundary are intermittent, we do not include it in the Register. If, however, a basin is shared by three or more nations, and tributaries which cross any of the boundaries are perennial, we include both the basin and all the countries within its territory, even if one or more of those countries only contributes intermittent streams. For example, Egypt is listed as riparian to the Jordan, even though no perennial streams cross its boundary with Israel. (These special cases are noted in the footnotes of their respective basins.)

8 Eight data layers form HYDRO1k. The six raster layers are the hydrologically correct DEM, flow directions, flow accumulations, slope, aspect, and compound topographic index (wetness index). The two vector layers include the drainage basins and synthetic streams. The traditional procedures for extracting drainage features are iterative and well established (Verdin and Greenlee, 1996). New methods for DEM surface drainage processing have been documented in Verdin and Greenlee (1996), and supported by Danielson (1998). HYDRO1k is available on the USGS's HYDRO1k Elevation Derivative Database.

9 To ensure accurate area representations, the first step was to project the DEM into an equal-area map projection. The second step identified real and artificial depressions (sinks) greater in area than a predetermined threshold, such as Lake Chad or the Dead Sea. Determinations were made as to which of the sinks were natural or spurious, by creating a sink mask and overlaying existing mapped hydrography. Once all depressions were verified, the DEM was filled using an approach developed by Verdin and Greenlee (1996), from which naturally identified sinks were maintained and spurious anomalies were removed.

10 The “Watersheds of the World” files included on the GlobalARC data set provided an excellent first approximation reference for this project as the HYDRO1k dataset was being developed, and we acknowledge the developers of this data set with gratitude.

11 We recognize the limitations of our process by reporting the size of basins, not as raw data as is common with digital data, but by rounding the last significant figure in basins 1-99 km² and the last two significant figures in basins 100 km² or larger.

12 The 1978 Register actually lists 215 international basins, but the Jurado is included in both North and South America. To avoid such ambiguity, we include the Jurado only in South America.

13 Total land surface of the earth = 148 940 000 km² (CIA World Factbook 1998); total land surface within international basins = 60 710 000 km²; percent of total land surface within international basins = 41.02%; total earth land surface excluding Antarctica = 134 940 000 km²; percent of total land surface within international basins, excluding Antarctica = 45.31%.

14 Our percentages of land surface within international basins, both within continents and globally, is vastly different than the 1978 Register for some continents, because that publication did not include island nations, either in its register or in its calculations of land surface. Our percentages for Asia, for example, are significantly lower since, apparently, the 1978 Register did not include the land surface of many Pacific islands in its calculations.

   info@icwc-aral.uz