CORDIS - Forschungsergebnisse der EU
CORDIS

G-MOSAIC GMES services for Management of Operations, Situation Awareness and Intelligence for regional Crises

Final Report Summary - G-MOSAIC (G-MOSAIC GMES services for Management of Operations, Situation Awareness and Intelligence for regional Crises)

Executive summary:

The G-MOSAIC project has responded effectively to the needs expressed by EU Member States and international organisations. Users such as these have provided positive feedback on the services delivered during the course of the project. The consortium received official letters of appreciation from the United Nations Department for Field Support and the United Nations Operation in Cote d'Ivoire (UNOCI), expressing their gratitude for the support provided to their respective operations.

Project Context and Objectives:

During the three years of the G-MOSAIC project, a set of pilot GMES services for Security applications have progressively evolved from research and development activities into pilot and / or pre-operational services. Five Security domains have been addressed by the services developed:

a)Natural resources and conflicts;
b)Migration and border monitoring;
c)Non-proliferation and treaty monitoring;
d)Critical assets, and
e)Crisis management and assessment.

These security domains fall into two categories: 'Intelligence and Early Warning' (a-d), and 'Crisis Management and Operations' (e).

The European Commission (GMES unit), in consultation with key European Union stakeholders, has further defined the scope of the Security dimension of GMES and - through interaction with users - has investigated service perimeters and a possible 'modus operandi' for the three principal Security domains indicated in the EU Regulation on GMES and its initial operations (2011-2013), namely: border surveillance, maritime surveillance and support to the EU External Actions .

Governance and Data Security policy are pre-requisites for the operational phase of GMES services in the Security field. To this end, a GMES Security Board was established, in order to take into account the recommendations on GMES Data Security policy adopted by the Council Security Committee (CSC) on the 26th of January, 2010.

Industrial partners, who have been involved primarily as service providers in GMES projects for Security, have progressively developed pilot and / or pre-operational services some of which have proven to be both sufficiently advanced and technically mature. In G-MOSAIC, this is particularly the case for the crisis management and assessment products and services.

During the course of the project, crisis management and assessment services were provided on the basis of round-the-clock availability (24 hours a day, 7 days a week). Service production was organised through 16 cooperative 'service chains', in which service providers worked closely with the European Union Satellite Centre to build the services up to the required level. The involvement and contribution of users, as well as the contributions from research and institutional partners in the G-MOSAIC consortium have been highly valuable, with regard to both the detailed technical requirements and the assessment and validation of the final products.

During the third and final year of the project, services in support to crisis management, in particular reached operational maturity. The G-MOSAIC project raised also several important points that should be taken into account in the further evolution of services in support to EU External Actions:

-The continuity of the service provision. GMES services in the Security field have established a strong and active base of users, who regularly utilise the services in support of their activities. Any discontinuity in the provision of services is likely to impact adversely both on user and stakeholder confidence in the GMES services for Security applications.

-Data Security policy. In the Security field, some of the data and information handled by, and exchanged between GMES service providers and users could be deemed sensitive, for political or strategic reasons. The need for a consensually-defined and properly implemented Data Security policy is therefore a necessary condition to be met by the services, in order to avoid the potentially detrimental use of sensitive data or products.

-Cooperative workflow production. Thanks to a common view and coordinated approach, the specialized knowledge of each partner converts often to a common Product Design that allows a true synergy of different fields of expertise. This was a very significant characteristic of the service development process within the G-MOSAIC project. The benefit of this type of service organisation is maximising the collective potential of the specific expertise of the consortium partners. The consortium supports the view that similar organisation of work practices should therefore be implemented in future GMES initiatives.

-Flexible product design, tailored to user needs. One of the major lessons learned through working with the GMES user) community in the Security field is that each crisis or event is unique, deferring in terms of both observation and information requirements. Consequently, each product calls for specialised customisation. For instance, with the Rapid Geospatial Reporting (RGR) service the main lesson learned is that the information created during a crisis has a contextual value for that crisis. The accent of the value was on the service itself, on the availability to react in short time providing the appropriate information to support decision-making. Further to this, the quality of the tailored products to user needs is enhanced as a function of the intensity of collaboration between service providers and users. This involves the definition of the areas of interest and of the observation modes.

