Global fire monitoring systems are crucial to study fire behaviour, fire regimes and their impact at the global scale. Although global fire products based on the use of Earth Observation satellites exist, most remote sensing products only partially cover the requirements for these analyses. These data do not provide information like fire size, fire spread speed, how fires may evolve and joint into single event, or the number of fire events for a given area. This high level of abstraction is very valuable; it makes it possible to characterize fires by types (either size, spread, behaviour, etc.). Here, we present and test a data mining work flow to create a global database of single fires that allows for the characterization of fire types and fire regimes worldwide. This work describes the data produced by a data mining process using MODIS burnt area product Collection 6 (MCD64A1). The entire product has been computed until the present and is available under the umbrella of the Global Wildfire Information System (GWIS).

The World Health Organization estimates that 100 thousand people in the world die every year from asbestos-related cancers and more than 300 thousand European citizens are expected to die from asbestos-related mesothelioma by 2030. Both the European and the Italian legislations have banned the manufacture, importation, processing and distribution in commerce of asbestos-containing products and have recommended action plans for the safe removal of asbestos from public and private buildings. This paper describes the quantitative mapping of asbestos-cement covers over a large mountainous region of Italian Western Alps using the Multispectral Infrared and Visible Imaging Spectrometer sensor. A very large data set made up of 61 airborne transect strips covering 3263 km2 were processed to support the identification of buildings with asbestos-cement roofing, promoted by the Valle d'Aosta Autonomous Region with the support of the Regional Environmental Protection Agency. Results showed an overall mapping accuracy of 80%, in terms of asbestos-cement surface detected. The influence of topography on the classification's accuracy suggested that even in high relief landscapes, the spatial resolution of data is the major source of errors and the smaller asbestos-cement covers were not detected or misclassified.

Land-cover/land-use thematic maps are a major need in urban and country planning. This paper demonstrates the capabilities of Object Based Image Analysis in multi-scale thematic classification of a complex sub-urban landscape with simultaneous presence of agricultural, residential and industrial areas using pan-sharpened very high resolution satellite imagery. The classification process was carried out step by step through the creation of different hierarchical segmentation levels and exploiting spectral, geometric and relational features. The framework returned a detailed land-cover/land-use map with a Cohen's kappa coefficient of 0.84 and an overall accuracy of 85%.

Accidental release of crude oil into the sea due to human activity causes water pollution and heavy damages to natural ecosystems killing birds, fish, mammals and other organisms. A number of monitoring systems are used for tracking the spills and their effects on the marine environment, as well as for collecting data for feeding models. Among them, Earth observation technologies play a crucial role and moderate spatial resolution satellite systems are able to collect images with a very short revisit time or even daily. This paper describes the use of MODIS (Moderate Resolution Imaging Spectroradiometer) data for monitoring large oil slicks with the Fluorescence/Emissivity Index (FEI) and Object Based Image Analysis (OBIA). Two case studies are presented: the Deepwater Horizon (2010) and the Campos Basin (2011) oil spill accidents. Results show that is possible to track the dynamics of the slick both for massive and long-lasting accidents and for smaller and very quick accidents. The main advantages of the method proposed are a straightforward implementation, a fast and semi-automated data processing and the capability of integration of daytime and nighttime acquisitions, as well as its adaptability to different sensors

No abstract