Figure Legend.

The satellite image map was produced during work with the Department of Mines and Water Resources Department in Asmara over a period of about 18 months, from images purchased by the Eritrean Interagency Consortium (EIAC) and by the Open University. Special thanks go to Martine Bilanou (Letense) formerly of EIAC, Haile Woldetensae of WRD Asmara, Dick Carlton of the Open University UK, and Beto de Souza of the University of Campinas in Brazil, for support, advice and assistance.
It is made from 12 Landsat Thematic Mapper images, and uses data from two infrared channels (Bands 7 and 4) that detect clay minerals and vegetation, and visible blue light. They make up the red, green and blue parts of the colour images, so explaining why vegetation looks green and the bare surface is much more colourful than in ordinary visible light. The full image is very useful for making inventories of vegetation, soild, rock types, underground water, and topography. A preliminary 1:1million scale print is available from Ghebremichael Temnewo ofMinistry of Land, Water and Environment, Water Resources Department, PO Box 1488, Asmara, for US$20. A reprint using the version here is planned. Proceeds from sale are intended for a variety of charitable causes in Eritrea. You should make enquiries about availability to Ghebremichael Temnewo on fax 291-1 124625.

Steve Drury, Oct. 1997.

Figure Legend.

This map was compiled by computer analysis of the ETOPO5 30 arc second (1 kilometre grid) data, which comprise topographic elevation values derived from published topographic maps of the world. I extracted the area of NE Africa containing Eritrea and used the image map to mask out bordering areas to emphasize Eritrea. The next step was to simulate the topography as if it was illuminated by lighting from the north-west (a standard procedure for cartographic presentation). The drainages are simulated by following the downhill path that water falling on every element of the topography would take. Neither the topography nor the drainages are accurate, because of the coarse resolution of the data (averaged for every square kilometre). Drainages are wrong for the area around the Danakil Depression, because the modelling assumes all water flows to sea level and the stops, and of course the Danakil Depression is below sea level. However, all the main geographic features are shown, including the basins of the Tekessie/Setit, Mereb/Gash, Barka and Anseba rivers.

Steve Drury, Oct. 1997.

---