The evolution towards operational services relies on the effectiveness of two parallel assessment processes, external and internal to the production:
1)The assessment of the service by the final user, who assesses the products received against the initial requirements and expectations, and evaluates the usefulness of the information and its potential for integration into the user's operational workflow.
2)The assessment of the service by the consortium partners, which occurs through a number of internal processes. These include the product quality assurance process developed within the different service chains, and the independent validation carried out by partners who are not directly involved in the product generation process.
-The cost-benefit analysis. As well as the technical assessment, service enhancement and improvement must also be evaluated in terms of cost. Thus, a rigorous analysis of costs and benefits should accompany decisions taken on the evaluation of the service portfolio. The stakeholders should be able to develop an accurate understanding of the total service costs.
-User involvement. The engagement of users for the Security dimension of the GMES was a very challenging - and ultimately successful - activity within the G-MOSAIC project, despite the fact that only three SLAs were eventually signed. The strategy for the involvement of users should consist principally of stimulating and maintaining the dialogue among user communities and project partners, including service providers. This can be achieved through a variety of methods, including direct meetings and common events. As a recommendation to future projects, and as a follow-up to the User Workshops organised during G-MOSAIC, more technical and dedicated workshops and training sessions should be organised in order to elicit feedback and encourage open discussions. Such user engagement activities will assist in identifying areas of improvement in the products and services, thus preparing for the operational implementation of GMES Services for Security applications.
-Synergies between GMES services in the Security and other GMES services, in particular Emergency . The need for the exploitation of synergies between the recently operational Emergency Management Service (powered by the SAFER project) and the community of Security users was identified in the early stages of the G-MOSAIC project. A protocol for synergetic work between the two services was designed and implemented. Synergies with parallel and past GMES projects and initiatives should be ensured in order to optimise resources, capitalise on past experience and ensure that proper continuity in the relationship with the User community is sustained.
-An external validation service: Within G-MOSAIC, validation was performed on a best-effort basis by consortium partners, internal to the production workflow. As a recommendation from G-MOSAIC to on-going and follow-up Security-related projects, a system of independent validation (at least external to the production) should be set upto provide production support as well as controlling quality. Ideally, the validation should be undertaken by an external agent and integrated into the production chain. In order for such an external validation system to have a genuine impact on the quality of the products, validation should be carried out before the delivery of the product to the user. There is a risk, however, that such an arrangement may delay the final delivery of products - especially in the case of rapid mapping activities.
-Governance. One of the main concerns emerging as an outcome of the G-MOSAIC project is the lack of a governance model for GMES Security-related services. The need to design and implement a sustainable governance structure has been expressed both in recommendations from the user community and by several key actors involved in setting up operational services for GMES Security applications. The BRIDGES project, started in January 2012, coordinated by the EUSC is contributing to this process by proposing potential governance models, implementation options, stakeholder positions and cost-benefit analyses. BRIDGES is building on the preliminary assessment work carried out within G-MOSAIC.
-Coordination of the user-related activities between FP7 projects in the framework of Support to EU External Actions should be ensured, in order to provide the user community with a general and complete view of the GMES Security capabilities and to ensure a coordinated approach to user engagement. This may, for example, lead to shared events (workshops and co-located meetings) and communication and dissemination strategies.
-Integrated and centralised imagery analysis. The development of an integrated and centralised architecture for imagery analysis is a recommendation from G-MOSAIC. This architecture should allow secure access for retrieving, viewing and analysing all available (spatial and non-spatial) information for a given site. The information provided should include satellite imagery, Geographic Information System (GIS) information, external databases and collateral information. Symbology and map legends should be customisable by the service providers. A centralised model will maximise the return on the costs and resources committed in respect of fully operational activities, minimise the points of failure and improve usability, both for users and for service providers.
-The high reliance on non-European data providers should be evaluated. Many of the products and services developed within G-MOSAIC remain dependent on such data, and clearly this needs to be addressed before the operational phase.

Project Results:

The design of the G-MOSAIC project was based on a review of available technologies and service elements in order to build up a pre-operational service portfolio aimed at supporting different thematic domains. The Service Portfolio was based on successful outcomes in the following activities: scenario definition, user needs collection, service operation and validation. The definition of scenarios led to the combination of specific technological applications into pilot services.

At the end of the validation and appraisal activities, performed by the service providers in cooperation with the Users, the G-MOSAIC Service Portfolio, in terms of services provision and observation modes, was consolidated.

Furthermore, the pre-operational deployment of services included a Service Level Agreement/Service Level Specification exercise, aimed at outlining the applicable legal models for service distribution.

The G-MOSAIC service portfolio is composed of the following sixteen services ('service chains'); reflecting five European areas of interest and related priorities ('security domains'):

-Natural Resources and Conflicts
Exploitation of Natural Resources
Population Pressure
Land Degradation
Illegal Mining
Illegal Timber Logging
Illicit Crops

-Migration and Border Monitoring
Border Area Monitoring
Monitoring Migration Routes and Settlements

-Nuclear and Treaties Monitoring
Monitoring of Nuclear Decommissioning Sites
Continuous Surveillance of Nuclear Facilities

-Critical Assets
Critical Assets Monitoring
Critical Assets Event Assessment

-Crisis Management and Assessment
Contingency Plan Preparation
Rapid Geospatial Reporting (RGR)
Damage Assessment for Post- Conflict Situations
Support Reconstruction Missions After Conflicts

G-MOSAIC services were based on change detection techniques, specifically aimed at producing the following service types:

-Periodic updates of cartography: applicable to services focusing on the seasonal observation of phenomena (for example, in the case of agricultural targets) or analysis of historical data (e.g. population distribution).

-On-demand surveillance: applicable to services triggered by users on demand, which can be further sub-divided into:

- Off-line mapping: one-off observations for which the information delivery is not subject to tight temporal constraints;
- Monitoring: Recurring observations, for which the time window, update frequency, and timeliness of service delivery are pre-defined;
- Rush mapping: one-off observations, for which the products are delivered with tight temporal constraints in relation to the time of activation (ranging from 12 to 48 hours). The service allows for a degree of flexibility regarding the customization of products.

Data and geospatial information within the G-MOSAIC project have been distributed to users and to the consortium with different levels of security, according to a project-specific classification system based on the sensitivity of the information:
-Public: Available to the project and to the general public for subscription and download.
-G-MOSAIC restricted - Level 1: This level refers to a dataset not included in the European Space Agency (ESA) catalogue but available to other projects for subscription. Such datasets may only be re-used in case other projects make a request for satellite data over the same area. Transparency is maintained with the project which originally requested it (ESA will keep the information reserved). The products are either adapted for the purpose of wider dissemination, or the publication is delayed until the information is “declassified”.
-G-MOSAIC restricted - Level 2: The dataset is not included in the ESA catalogue. Information about the area of interest can be disseminated, but the distribution of products is non-public and subject to the restrictions specified by the user.

G-MOSAIC SECURITY DOMAINS
The series of the five security domains within G-MOSAIC were developed with the cooperation of users, who participated by providing requirements, contextual data and information and by assessing and evaluating the services. The domains bring together different services within the context of specific geographic or thematic boundaries. What follows is a short description of the G-MOSAIC Security domains, with some examples of the products delivered.

Natural Resources and Conflicts

This domain is focused on exploring the concept of early warning crisis indicators of regional conflicts. The exploitation of natural resources (such as timber and minerals), increases in population and the degradation of land are amongst the phenomena which have been examined in relation to occurrences of regional conflicts. The core information used in this service is land use and land cover change, which is combined with socio-economic data and local conflict information.

