Premium
A combined algorithm for automated drainage network extraction
Author(s) -
Chorowicz Jean,
Ichoku Charles,
Riazanoff Serge,
Kim YounJong,
Cervelle Bernard
Publication year - 1992
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/91wr03098
Subject(s) - preprocessor , algorithm , digital elevation model , computer science , extraction (chemistry) , drainage network , drainage , channel (broadcasting) , data mining , artificial intelligence , remote sensing , geology , ecology , chemistry , chromatography , biology , computer network
Based on existing principles of automated drainage network extraction we have developed two methodological algorithms, the “profile scan” and “hydrological flow modeling,” and used them to extract networks from digital elevation models (DEMs). The “hydrological flow modeling” algorithm specializes in the extraction of well‐connected hierarchically arranged skeletal channel networks. On the other hand, the channels extracted by the “profile scan” algorithm lack adequate connectivity, but this algorithm is suitable for the extraction of wide valley bottoms and other flat areas. A combination of the two algorithms gives a more versatile algorithm capable of yielding networks which are not only well connected but also portray the surface character of the drainage network thus generated. The good functioning of our algorithm is not inhibited by the presence of pits in the DEM. There is therefore no preprocessing of the DEM prior to drainage extraction. Rather, at the end of extraction, isolated spots are eliminated since these are the ones most probably representing artifacts.