Mar 30 2010
Taking data from the LIDAR sensor and a digital photogrammetic camera, researchers from the Universidad Politécnica de Madrid's School of Computing have developed an application for generating digital elevation models (DEM) and classifying land that can be applied to minimize the effects of flooding.
This tool is also useful for studying the best location for public works and residential areas and for controlling and managing water resources. The Spanish companies Inclam and Heli Ibérica Fotogrametría have partnered this project.
Floods
The numerical simulation of swell waves is a conventional way of providing flood risk maps. Computer applications like these are able to realistically simulate the variable situations of river basins, as they are able to cover all factors, including fluid mechanical properties.
Some of the tools that are now used to develop 2D hydraulic models have a number of defects. For example, many do not properly account for land relief. Others use satellite images whose resolution is unsatisfactory for land classification. Additionally, today's hydraulic applications do not use either specific coefficients, like the Manning roughness coefficient, or data captured by digital cameras during photogrammetic flights for land classification.
On these grounds, applications that integrate the data generated by hydraulic simulation models in geographical information systems need to be developed to output decision making support criteria for land planning and development.
In this context, the Spanish companies Inclam and Stereocarto, in partnership with Dr. Águeda Arquero and Dr. Estíbaliz Martínez, from the Department of Computer Systems Architecture and Technology, and Dr. Marina Álvarez, from the Department of Computer Languages and Systems and Software Engineering, at the UPM's School of Computing, have developed the LHIDRA-MANNING project: an automatic system for calculating the Manning roughness coefficient and digital elevation models from aerial images and LIDAR data for use in hydraulic modelling.
Technical optimization
As part of this project, the partners have optimized some of the traditional remote sensing image classification techniques for application to images taken using a large-format photogrammetic camera that generates a panchromatic (PAN), multi-spectral image with four spectral bands. The image has a spatial resolution of less than 20 cm and a radiometric resolution of 12 bits per channel. Because of the altitude at which the data are collected, the image has a better geometric resolution than images output by satellite sensors and requires no atmospheric correction.
The information supplied by the digital camera is supplemented by data captured by the LIDAR system to calculate the DEM, outputting a better altimetric accuracy (errors of less than 20 cm) and a better spatial resolution (decimetric precision).
This new system was presented at the latest Water Engineering Conference held in Madrid, where LHIDRA-MANNING, the automatic system for calculating the Manning roughness coefficient and digital elevation models from aerial images and LIDAR data for hydraulic modelling, was discussed.