Research inventory
| Personal data | Research themes | Ongoing teaching | Publications |

Person in charge of the Unit : Oui
ANAGEO (Research Group in Geospatial Analysis) is led by Prof. Eléonore Wolff and part of the Institute for Environmental Management and Land-use Planning (IGEAT), Department of Geosciences, Environment and Society, Université libre de Bruxelles (ULB). ANAGEO carries out research in the field of man-environment interaction, with a special focus on the study of land use/land cover, its evolution and its determinants, using remote sensing and geospatial analysis. The group has a strong background in deriving geoinformation from satellite and airborne remote sensing imagery - notably through extensive expertise in GIS-based modelling and Object-Based Image Analysis (OBIA) for large-scaled applications entailing big data management. More recently, the group started working and developing deep Convolutional Neural Networks (CNN) for different remote sensing applications. It has been involved in several pluri-disciplinary international projects as coordinator or WP leader. The research topics that are addressed by ANAGEO revolve mainly around the following themes: monitoring of urban areas in space and time, land planning, natural resources management, population estimation, spatial epidemiology, and humanitarian action (including Mine Action). ANAGEO has a long experience in VHRRS and HRRS remote sensing and spatial analysis of urban areas in Europe and in developing countries, more particularly in Sub-Saharan Africa.
GeoRisCA (2012-2016) is a scientific project funded by the Belgian Federal Scientific Policy (Belspo).The main objective is the assessment of the georisk in the Kivu region (DRC, Rwanda, Burundi), by analysing and combining seismic, volcanic and mass-movement hazards as well as the vulnerability of the population, the infrastructures and the natural ecosystems, in order to support risk management.The project is coordinated by the Royal Museum for Central Africa, in collaboration with the Université Libre de Bruxelles, the Université de Liège, the Vrije Universiteit Brussel and the European Center of Geodynamics and Seismology. It is conducted in close collaboration with local partners from the three targeted countries (Burundi, DRC and Rwanda).
VALIURB: Urban sprawl and ecological corridors: Consolidation of VHRS remote sensing methods
The objective of this study is to analyse the evolution of built-up surfaces and vegetation in urban and suburban areas of Brussels, Strasbourg and Rennes from high and very high spatial resolution images in order to characterize urban changes on the one hand and precisely identify and characterize ecological corridors on the other hand. The methods developed by the laboratories of Strasbourg (LIVE), Rennes 2(LETG-COSTEL) and ANAGEO will be exploited in order to define a multiscale methodological approach as consolidated and reproducible as possible. This project is co-funded by the Belgian Science Policy Office and the CNES and will use images of the new Pleiades satellites.
ARMURS - Automatic Recognition for Map Upadte by Remote Sensing
Topographical data producers are currently confronted the need of a faster updating method. Although a state-of-the-art technique exists, no automated tool predicts and locates changes. The general objective of the project is to capitalize partners' knowledge in the development of a demonstrator to assist data producers in updating more efficiently their topographic database by using state-of-the-art image processing and statistical analysis techniques. Data will include remote sensing images, together with socio-economical data. To achieve this general objective, two main axes are distinguished: - At a regional scale, the objective is to analyze topographical databases from multiple sources such as satellite images, demographic database, or economic database in order to predict information about the localization of changes in man-made structures (such as houses, roads, etc.). - At a local scale, for areas of predicted changes the older databases will be compared with recent very high resolution images in order to detect and recognize differences (errors in the topographical database, building construction, road construction, etc'). Several types of man-made structures are distinguished (house, road, etc.)
TEMPORAL STUDY OF SOIL IMPERVIOUSNESS EVOLUTION IN THE BRUSSELS-CAPITAL REGION
Imperviousness is directly related to urban development. It can be defined as the sum of roads, car parks, sidewalks, rooftops and other impermeable surfaces of the urban landscape. Impervious surfaces no longer allow rainfall to soak into the ground. Consequently, most rainfall is directly converted into stormwater runoff. Aside from causing water pollution and other environmental problems, this can be too much for the existing drainage system to handle and lead to floods. Up to now, the usual response to floods in Brussels lay in the improvement of the drainage system and the digging of stormwater reservoirs, but increasing awareness has led the Ministry of the Brussels-capital Region (MBCR) to consider changing its policy as regards imperviousness. The first step in this process consists in evaluating the evolution of the situation, as a basis for further analysis. The objective of this study is to measure imperviousness evolution from the 1950s up to now in the Brussels-capital region, with extensions following the limits of the watersheds (i.e. about 200 sqkm in total). In this framework, the percentage of imperviousness is estimated in 1-ha grid squares.The use of high- and very high-resolution satellite remotely sensed data is the fastest and most efficient way to estimate the extent of imperviousness. However, because high-resolution data have been available since the 1980s only, and very high-resolution data since 1999 only, the study period is divided into 1955-1985 and 1987-2005. Different data types are used within these two periods, according to their availability and relevance.The results of the study will be : first, imperviousness maps at each date considered; second, maps displaying the evolution index of the impervious surfaces percentage; and third, statistics and/or maps aggregating the percentages and/or their evolution on the level of the different spatial entities of interest.
