Final Report Summary - MELODIES (Maximising the Exploitation of Linked Open Data In Enterprise and Science)
Executive Summary:
The MELODIES project was conceived as a direct response to the European Open Data Strategy. Its principal objective is to demonstrate that the release of Open Data simulates innovation and growth, by enabling researchers, industry and government to develop new and improved services that benefit society in a variety of different ways. The project has achieved this by developing eight new real-world services, based upon Open Data and aimed at different users within the environmental sector. The services are underpinned by a cloud computing platform, which is designed specifically for the purpose of managing and processing large amounts of environmental data (e.g. from satellites). These services were:
* “Emissions Inventory Service”: produced information on land cover and soil moisture for use in estimating greenhouse gas emissions in the UK.
* “Urban Accounting Service”: combined, visualised and analysed open data about urban environments to support urban planning and management.
* “Land Management Service”: used satellite data to optimise the economic and ecological benefits of farming.
* “Ocean Status Assessment Service”: produced information on “Good Environmental Status” of oceans and seas, for compliance with the Water Framework Directive.
* “Desertification Indicators Service”: used satellite data to calculate indicators of desertification and land condition, to support monitoring required by the United Nations Convention to Combat Desertification.
* “Crisis Mapping Service”: developed a rapid mapping application for disaster response, and a European Historic Floods database for users in the insurance industry.
* “Marine Transport Service”: used data and modelling to improve predictions of when ships will arrive into port, avoiding costly and inefficient delays
* “Groundwater Modelling Service”: use models and citizen observations to understand and manage groundwater bodies
Many of these services have become operational at some level during the project. All the remaining services have concrete plans for further technical and business development, ensuring their continuation beyond the timeframe of the project.
During the course of the project, the Sentinel satellites started delivering open data, and we were able to take advantage of this and become early adopters of this extremely valuable source of data.
A key research theme throughout the project was the use of Linked Data. “Linked Data” describes a set of best practices for describing data and publishing it on the Web, in a manner that permits different data sources to be explicitly linked with each other. In this way, it is possible to break out of “data silos”, in which different datasets are managed and accessed completely independently – these “silos” mean that it can be very difficult to merge data from different sources.
The project was very active in disseminating its results to a wide audience, through over 150 meetings and workshops, the project website, a popular blog and social media. At the end of the project we held a three-day event in Brussels, in which we presented our results to a wide audience through talks, demonstrations, posters and fact sheets.
Project Context and Objectives:
The European Open Data Strategy establishes important new principles that ensure that European public sector data will be released at no cost (or marginal cost), in machine-readable, commonly-understood formats, and with liberal licenses enabling wide reuse. These data encompass both scientific data about the environment (from Earth Observation and other fields) and other public sector information, including diverse topics such as demographics, health and crime. Many open geospatial datasets (e.g. land use) are already available through the INSPIRE directive and made available through infrastructures such as the Global Earth Observation System of Systems (GEOSS). The intention of the Open Data Strategy is to stimulate the growth of research and value-adding services that build upon these data streams; however, the potential value inherent in open data, and the benefits that can be gained by combining previously-disparate sources of information are only just starting to become understood.
The MELODIES project (Maximising the Exploitation of Linked Open Data In Enterprise and Science) is a direct response to this Strategy. Its principal objective is to demonstrate that the release of Open Data simulates innovation through the provision of new services for users in research, government, industry and the general public in a broad range of societal and environmental benefit areas.
Open data are highly diverse and can be very large (scientific datasets can reach the scale of petabytes) and therefore require significant computing capacity and informatics expertise to exploit them to their full potential. This capacity is often beyond the reach of businesses (particularly SMEs), requiring large capital investments as well as highly specialist skills in several aspects of computing. Therefore, another key project objective was to build a computing platform that provides the service developers with the high-level computing and data management capabilities that they required. This allowed them to focus on the design and evolution of their service, enable collaboration on common tasks and contributes towards the long-term sustainability of the services.
Linked Data is an emerging paradigm for data publication that is intended to ensure that diverse datasets are joined up, and that data can be reused by the widest possible audience, beyond the specialist user community. The MELODIES project applied Linked Data methodologies to test their efficacy in developing real-world services, building heavily on previous European investments in the field.
