symbolization and generalization to map water pipe data flow and water quality at different scales
ISBN 978-85-88783-11-9
Authors
1Anne, R.; 2Pham, H.
1IFSTTAR Email: anne.ruas@ifsttar.fr
2IFSTTAR Email: ha.pham@ifsttar.fr
Abstract
Mapping drinking water flow is a real challenge not only to detect water leaks but also to control the quality of the water. In France, 900000km of water pipe delivers 99% of the population. A recent law imposes to map these pipelines on a known geographical system with a planar positional accuracy from 0.4 to 1.5m according to the age and type of the pipeline. Wireless sensors and models based on computational fluid mechaniscs (CFD) allows to study flow and to reconstitute in each point parameters such as velocity, pressure and a chimical concentration. This information can be used to detect leaks and to control the concentration of chlorine or other chemical products. However this information is not easy to map on GIS because the small width of water pipe and the very high quantity of points necessary for data flow computation. In this research work we propose solutions to map this information at different levels of detail with other information such as roads and buildings. We first propose to use area symbols instead of punctual symbols to improve zoom-in visual effect. We also propose to generalize initial water data for zoom-out processes. We use the axis of the water pipe as basic geometry and we segment it. We then compute a generalized value of pressure, velocity or chemical concentration for each segment with specific function adapted to the property we wish to enlighten. We propose a conceptual data schema that describes the required information to map this data at different level of detail. The solution has been fully implemented on experimental data and illus-trated by means of a dedicated web mapping that proposes a set of GIS functions such as the selection of the data, zoom-in zoom out but also the animation to see the propagation of a chemical product on the water pipe