Flooded areas assessment by integrating hydraulic flood analysis to the detailed flood maps generated with a multi-temporal image segmentation approach using Cosmo-Skymed

Abstract

ABSTRACT The support of Earth Observation Systems to flood monitoring and damage assessment has been evaluated by integrating skills and knowledge of hydrologists and experts on image processing. Detailed flood maps are generated from the analysis of images pairs acquired by Cosmo-Skymed on the same area at different times. The adopted methodology applies image pre-processing and segmentation techniques with a multi-temporal approach. After noise reduction by despeckling, the user manually localizes few water points in one image. From selected points, segmentation distinguishes the classes of flood, permanent-water and no-change areas. Then, an anisotropic scanning analyzes the images to define its content. The resulting connected regions are converted into vectors to be entered as constraint for a Physically based 2-D hydraulic flood model. The model recursively varies the unknown boundary conditions to match at best the areas extracted from Cosmo-Skymed. The product is a hydraulic consistent report of the flooded area including information on water depth and velocities. By combining these information with vulnerability maps, extracted from optical satellite images with a supervised approach, an es timation of the damage is provided. The reported results refer to the monitoring of the flood event occurred in the Scut ari Lake area (Albania, January 2010). Keywords: flood maps, image processing, multi-temporal, COSMO-SkyMed, segmentation, hydraulic model, water depth, water velocities. 1. INTRODUCTION Flood hazard is one of the central themes of environmental monitoring that can be supported with the information provided by the Earth Observation Systems. During flood events, the SAR (Synthetic Aperture Radar) images are particularly useful in flood monitoring and detection. Thanks to their all-weather capability and their cloud-penetrating properties, the integration of radar images and their analysis into a monitoring system greatly improves its efficiency. Then, the advantages obtained with the exploitation of EO data have been verified in this work by combining the knowledge of hydrologists, environmental scientists and experts on image processing. So an integrated method for the detection of the floods hazard and damage assessment has been developed. In this paper, some applications of image processing and segmentation of SAR images are presented, in order to generate detailed flood maps. A multitemporal image-analysis is applied to pairs of SAR images acquired on the same area at different times. A multi-temporal image segmentation generates maps of flooded areas. The proposed me thod is able to distinguish connected regions of flood, permanent-water and no-change areas. The methods have been tested on images coming from Cosmo/Skymed mission. The acquisitions of Cosmo Skymed in Stripmap mode covered the area of the Scutari Lake (Albania) after the flood event occurred on January 2010. In order to develop a hydraulic consistency of the flooded areas detected from Cosmo Skymed acquisitions a 2D hydraulic model have been developed for the specific purpose. The model is easy to apply on large regions with limited amount of information other than the sa tellite-retrieved flooded maps. An iterative scheme allows to explore boundary conditions of the model in order to provide the best estimate of the inundated areas with additional information on the