GRAZEO is a 2-years project financed by the Belgian Science Agency. It is a spin-off of the EPISTIS project (Remote Sensing Tools to Study the Epidemiology and Space/Time Dynamics of Diseases ) where several models were developed using remote sensing and spatial modelling to stratify the risk for foot-and-mouth disease (FMD) transmission between buffaloes and cattle in Kruger National Park (KNP) and its surroundings. In GRAZEO however, the emphasis will be put on the role of the natural vegetation as a key factor of habitat suitability for large herbivores. The abundance and quality of the forage source is known to be a key driver of feeding patterns and distribution of livestock and wild grazers in savannah rangelands. Therefore, the main aim of GRAZEO is to explore the potential of the new generation of very-high spatial and spectral resolution sensors, such as WorldView-2, for (i) developing methodologies for mapping grass patches, grass biomass and quality indicator (nitrogen concentration), as well as tree species communities as complementary forage quality indicator, at a scale compatible with savannah spatial heterogeneity and animal movements; (ii) Investigating how these improved inputs, along with adapted modelling processes and multiple scenarios testing can contribute to improving the buffalo-cattle contact modelling outputs.
MORECA: Monitoring of large scale small holder reforestation projects for carbon finance mechanisms
In the Virunga National Park the main cause of deforestation is illegal logging for charcoal production and for fuel wood collection. To tackle this problem, the current EcoMakala project (EU, 2007-2012) aims to reforest at least 2000 ha of smallholder lands (0,5 to 5ha). All around the world similar projects exist and help to fight against deforestation. To be eligible as an activity under the 'Clean Development Mechanism' of the Kyoto Protocol, and therefore be eligible for funding, a conclusive proof that the areas were not carrying forest (as defined by UNFCCC) from 1990 onwards has to be delivered. Satellite images at high resolutions (such as SPOT and LANDSAT) existed since the 80's and can specify the land cover in the past.The MORECA project in therefore born to (i)Test the feasibility of applying the UNFCCC definition of forest on archival images such SPOT-1 at 20m of resolution.(ii)Define the eligibility of the plots (based on the UNFCCC's forest definition for the DRC and the basis year 1990); (iii)Monitor the plantations with very high resolution images.
Our aim is to increase the general public concern on Climate Change and invite people to commit themselves to reduce their greenhouse gases emissions. The general tone of the campaign will be light, fascinating and motivating. It has to create affective positive links with the implementation of the solutions, especially for youth. The CC concerns and messages will therefore be given to the public by existing individual animals tracked by satellites and individually identified (each has a first name). Their messages will be broadcasted on various TV channels across Europe. The project includes : 1. The production and distribution of 35 TV clips (1 or 2 minutes). These TV clips will use satellite-tracked animals as Trojan Horses to explain Climate Change concerns. 2. A website will include (1) a flash programme where people may commit themselves to reduce their CO2 emissions, their commitments are registered in a database and the statistics of the engagements are shown, (2) news from the satellite-tracked animals in the form of these animals blogs, (3) more info on CC, (4) more info on behavioural changes to reduce GHG emissions, (5) the TV clips, (6) partners, contacts, links, various. 3. Viral communication/viral marketing will increase the visibility, the recruitment and the spread of the message by the use of viral videos (or virals). We call 'virals' the small programmes (flash, pps, mpeg, ...) that are spread throughout the internet as attachment to emails sent by people to each other on a voluntary basis. Most often because they are attractive, funny and convincing. 4. Additionally, further communication and spread actions will allow to distribute these products across Europe and outside Europe, during and after the programme duration. These include Press conferences and Press releases at European and national levels. Contacts and follow-up after the end of the programme. Basing our estimates on other campaigns, we target (cautious estimates) : - 10 TV channels in Europe broadcasting the TV clips - Cumulated audiences of over 50.000.000 TV viewers - Over 100.000 European citizens committed to reduce their GHG emissions - Over 150.000 European citizens receiving the virals (this is hard to check for reporting however)
Up-to-date knowledge of land cover is an important tool for the various planning authorities with responsibilities for the management of territory. The geospatial objects are changing over time and the land cover information (vector geospatial database) has to be up-date periodically.The vector geospatial databases at a local scale in Belgium (e.g. Agriculture Directorate-General crop field database, National Geographic Institute 1/10000 topographic maps, Projet Informatique de Cartographie Continue of the Walloon Ministry of Transport and Equipment) were carried out by digital cartographic production process based on aerial photographs, and the update of these databases is now at the agenda.The VHR satellite data are, since 1999, capable to meet the mapping and monitoring needs of municipal and regional planning agencies. Indeed, these data from the sensors Ikonos, QuickBird, OrbView-3, and in near future, the Pléiades-HR French sensors, have spatial resolution lower than 5 m in multispectral mode and lower than 1 m in panchromatic mode.In this context, this post-doctoral research project suggests to use region based classification of VHR satellite data in different change detection processes for updates of vector database.