Environment and natural resources can play a role in the onset, duration, and termination of conflicts. Conflicts are not caused by environmental factors only. However, they can play a key role in the dynamics of conflicts. This role is complex and has an impact only in combination with other socio-economic factors. Generally speaking, it is often the lack of vital goods (such as food, but also access to land, work or housing) or the availability of valuable goods (such as diamonds, gold, etc.) coupled with poor economic conditions (poverty) and weak constraints (weak government, corruption, strong foreign influence, etc.) that could contribute to (violent) conflicts. Earth observation (EO) data could help to measure both environmental and certain socio economic factors. This could ultimately lead to the development of crisis indicators.

The indicator development in this domain tried to follow a modular top-down approach with three levels from global to sub-national. The indicators are developed for 'hot-spot' countries that are characterized by a high risk of armed conflict as defined in previous work based on structural indicators. Then three different service chains (exploitation of natural resources, population pressure and land degradation) aimed at providing conflict related information for these countries at sub-national level. The information is derived 1) from multi-temporal analysis of EO data aiming at identifying possible flash points for crisis within a country, and 2) through situation monitoring using information about conflictive events, population, socio-economic data, where available, and other relevant datasets.

The following three indicators were developed:
1.Exploitation of natural resources
2.Population Pressure
3.Land Degradation

Land use/land cover information was derived from multispectral Landsat and DMC satellite time series.

The approach was based on an adapted FAO Africover classification scheme in order to allow cross comparison and to assure continuity for other purposes. The biggest challenge for the service chain was the lack of appropriate satellite data. Despite search in GMES and other global archives, it was not possible to set-up a regularly spaced time series of satellite data. The achievable time series includes time slices from 2003, 2008/2009 and 2010 for the north Kivu province. Full coverage of both Kivus was achieved only with the data of 2010. This led to the production of land use/land cover maps and their changes over time. The recent land use/land cover maps were validated in the field by local experts, which were trained during a validation workshop in Goma organised by the project.

The combination of land use/land cover change and conflict information did not reveal any links. Moreover, the analysis did not show any link between conflict locations and the mining positions nor the population density. The reasons for this are manifold, but the most important is probably the insufficient EO data coverage to make a regional assessment, where land use/land cover changes in a period of 5 years are compared with more or less continuous conflict information. However the absence of a demonstrated link between conflict locations and the variables examined (land cover changes, mining positions, population density) does not mean that it does not exist; with more and better quality data the situation might well be different. Although the objectives were not fully achieved, the products were highly appreciated by local and international NGO's working in the area.

Worldwide, more people currently live in cities than in rural areas. The appeal of the city remains strong in particular in Africa, where currently the highest growth rates are observed. But the growth is mostly unbalanced and the gap between the rich and the poor continues to grow. In the long term these inequalities may pose threats to social and political stability ultimately leading to violent conflict. Earth observation (EO) data can help to measure the growth of a city over time and to map the urban inequalities. The service 'Population Pressure' aimed at providing refined geographical and thematic information related to changes in the city structure for the identification of potential conflict hot spots in larger urban zones.

The service was developed for Harare, the capital of Zimbabwe and it is connected with the 'Land Degradation' Service Chain. Zimbabwe has a recent history of violent, sometimes racially-motivated conflict over land tenure and for control of its government. These issues are shared with several other countries in the Southern African region. Despite the establishment of a coalition government, the political situation remains tense. The international community has consistently condemned political violence carried out by the ruling government, especially after an urban slum demolition initiative in 2005 called 'Murambatsvina', which left an estimated 700,000 people without jobs and/or homes. The humanitarian consequences of the operation were enormous.

Satellite imagery is an alternative source of independent information for monitoring processes such as Murambatsvina. In this spirit, the service provided built-up maps (2004, 2006, 2009) based on SPOT 2 and 4 as well as RapidEye satellites. Such detailed mapping of built-up areas is key information needed in any crisis at any phase, because it is a direct proxy for population. The built-up area of a city can be automatically extracted from satellite data. In analogy to the well-known compact and robust NDVI, a general index related to the presence of built-up based on image texture was used.

Subsequently, the built-up index was used to derive built-up change maps for 2004-2006 and 2004-2009. Multi-temporal built-up area mapping allows the mapping of changes over time to illustrate the urban development. In the case of Harare, Zimbabwe, some noticeable areas occur in the urban fringes. The changes can be interpreted using the base imagery. EO based change detection allows synoptic analysis of large areas for the identification of change hot-spots for further analysis.

Besides mapping the extent of built-up areas, the service provided products, which characterised an urban area based on its spatial variations according to building types, size, height and material, as well as social differences and usages (residential, industrial, commercial).

Characterising the built-up area based on its physical characteristics allows the inference of links to population occupancy and economic status (e.g. poor slums, middle class residential). Possible application areas are in urban planning and monitoring of urban development. Furthermore, by providing refined geo-spatial information, it can support the preparation of censuses. The information provided was derived from very high spatial resolution GeoEye-1 satellite data.

The combination of the settlement information with conflict data (ACLED) did not reveal any specific link.

The reported conflicts are mostly linked to the general unstable political situation and do not specify specific locations within the city. Nevertheless, when combined with other socio-economic information it is possible to identify future conflict potential.

The 'Land Degradation' service uses, similar to the 'Exploitation of Natural Resources' service described above, land use/ land cover information derived from EO data in combination with conflict and other socio-economic data to derive information about hot spots of politically influenced land degradation as a potential source of conflict in the future. The service was developed for two areas of interest. First, an area in central Zimbabwe south of the capital Harare, and as such linked to the area of interest for the population pressure service, and second the trans-boundary watershed of the Umbeluzi river between Swaziland and Mozambique.

