The seismic vulnerability analysis of urban environments is an operational issue that concerns the comprehensive knowledge of both building structural features and soils geophysical parameters, especially when considering areas that are prone to hydrogeological and seismic disasters. The protection of such environments, together with the population growth and the urbanization processes, requires a multi-disciplinary approach aiming at providing both an effective assessment of urban resources and synthetic parameters for managing post crisis events, restoration activities and search & rescue operations. Within such a framework, airborne Light Detection and Ranging (LiDAR) and Hyperspectral sensors have demonstrated to be powerful remote sensing instruments, whose jointly use allow providing meaningful parameters to describe both the topographic settings of urbanized areas and the buildings properties, in terms of geometrical, spectral and structural features. Based on this rationale, in this study, the operational benefits obtained by combining airborne LiDAR and Hyperspectral measurements are provided to support the seismic vulnerability assessment of urban seismic areas. The digital elevation model as well as the building height and the shape of the observed area are gathered by using airborne LiDAR measurements. Spectral and structural information of urban buildings are provided through the supervised classification of IMSpectorV10E VNIR (wavelength range between 400 and 1000nm subdivided into 503 bands) measurements acquired by the IPERGEO sensor. The objective is to combine the different products provided by LiDAR and Hyperspectral image processing analysis within a Geographic Information System (GIS) platform, to evaluate the intrinsic properties of buildings (e.g. perimeter, covered area, height and type of roofs) together with the topographic features of the surrounding area (e.g. the surface height and slope) for providing synthetic parameters and thematic maps useful for seismic assessment and mitigation purposes, such as: (i) the identification of steep slope areas, (ii) the analysis of building roof typology for supporting the evaluation of structural load conditions, (iii) the detection of critical structures (e.g. asbestos buildings), (iv) the identification of primary roads (in terms of escape or access routes) for supporting search and rescue operations, (v) the analysis of main road conditions after building collapses. Meaningful experimental results, gathered for the historical center of Cosenza city (Italy), allow demonstrating the benefits of the proposed approach for both seismic assessment and mitigation purposes.
The present work is supported and funded by Ministero dell'Università, dell'Istruzione e della Ricerca (MIUR) under the project PON01-02710 "MASSIMO" - "Monitoraggio in Area Sismica di SIstemi MOnumentali".
The roof type classification product looks very nice. There is, however, unfortunately no use case presented. What do you intend to use it for? For structural earthquake vulnerability assessment the roof type alone is surely not enough (if even significantly relevant at all). For other hazards such as windstorm, however, roof type is indeed a major input variable for vulnerability assessment, so there is clearly a lot of value in this..
I'd be interested also if you did some accuracy assessment on the ground with regard to how well the automatic roof type classification based on the hyperspectral data works.
First of all, I would specify that the products shown in this paper are included in the PON MASSIMO project funded by Italian Ministry of Education, University and Research, a multidisciplinary research with the participation also of experts in structural, geological and seismological field.
I try to reply to the your questions.
The road risk block index has been quantified, in this preliminary analysis, as function of the built volumes, in the hypothesis of a total collapse. In particular, an analysis based on the geometric characteristics of each building have been performed to estimate the fraction of material due to bearing structures, partition walls (also, considering the openings), plaster and floor slabs. However, I agree with your observation; in fact, a next improvement of this work is the introduction of further analysis parameters, such as the construction time of buildings for the extraction of materials used and the code technique adopted.
The roof type classification, coupled with other information, can be used for many purposes. For example, in framework of the South Italy , the identification of shingles and brick roofs can be used for a first screening phase in order to select the more ancient buildings. Also, the type of roof can provide information on the weight and slope of the coverage associated to typical methods of construction, which will be used together with information on the local seismic response, the geometry of the structures and construction materials to assign an attention index for each building on a large urban area (this is one of the aims of the whole research project). Moreover, the identification of potentially dangerous materials (e.g. the asbestos), in case of damaging due to the seismic shaking, can be taken into account to evaluate the environmental risk, therefore also the economic exposure.
With regard to the accuracy assessment of the roof classification, some samples are selected as ground truths for each class and a comparison in terms of overall accuracy among different classification algorithms can be found as poster, presented by same authors, to the 9th EARSeL SIG Imaging Spectroscopy workshop (Luxembourg, 14-16 April 2015). Also, we are working specifically on this topic, to assess the capability of the use in urban environmental of airborne and satellite hyperspectral sensors.
