Channel network extracted from a LiDAR DTM in a mountainous forested basin, Tijuca’s Massif, Rio de Janeiro, Brazil.
ISBN 978-85-88783-11-9
Authors
1Barbosa, L.S.; 2Avelar, A.S.
1UFRJ Email: leobarbosa@live.com
2UFRJ Email: andreavelar@acd.ufrj.br
Abstract
The channel network is extremely important to understand the various geomorphological processes. So the survey with increasing detail becomes more important for the further development of research in hydro-geomorphology. The Massif da Tijuca, in the city of Rio de Janeiro, southeastern of Brazil is an area of significant environmental value because it contains the Tijuca National Park and due to historical processes of population density on its slopes and surrounding areas , in which case the slopes presented an intense process of occupation. These environmental and social factors increase the need for more detailed base maps for better support the planning of these areas. Light Detection And Ranging (LiDAR) is an active remote sensing of light beams (laser), which to reflect on the surface enables get distance and other information due to the reflectance of the object, generating topographic data with high resolution for large areas (James et al. 2007). This type of survey can help identify channels, erosional features (such as ravines and gullies), and improve the topographic, hydrological and ecological modeling. After obtaining an image for the LiDAR of Tijuca’s Massif was used for the initial data treatment the Global Mapper 15 software, by generating a cloud of points in LAS format (Lidar Data Exchange File Format) which can be converted into a TIN. However, were interpolated only the points classified as ground, because due to the high resolution survey is necessary to differentiate the points raised concerning vegetation, buildings and other features claims on the ground surface points. The TIN has been generated from the point cloud and then the MDT to a grid (1m x 1m). The channel network was drawn automatically (ArcHydro10.1) adopting a criterion established for the same study site by Araújo (2013) in which the network channels is defined with minimal contribution area, 0.5% and 0.25% of the total basin area. Due to the high resolution of the DTM was also used 0.05%. Finally, the slope position classification for defining the valley bottom areas was performed using Topographic Position Index tool (TPI) proposed by Weiss (2001) with 8 different combinations of rays. The result of the TPI, provided 8 different maps, but when compared with the Dias (2010) study, it shows the same limitations in some areas, particularly on the spatial continuity of the tops and valley bottoms. After the various combinations of radius showed the best result was the sum of the TPI value with the radius of 50, 70, 80 and 100 meters. But the important thing to be extracted from this analysis are the appearance of convergence areas (valley and lower slope). The automatic channel network elaborated showed the best results with 0.05% of the area of the basin and as Araujo (2013) study showed that the automatic extraction of channels in the same field study showed good results when compared with the network reconstituted channels 1: 10,000. However, with the more detailed DTM used in this work to drain had a strong influence of roads, landfills, outcrops, and does not fit well in flat areas. Because of these errors the drainage network has demonstrated the need for field adjustments with DGPS and photo-interpretation. Although research be at an early stage, the LiDAR survey already shows great potential for the extraction of detailed morphometric parameters with a resolution of 1m x 1m (Barbosa et al. 2014). The results showed that a LiDAR DTM can be used to improve quality of geomorphological mapping in Tijuca’s Massif, but some methodological limitations still need to be adjusted.
Keywords
Channels; LiDAR DTM; ArcHydro