The information content theory for the estimation of the topographic index distribution used in TOPMODEL

✍ Scribed by GIUSEPPE MENDICINO; AURELIA SOLE

• Publisher: John Wiley and Sons
• Year: 1997
• Tongue: English
• Weight: 513 KB
• Volume: 11
• Category: Article
• ISSN: 0885-6087
• DOI: 10.1002/(sici)1099-1085(199707)11:9<1099::aid-hyp547>3.0.co;2-f

✦ Synopsis

This paper analyses the signi®cance of the entropy concept in the topography parameterization within the model TOPMODEL proposed by , by means of the hydrological behaviour of an experimental basin in southern Italy. For a signi®cant number of ¯ood events recorded at the basin outlet, the performance of TOPMODEL for dierent spatial distributions of the topographic index, ln(a/tan b), has been observed. Performance is related to the information content estimated as an entropy measure, corresponding to each of the spatial distributions of the topographic index, with the aim of identifying the procedures most suitable to represent the hydrological process of rainfall±runo.

The results obtained have shown that for ¯ood events corresponding to brief, heavy precipitation, some procedures provide better performances than others. Moreover, these improvements are justi®ed by greater information content in the corresponding spatial distributions of the topographic index.

Finally, TOPMODEL performances for some procedures have been analysed, varying the resolution scale of the topographic index. For analogous hydrological performances, scale change produced variations in some of the subsurface hydraulic parameters. These variations were proportional to a spatial variability measure of the topographic index distribution, derived from the corresponding information content.