Landform identification of the low-order filled valleys by using digital aerial photogrammetry and morphometric analysis: a case study of the Xindian River, northern Taiwan

Authors: Wen-Yu Lin*, National Taiwan Normal University, Su-Min Shen, Department of Geography, National Taiwan Normal University
Topics: Geomorphology
Keywords: low-order filled valleys, morphometric analysis, digital aerial photogrammetry
Session Type: Poster
Day: 4/12/2018
Start / End Time: 1:20 PM / 3:00 PM
Room: Napoleon Foyer/Common St. Corridor, Sheraton, 3rd Floor
Presentation File: No File Uploaded


Taiwan, a highly populated mountainous island in a tectonically-active and monsoon/typhoon-dominated settings, has suffered dramatic loss of lives and properties in many low-order stream valleys due to debris flows, various slope failures and flash floods over last several decades. The limited gentler land surfaces in these low-order valleys, on which the settlements and farmlands are usually located, are mainly formed by debris flow and/or mass-wasting processes. It is doubt, however, whether any governmental agencies have an overall pictures of where are the numerous lower-order valleys located and to what extent have these valleys been developed. Thus, this research attempts to develop a (semi-) automatic procedure to identify the filled valleys with stream order lower than the 5th, of which the valley beds are characterized by debris flow fans and filled deposits. The study area is in within the Xindian River with outcrops of the Tertiary sedimentary rocks in northern Taiwan. Morphometric analysis is firstly applied to automatically identify the “candidates” of the filled valleys and then digital aerial photogrammetry, UAV and ground observation is conducted for the three most significantly filled valleys. The results of various resolution of DEM (20 x 20, 6 x 6 and 2 x 2 m) are also compared and the validity of the proposed procedure is also discussed.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login