The service chains included in the 'Crisis indicators' activities were able to produce relevant background information for conflict prevention, even if it was not possible to develop crisis indicators based on land use changes. This is owed to the complexity of the interactions between land use/land cover and conflict. It was aggravated by the often scarce data coverage (in particular for Kivu/DRC). Nevertheless, maps combining conflict and land use information provide an added value for decision makers at different levels. In addition, the activity demonstrated the feasibility of validation on the ground even in crisis situations. This was possible thanks to links established with local/regional entities, both governmental and NGO's.

The security domain Natural Resources and Conflicts addresses also the identification and characterisation of potential illegal activities (Illicit Mining, Timber Logging and Crops), in relation to their role in the development of conflict.

Illegal Mining. Exploitation of natural resources in most cases impacts the environment, often even spoiling natural habitats and affecting people's livelihoods. In special cases, the exploitation and trade of minerals can even fuel armed conflicts, as is the case in the East of the Democratic Republic of the Congo (DRC). The country is said to be one of the richest in the world in terms of mineral resources. Economically, however, it is one of the poorest. The lack of precise geographic information is a critical issue in the prevention of, and response to, on-going crisis situations relating to natural resource exploitation. Mining areas in the eastern DRC are often very difficult to access because they are widely dispersed, far away from road infrastructure, or regarded as unsafe. Field studies are fraught with difficulty and danger, as extremely bad road conditions hamper access and researchers studying the area run the risk of violence from rebel groups and militias. In this context, where armed conflicts and the militarisation of the mining sector have made traditional field assessments almost impossible, the use of remote sensing techniques has proven beneficial for the detection of widespread mining sites. Satellite monitoring thus provides a tool to complement field-based monitoring studies.

The pilot service for the monitoring of 'Illegal Mining' developed within G-MOSAIC provides conflict researchers with relevant information about mining areas and their surroundings, in order to contribute to focused reactions during conflicts, and to support the rapid identification of meaningful geospatial context information.

In the beginning of the project different areas of interest were defined together with the users (BICC, EEAS, Global Witness, IPIS and Resource Consulting Services). For an analysis at regional scale three large areas (spatial coverage: 2500 km²) were identified to get an overview of the situation on the ground and to detect hot spots of mining activities, which are subsequently analysed at local scale (spatial coverage: 100 km²). The study sites differed in landscape and topographic characteristics, in order to allow the development and testing of robust and transferable analysis approaches.

For the pilot service different product types were developed:
(1) Geographic Reference Maps, providing basic geo-information such as road and river networks and settlements,
(2) Potential Mining Maps including change detection, highlighting those areas where mining activities might be taking place and
(3) Information Dossiers providing detailed information on product generation and major results.

Within the service chain evolution several intermediate products were developed such as the phenological analysis, i.e. the use of medium resolution Earth Observation data (specifically MERIS) to detect anomalous decreases in vegetation vigour as indicators of mining activities.

For the extraction of potential mining sites a combination of semi-automatic object-based image analysis (OBIA) and further geospatial refinement was applied. The workflow was organised in two stages:
(1) a transferable feature extraction scheme for the detection of bare soil areas, indicating potential mining sites, by using very high spatial resolution satellite imagery, and
(2) further refinement after the initial classification by using relevant ancillary information (e.g. vector data such as roads, rivers and settlements) to reduce the number of false alarms.

A special focus was set on defining a classification rule set based on stable image characteristics, to ensure transferability to other natural environments and to different sensor systems.

In summary, the activities of the pilot service 'Illegal Mining' demonstrated the applicability and usability of Earth Observation techniques for supporting the monitoring and documentation of mining activities in the DRC. New developments should place an increased emphasis on in-cooperating users' feedback and further specified needs such as the provision of detailed information on the mining activities, e.g. estimation of extracted volume.

Illegal Timber Logging. The Democratic Republic of Congo (DRC) comprises the greatest extent of tropical rainforests in Africa, covering over 100 million hectares. The country has been exposed to industrial logging since the 1920s, which gradually developed until the majority of DRCs forests were under large-scale logging concessions. By the 1990s, the main cause of deforestation was industrial activity, including the extractive industries, large-scale cattle ranching, and extensive agriculture. Illegal logging is a major driver of deforestation, and the cause of severe social, economic and environmental damage. From 1990 to 2005, forest cover in the DRC decreased by nearly 3%. The United Nations estimate that at the present rate of exploitation, more than two thirds of the Congo Basin forest could be lost by 2040.

The overall scope of the pilot service 'Illegal Timber Logging' developed within G-MOSAIC is to provide information support for the surveillance and situation awareness in the field of natural resource exploitation. Emphasis was set on the evaluation of nature and extent of logging in order to support intelligence and early warning. Together with the users (BICC, EEAS, Global Witness, IPIS and Resource Consulting Services) different study areas in the provinces Bandundu, Équateur and Orientale were selected, covering industrial-scale logging in a concession and clearings for small-scale swift farming.

For the pilot service different product types were defined:
(1) Geographic Reference Maps, providing basic geo-information such as road and river networks and settlements,
(2) Timber Logging Maps, highlighting those areas where logging activities are taking place and
(3) Information Dossiers, providing detailed information on product generation and major results.

Within the service chain generation several intermediate products were developed such as the phenological analysis, i.e. the use of medium resolution Earth Observation data (specifically MERIS) to detect anomalous decreases in vegetation vigour as indicators of logging activities.