The aim is to understand how the peasantries are integreted into the dynamics of territorially-based power construction which are linked to the agricultural colonization of the Cambodia's Northwest uplands.The research focuses on both the behaviors of the peasantries against the power systems that control the access and use of land resources in the area, and the contrasted economic rationalities of the peasantries involved in a process of rapid class differentiation.
CASIM: Mapping impervious surfaces in Wallonia
In order to meet the growing need for environmental protection, management of natural hazards, natural resources conservation and sustainable agriculture imposed by the European Union, the Walloon Region wishes to enrich its database of geographic reference data with the large-scale mapping of impervious surfaces. Additionally, as these impenetrable materials such as asphalt, concrete, brick, and rooftops are considered to be a significant alteration of the quality of the soil as a natural resource (COM 2006, European directive on soil protection), knowing the exact extent of these impervious surfaces is very important. The CASIM project aims at mapping these impervious surfaces herein considering different mapping methodologies. Furthermore, at the same time the value of the existing aerial ortho-images, LIDAR data and different thematic data produced or acquired by the Walloon Government is assessed for mapping these surfaces. In detail, this study involves two tasks:(i) Mapping of impervious surfaces with existing data, managed, produced or acquired by the Public Service of Wallonia, including the land use map of Wallonia, the digital soil map, the PLI, ...(ii) Assessing the feasibility of mapping impervious surfaces in Wallonia with a combination of aerial ortho-photos and LIDAR data. For this a pilot site was selected in the basin of the river Houille.Different methods of segmentation and image classification will be identified and tested.
POPSATER - Population Estimation by Remote Sensing
It is essential to many projects to have some knowledge of population data (socio-economic manpower, distribution'). However most developing countries do not have or have little information concerning their population. High- and very high-resolution satellite images represent a new opportunity to improve population knowledge. The objective of this project is to develop, in rural and urban areas, a new method of population estimate based on satellite images, to highlight its cost and to compare it with the costs of censuses. Two types of approaches are considered according to the different case-studies. When the dwellings are identifiable on the image thanks to their shape, their texture and their spatial arrangement a direct approach is considered. It consists in multiplying the number of dwellings by the average population per dwelling. If the dwellings are too small to be identified and counted, the population estimate will be carried out using a zonal approach: the surface of the built-up area is measured and multiplied by the density of inhabitants. In order to obtain precise data, a questionnaire was drawn up and submitted to resource-people in order to quantify various parameters relating to the households and their lifestyle. The objective of the project is to set up a replicable method that can be used easily and quickly.
EPISTIS - Remote Sensing Tools to Study the Epidemiology and Space/Time Dynamics of Diseases
This four-year project financed by the Belgian Science Policy is aimed at developing remote sensing tools for improving the analysis of the space/time dynamics of disease transmission. Two case-studies have been selected, the first being focussed on a vector-borne disease and the second on a highly contagious disease, i.e. (1) the presence and spread of Bluetongue (ovine catarrhal fever) in Italy and in the Mediterranean basin and (2) the dynamics of transmission of foot-and-mouth disease at the fringe of the Great Limpopo Transfrontier Park in southern Africa. Although these case-studies target specific animal diseases, the innovative approaches that will be developed and implemented will be generic and therefore applicable to other diseases, including human diseases, having a similar epidemiological setting. A preliminary epidemiological analysis will ensure the collection of the necessary data on disease occurrence, vectors, potential hosts, wildlife, cattle and human population. Data derived from remote sensing - from low to very-high spatial resolution - will serve on the one hand as an information source at the epidemiological level and on the other hand as input variables in spatially-explicit models on different scales: distribution models on the country or regional scale, landscape models and spread models on the local scale and models based on very-high resolution in focus areas requiring a higher level of detail. All these models will be integrated in a space/time information system (STIS) so as to allow a more in-depth and more integrated analysis than what has been done up to now in this field. The results will help assessing the risks of future spread, show the effectiveness or ineffectiveness of existing barriers and improve the management and control of diseases.