The overall methodology used for the identification of logged areas and related changes is based on both, optical and radar satellite data. Since optical radiation does not penetrate through clouds, a special emphasis was placed on the combination of optical and radar data. Radar-based analysis provides information about logged areas in a region irrespective of cloud coverage, whereas detailed information about land cover changes is obtainable using optical satellite data. Change detection based on SAR images covered pre-processing, speckle filtering and additional false alarm reduction. The detected changes flow in an information management system that is able to query changes, view changes over longer periods of time, combine changes with other geospatial information, and to help interpretation. The information extracted by using optical data was derived by means of multispectral land-cover classification and image differencing analysis. In this respect, the use of vegetation indices derived from satellite data (such as the Normalized Difference Vegetation Index, NDVI) proved to be useful in identifying vegetated and non-vegetated areas. Five types of transformation were described and mapped accordingly, i.e. seasonal variation in natural vegetation, deforestation, clearances due to the agricultural cycle, regrowth after 'slash-and-burn', and the growth of secondary forests.

To summarize, the products developed within the 'Illegal Timber Logging' pilot service, demonstrate the combination of different sensor systems and algorithms which can fulfil the requirements for the localisation and identification of logging areas. Nevertheless for a more detailed and in depth analysis, further research should include the analysis of very high resolution satellite imagery. Furthermore advanced validation in the field will support method development and evaluation.

Illicit Crops. Many countries around the world are affected by the devastating consequences of drug abuse and illicit trafficking: adverse effects on health; an upsurge in crime, violence and corruption; the draining of human, natural and financial resources that might otherwise be used for social and economic development; the destruction of individuals, families and communities; and the undermining of political, cultural, social and economic structures, as expressed in the United Nations Declaration on the Guiding Principles of Drug Demand Reduction.

The production of the poppy or opium plant was introduced in Peru some years ago. Currently, there are poppy plantations in the departments of San Martin, Amazonas, Huanuco and Cajamarca (Cortegana, one of the study areas within this service).

Colombia is the location of approximately 43% of the world's coca cultivation, although this percentage is decreasing mainly due to eradication. The majority of the cultivated area (79%) is distributed across 8 departments, one of them corresponds to a study area within this Service, Nariño. This department has the highest amount of land under coca cultivation with 24% of the total coca cultivation in the country. In contrast, the department of Antioquia represents 7% of the 2009 total coca cultivation in Colombia.

The core products developed during the G-MOSAIC project are:
-Potential areas of coca cultivation maps (Colombia) using optical or radar data, which provides indicative locations of potential coca plantations.
-Potential areas of poppy cultivation maps (Peru) using optical data, which provides indicative locations of potential poppy plantations.
-Mature Coca Index (MCI) using optical data is an index expressing the extent of mature coca plants on the ground.
-Phenological analysis, detecting anomalous decreases in vegetation vigour as an indicator of illegal activities.

Migration and Border Monitoring

The Security domain 'Migration and Border Monitoring' objectives could be synthesized as follows:
-identification and monitoring of main migration routes, along terrestrial border and around coastal areas of the uncontrolled flow of migrants through a monitoring of migratory routes extending from unstable countries to third countries borders
-identification of migration origin and destination country, migration routes
-develop an early-warning system in order to anticipate and monitor destabilizing mass population movements.

In particular the service chain Border Area Monitoring puts a focus on the monitoring of border regions and the observation of border crossings and related infrastructure along the borders.

All products offered by the Border Area Monitoring service aim at providing monitoring and surveillance services and to support situation awareness in border areas and in areas and/or along routes subject to migration and temporary settlement/urban extensions. The products provide a significant intelligence and decision support in order to help stakeholders involved in issues related to border control, migration and (temporary) settlements.

The service chain Monitoring Migration Routes and Settlements puts a focus on the monitoring of migration routes, settlements and population movements. This includes the observation of long range migration routes and of settlements along the routes.

All products offered by the Migration and Border Monitoring service aim at improving (partially already) existing monitoring and surveillance services and to support situation awareness in border areas and in areas and/or along routes subject to migration and temporary settlement/urban extensions. The products provide a significant intelligence and decision support in order to help stakeholders involved in issues related to border control, migration and (temporary) settlements.

For both the service chains included in the Migration and Border Monitoring security domain, the services can be offered worldwide; within GMOSAIC the focus was on Eastern Europe and Central Africa.

Users come from various domains, but mainly public, civil security and military stakeholders are addressed.

The main Science and Technology (S&T) results and foreground of the activities developed within this Security domain are as follows:
-Continuous adjustment and improvement of the services' workflows with special focus on the implementation of the appraisal's period user feedback
-Contributing to the project's use on Eastern Europe and Central Africa with selected elements of the core service chains
-Validation of products by the users and advancing the services into operations

With reference to the individual products generated some specific technical achievements are summarised hereafter:
'Border Permeability Map' (BPM) and 'extended Border Permeability Map'

The BPM provides a description of the ease with which a border can be crossed, taking into account the topography and the land cover along the border as well as the surveillance measures on site (Border guards, check points etc.)

In the extended version of the BPM additional information about the trafficability is taken into account. By using high-precision data and additional parameters such as snow cover or slope direction a more precise modelling is enabled than using only the BPM.

The BPM is produced on the basis of three different criteria / indices: 'Walk', 'Hide' and 'Secure'. These three indices characterize a site and / or a region in terms of its suitability to be passed by foot (walk), for a hiding place (hide) or for being crossed discreetly (secure). For example the BPMHide product displays the possibilities available for a migrant to hide along his way towards the border. By experience there will be more suited locations for a migrant to hide further away from inhabited and in sparsely and / or inaccessible areas. For determining the indices and finally drawing the map the following information layers are analyzed: Land cover, DEM and slope, night-lights, and population density.

Finally, the 'extended BPM' consists of a BPM plus a so-called 'Trafficability Map'. The latter is based on a trafficability analysis which aims at combining and modelling parameters that influence the ability of vehicles getting moved / driven ('speed factor' dependent on terrain conditions). The parameters considered for the trafficability analysis are terrain slope, soil type, forest density and land cover type.

'Reference map (Basic) ' and extensions 'Population density map' and 'Road network map'

These product types correspond basically to a topographic map featuring the main features on the earth's surface. Modifications have been specified with special focus on population and network: For the first the information is derived from the grid world population database and a recent land cover map. For the latter ancillary data is used to better locate roads and to better assess their accessibility. The main results of the mapping process, for example applied to the border area DR Congo /Burundi, are represented by a digital geo-database containing topographic features extracted by visual interpretation of SAR imagery and physiographic features extracted by semi-automated contour line and spot height derivation using elevation data (here SRTM DEM was used). Eventually, four maps at the scale of 1:50.000 in DIN A1 page format were created. An overview map with a scale of 1:100.00 in DIN A1 format was produced as well. Each map layout is available in two versions: One version displays the extracted features only. A second version displays the extracted features on top of the imagery used as main data source (here TerraSAR-X imagery was used). The results achieved this way are based on standard feature extraction rules (here: FACC for a scale of 1:50.000).

Users highlighted the methodological shortcoming in regards to the changes in weather condition.

They made several recommendations and suggestions in order to improve the products (to use an additional annual/monthly weather condition information: the precipitation forecast, the soil moisture, etc). Finally , Users were highly interested in the analysis and they encouraged to extend the analysis with aid of additional dataset, making them more suitable to the changeable climate conditions over the whole Europe.

Nuclear and Non Proliferation Monitoring

The Nuclear and Treaties Monitoring security domain , comprises the monitoring of known Nuclear areas of the world where treaties have been established for the control of the state's nuclear activity, the dismantling of the area, etc.

Thus, the 'restricted' GMOSAIC security services dedicated to nuclear activities have been included in this topic.
This domain supports efforts for preventing and combating the proliferation of Weapons of Mass Destruction and verifying compliance with the relevant treaties. There are two services in this domain:

-Monitoring of Nuclear Decommissioning Sites, with the objective of monitoring the decommissioning of nuclear sites in order to verify compliance with nuclear non-proliferation treaties.
-Continuous Surveillance of Nuclear Facilities, with the objective of verifying that nuclear material and facilities in the selected sites are used for civil purposes, and the assessment of whether the initial declaration of material and facilities was complete and correct.

The pre-operational products developed within this security domain were based on Multitemporal Coherence Maps , interferometric analysis and change detection maps. The maturity of these products should be further improved in on-going of future GMES Security initiatives.

Critical Assets
The activities performed in the 'Critical Assets' Security domain are in alignment with the provisions of the European Security Strategy (ESS). Thepolicy frames accounted for service definition have been: EPCIP [COM (2006) 786, 787], EU civil protection: COM 2008/73/EC, Council Decision 2007/779/EC, and CSDP Common Security and Defence Policy (SEA).

The Strategy identifies five principal threats to European Security: international terrorism, the proliferation of weapons of mass destruction, regional conflicts, state failure and organised crime.

Specifically, in the European context, critical infrastructures are intended as the physical technology facilities, networks, services and assets that, if disrupted or destroyed, would have a serious impact on the health, safety, security or economic well-being of citizens or the effective functioning of governments in the Member States.

Critical assets comprise a wide variety of elements, either manmade structures or natural.

Natural Critical Assets not only to extreme natural hazards leading to natural and humanitarian chaos in need of reconstruction, but also to singular geomorphological elements that, under certain circumstances can become 'critical': wells in the dry Sahel season, gullies sheltering illegal route passages, family farmlands irrigation ponds, cliffs or shallow water tables disrupting access, etc.

The catalogue of natural critical assets has been carried out by GMOSAIC team with the purpose of using those features as background reference. In other words, no environmental monitoring of natural assets or phenomena is intended when extracting natural assets features.

An example may clarify; a flood plain is considered critical and recorded as such in so far as, during the rainy season it may affect the local road network. Hence, the flood plain is thus depicted as a natural critical feature into the geodatabase. The analysis of specific flooding events or rhythms falls beyond the objectives of the service.

Man-made critical assets are widely acknowledged as objects associated with property and economic activities, generally coincident with civil engineering critical infrastructures.

Close attention should be noted upon other man-made critical assets, not always linked to civil works, such as frontiers and crowd gathering events. Overall, there is a wide coincidence between man-made critical assets and what is commonly addressed to as critical infrastructures (civil works). Quite often, natural and man-made critical assets concentrate in space giving rise to the so called critical areas, herein understood as territories where the adjacency of singular criticalities add up to greater dangers, risks, vulnerabilities and potential chained impacts.

The products offered within this Security domain are aimed at improving existing monitoring and surveillance services and support situational awareness in border areas and along routes subject to migration and temporary settlements or urban extensions.

Even if the main critical assets referred to in the normative frames mentioned above , the focus on man-made civil engineering infrastructures, the service developed within the GMOSAIC project combined natural critical assets and man-made critical assets. The perception of criticalities linked to specific assets varies across countries, time and relevance of economic activities. GMOSAIC mostly services security needs outside Europe. Critical assets are rather different in African, south American or Asian contexts to those defined for Europe or north America; differences lay in volume, concentration, maintenance, construction materials, assigned security buffers, etc. Hence, EO data analysis for critical assets monitoring in developing countries requires skilled image interpreters for feature extraction.

The countries monitored during the project duration were: Haiti (Gonaïves area and Port-Au-Prince area earthquake monitoring), Colombia (various activations for the monitoring of inundations during 2010), Chile (earthquake monitoring) , Guinea Bissau, Burundi,Democratic Republic of Congo Libya (activated within RGR mapping during the Arab spring), Bulgaria ,Russia , Poland/Ukraine .

In the context of critical assets protection, the services were intended to provide support for critical asset monitoring in two different scenarios:
-Monitoring critical assets to control and prevent possible events or actions that may threaten citizens security
-Post-crisis assessment to support follow up activities, after the detection of security breaches

During the project duration, the service was activated for 9 countries worldwide.

The core products delivered (for 20 activation sites for Critical Assets Monitoring and 7 activation sites for Geospatial support for crisis) were:
-Baseline and update maps of natural critical assets (by optical and sar images processing)
-Baseline and update maps of man-made critical assets (by optical and sar images processing)
-Change detection of critical assets (by optical and sar images processing)
-Low resolution thermal anomaly detection

G-MOSAIC has required a previously non-tested flexibility, on the side of Service Provider. End users have requested a much faster service than ever provided before, and in a variety of environments beyond any planning. For instance, the need to service ever cloud covered areas in the equator-tropical world, has tested the limits of VHR optical data, cloud masking or multi-temporal cloud-free mosaics, together with the benefits of the coordinated use of SAR data.

Yet, another lesson learnt may be catalogued under the subject of 'scale'. In this case, scale refers to the following aspects:
-The 'institutional scale' of the end user: GMOSAIC work package providing EO services for critical assets has handled requirements from international agencies as well as from national administrations. The range of requirements varies considerably but, in the context of maximizing the use of VHR EO data for security purposes, no end users should be despised as 'small users', just by the fact of not enjoying the status of a large international agency. The national demand market share for the security of national critical assets is particularly high in those geo-political areas that are consolidating democratic processes.
-The actual 'spatial scale' of the critical asset under analysis. The experience of over 20 activations handled throughout GMOSAIC shows that the spatial coverage demanded may vary from large geographic regions (e.g.: water sheds regulated by dams) to just few hundred meters occupied by the critical infrastructure at stake.
-Security domain and cross cutting aspects: Critical Assets are ever-present cross cutting elements to security, safety (and emergency response). Relevance and specificity requires individual analysis, extraction methodology, legend and GDB structure definition.
-User involvement and requirements collection: importance of large institutional end users but National end users not to be despised.
-Validation activities: the field validation has been not always carried out due to budget constraints.
-Non EO data supply: end users were not fully open to share in situ data.
-Liability and SLA: positive reception of services, but users demonstrated to be reluctant to commitments.

Crisis Management
The services provided in this Security domain aim at providing geospatial intelligence in response to crisis situations. 'Crisis' in this context refers to wars, civil conflicts or natural disasters leading to humanitarian emergencies. The services developed by G-MOSAIC in the context of crisis management and assessment address different phases of the crisis cycle:

-Planning and preparedness, responding to user needs related to a 'pre-crisis' situation. The Contingency Plan Preparation service supports planning for civil evacuations and the preparation of operational plans for strategic missions, through the provision of geo-spatial information (e.g. city maps). The service has been designed for situations of forthcoming crisis, addressing areas in which there is a need for imminent action;

-Crisis response, through the provision of information oriented towards rapid response (e.g. RGR). The service provides information on transportation networks, logistics, facilities (e.g. hospitals), critical assets and infrastructure (e.g. airports, harbours), highlighting helicopter landing areas and areas appropriate for use as population gathering sites;

-Post-conflict damage assessment and reconstruction, in order to improve the rehabilitation process, notably through the provision of change detection products.

The first phase address the needs related to situation of forthcoming crisis, in the area in which an imminent action is foreseen. The main Users are the national foreign affair ministries which are in charge of repatriating EU citizens from crisis areas outside EU lacking of embassy or a consular post, according to the EU policy concerning consular protection in third countries. In the occurrence of a crisis contingency plans are prepared and maintained in the areas of interest, providing information for planning civil evacuations and preparing strategic military interventions. A typical situation is the evacuation of EU citizens from an hazardous city, with a monitoring service of road network and some critical assets - bridges, airports, city exit-ways. This evacuation plan could be combined with a parallel military intervention for crisis management.

The contingency plan preparation was based on the delivery of the following products:
- Urban map, derived from an optical satellite orthoimage with geo-spatial information in support of emergency planning for evacuation and intervention for the greater city area.
- Multitemporal Coherent analysis based on SAR, false colour composite images obtained from linear combination of the SAR detected interferometric coherence of two CSK images (GSDS 3 m) acquired in interferometric mode, including a report on the MTC Map analysis results focusing on land use/land cover changes.
- Detailed Land Cover and City Map, produced from very high resolution satellite orthophoto, IR band, 50 cm spatial resolution.

These service chain products have been provided on several areas of interest: Harare (Zimbabwe), Asmara (Eritrea), Tyre (Lebanon), Al Daraweh (Egypt), Aden (Yemen)

A test with updated City Maps extracted from VHR satellite imagery was performed on Pretoria College in Pretoria area, where the Italian football team was based on the occasion of the South Africa 2010 World Cup Championship.

The RGR Service was oriented to provide operational response, however it was implemented in a Research and Development (R&D) environment.

The service evolved during the project, developing a basic portfolio adjusted to the user needs, which included the following products:

-Reference maps, including several informative layers as transportation networks, logistic facilities (hospitals and first-aid centres, security forces, fire brigades, administrative buildings, educational and sport areas, etc); critical assets and infrastructures (airports/airfields, harbours, industries, communications lines, helicopter landing areas, population gathering areas)
-Rapid damage assessment : providing qualitative estimation of the damages, spatial distribution of damaged assets and infrastructures
-Radar analysis: related to major changes (detectable regardless of weather conditions)
-Activity analysis and reporting: Interpretation / explanation of the nature of changes
-Evacuation support: Several features related to evacuation activities, such as potential helicopter landing areas and evacuation routes from gathering areas
-Critical infrastructure analysis: Airports, harbours and industrial installations, such as refineries
-Automatic change detection: Detection of changes in large areas using trained algorithms

The experience gathered during the several activations carried out, contributed to demonstrate the following results:
-The service was able to respond operationally to the needs of the user.
-The products delivered by the service were fully operational and highly appreciated by the users.
-The RGR service contributed to the Research and Development (R &D) community, developing concepts of geospatial validation.
-The RGR service established the background of a cooperative and coordinated approach of rapid geospatial reporting to be developed in future iterations of this type of services.
-The RGR Service implemented protocols to keep information confidential whenever requested by the user.
-The RGR Service played an active role in coordination with other projects and service of GMES through approved (EC, REA, GMES Projects involved) guidelines for synergistic approach on complex emergencies

Finally, after a crisis a detailed analysis of the damage resulted by the crisis is needed.

Wars or civil wars have dramatic humanitarian and financial consequences to the affected countries. The damaged infrastructure (e.g. houses, streets, industry, electricity and water network), hinders the normal life for the population for a longer period even after the combat operations. The impact on the society is also through indirect damages e.g. to the normal flow of goods and services. Post-conflict situations therefore call for the assessment of damages to infrastructure and buildings. Such assessments, which are the key output of the pilot service 'Damage Assessment for Post-Conflict Situations' (DAP) supply core information for the overall assessment of the severity of a conflict and its consequences, and support needs assessment and recovery planning. On the other hand, the 'Support to Reconstruction Missions After Conflicts' (SRM) pilot service offers e.g. (1) longer-term monitoring of reconstruction operations, which allows donor agencies to ensure that funds are being correctly allocated and utilised and (2) up-to date and reliable information for the repatriation phase of refugees, the awareness of the environmental circumstances in the refugees' area of origin and potential long-term environmental impacts of the refugee/IDP camps.

In the beginning of the project different scenarios were defined together with the users from EEAS, BmEIA, and UNEP. For four areas of interest (damage assessment in Georgia and Haiti, reconstruction monitoring in Indonesia, environmental impact assessment of a refugee camp in Chad), satellite data of various time slots (before conflict, directly after conflict, sometime after conflict, i.e. data one year apart) were ordered. Two service chains were developed and applied to the different data sets.

For automatic change detection a high co-registration accuracy is needed, which can be achieved by automatic co-registration processes by now, unless the poor data quality (e.g. cloud coverage) hinders the processing. The main step is the change detection of the different time slots and the interpretation of the results. Main technique in crisis mapping is still visual (manual) interpretation of the data. Automatic methods are developed as described below. Finally, the results are presented as maps, web-based geoinformation (e.g. kml-files) and explanatory reports, where the content can be modified according to user requests. Depending on the level of detail the damage assessment maps show the extent and location of damages to buildings and infrastructure (e.g. bridges).

During the project, new change detection methods were developed and analysed. Different approaches were followed: pixel-based as well as object-based methods on optical image data in 2D, 3D (stereo data and derived digital elevation models), 2D change detection of SAR data, as well as post-classification change detection. Some of the methods had to be discarded due to the poor accuracy of the results. Other methods led to promising results, scalability to larger areas of interest and further automation are main goals of the service development.

Further to the provision of pre-operational pilot services and products in support of GMES security activities, the G-MOSAIC project has demonstrated the long term sustainable provision of the services required by end users, by providing:
a) study of service interoperability and data exchange standards
b) study on architectural models and prototype suitable solutions for security services provision, addressing specific requirements related to confidentiality of sensitive data and information handling
c) a funding model for GMES Security Services

a)The activities related to the information exchange analysis, were aimed at maintaining a link between G-MOSAIC and the current European and international activities in interoperability and standardisation of geo-information in order to support GMES Security Services definition with state of the art knowledge on the structure of the geo-information.

b) The G-MOSAIC architecture was made up of the following components:
-WEB-GIS interface for the service domain information and data access; the EO data provided through the GMES Data Access System were processed into the G-MOSAIC service chains and the final products published through the Common Portal, with integration of external resources (external WEB-GIS, Globe Viewer facilities/other clients at users premises, in-situ systems, etc).
-Data mining interface (based on text or geographic based) for information search and retrieve.
-SSC: performance monitoring system.
-SHG (Shattering Grid): it is able to apply, according to the access level of the user , data filtering of maps (masking or blurring of areas with confidential contents) and vectorial data (hiding attributes and generalizing features).

Furthermore, the architecture was integrated with:
-Service Platforms, specific tools not included into the Common Portal acting as web service sources for the Common Portal, exporting to it the information according the OGC WMS standard by HTTP(S) requests/responses.
-SIL (Security Interoperability Layer ) which acts as a centralized security layer.

c) The feasibility of future GMES services for security applications in the domain of EU External Actions is dependent on an accurate assessment of the costs which such services will incur, both initially (as one-off investments to set up service infrastructure) and in the longer term (as regular and recurrent costs of maintenance). Following a methodology similar to that used within BOSS4GMES to assess the service costs of the Land, Marine, Emergency Response and Atmosphere services, the G-MOSAIC project has produced a preliminary analysis of the currently disclosed costs of GMES services in the Security domain.

Potential Impact:

The main results of the G-MOSAIC project relate to:
-the dissemination of the knowledge on GMES potential impact on Security related User Community and to contribute to building a political consensus on GMES Security Services
-direct involvement of users in scenarios definition, to be continuously updated based on feedback collection, in order to guarantee the integration of geo-information into their workflow and decision making process

The G-MOSAIC project contributed to the GMES sustainability, by:
-assessing a sustainable provision and funding model for GMES Security Services
-studying architectural models and prototype suitable solutions for security services provision, addressing specific requirements related to confidentiality of sensitive data and information handling
-providing effective involvement of user organizations in the services/products definition and deployment
-adopting a collaborative scheme between different Service Providers inside the Service Chains
-demonstrating the maturity of security services for operational use
-establishing an effective synergy with Emergency Response Service for a seamless response to the various User requests covering both emergency and security domains
-increasing visibility of the GMES Security dimension

List of Websites:
http://www.gmes-gmosaic.eu
218822-final-report-1176376.pdf