In summary, the main objectives of the project were:
1. To combine multiple open data streams (including Earth Observation and government data) to develop eight new innovative services covering a range of domains including agriculture, urban ecosystems, land use management, marine information, desertification, crisis management and hydrology. These services combine Earth Observation data with other open data sources to produce new information for the benefit of scientists, industry, government decision-makers, public service providers and citizens.
2. To apply Linked Data and Semantic Web approaches to enhance the services and enable multiple sources of data to be discovered, interpreted and cross-referenced, thereby generating new knowledge;
3. To ensure that the services are sustainable beyond the lifetime of the project by considering cost-effectiveness, improving data quality and maximizing their user base, whilst also considering the direction of travel of wider initiatives;
4. To deliver reusable technology, building upon previous e-Infrastructure investments, that facilitates the exploitation of open data by both the MELODIES services and by future services outside of the project;
5. To implement a comprehensive and coherent strategy of user engagement and dissemination ensured that the project develops in response to real user needs;
6. To demonstrate the value of open data and stimulate the release of more open data by data providers.
Project Results:
= Overview =
The overall structure of the project is as follows. The eight services (applications) are developed in workpackages 3 to 10 respectively. Some of the eight services are be further divided into “sub-services”, reflecting distinct streams of activity within their workpackage. Workpackage 2 encompasses the development of the technical platform, which underpins the services. Workpackage 11 includes activities relating to ensuring the sustainability of the services, and workpackage 12 includes dissemination activities; both of these are cross-cutting workpackages in which all partners have participated. Workpackage 1 encompasses project management activities; we do not describe these further here.
The main results of the project are given here, organised by workpackage and focusing on the technical workpackages (2 to 10 inclusive). This information is suitable for wide publication. In some cases, for commercial reasons, not all information about the results can be made public. In these cases, further information can be found in Section B.
The text below contains many links to the MELODIES website, where presentations, videos and other dissemination material can be found. The reader is also referred to the introductory video, poster and factsheets that are attached to this Report.
= WP2: Technical Platform =
The primary purpose of the project’s Technical Platform was to provide underpinning technology to support the development and deployment of the services. However, this workpackage yielded a number of technologies of its own, which will be exploited further in future projects.
Cloud computing platform: The IT infrastructure that forms the basis of the Platform is a cloud computing platform, built and operated by Terradue. One key feature of this is the ability for service developers to start their operations at a small scale (using modest computing resources), and then easily scale as their requirements grow. In return, the service developers in the MELODIES project have provided very valuable requirements (e.g. for access to particular software packages and/or key datasets) that have helped Terradue to improve their platform. As a result, the cloud platform has been significantly improved within MELODIES and is now being made available for much wider use, both within the context of research projects (e.g. NextGEOSS and CoReSyf) and commercial use. The platform will therefore be sustained through a mixed model of R&D and commercial revenue. More information can be found at http://www.melodiesproject.eu/content/melodies-shared-platform.
Linked Data tools: The use of Linked Data techniques was a key technical research element of the MELODIES project. (“Linked Data” describes a set of best-practice approaches for publishing and processing data, using widely-adopted open standards, allowing previously-separate datasets to be brought together more easily.) This approach is not specific to geospatial or scientific data: many kinds of data (e.g. library catalogues, socioeconomic data) are published in this way. Therefore one challenge of the MELODIES project was to develop technology that applies Linked Data techniques to complex and high-volume Earth Observation data. The University of Athens led this part of the project, and have developed a suite of software tools for handling geospatial Linked Data. These range from tools to convert data from existing formats into Linked Data (GeoTriples), through database tools (Strabon and Ontop-spatial) to Linked Data visualisation tools (Sextant). All are available as open source software and can be exploited by anyone. Development of these tools is ongoing through other research projects, including BigDataEurope. More information can be found at http://www.melodiesproject.eu/software-tools.
CoverageJSON data format: During the project we observed that there was a gap in the capabilities available to the developers of websites and apps who wish to consume scientific data. Existing data formats (e.g. NetCDF, HDF, GRIB, XML) were either highly complex or too inefficient to be used in these environments. Therefore we developed a new data format, known as CoverageJSON, which is designed as a format for encoding many kinds of scientific and Earth Observation data in a manner that is friendly for web and app developers. (In web development, the pre-eminent data format is JSON, the JavaScript Object Notation.) The format is based on the concepts and standards from ISO and OGC and its specification is published openly on the web (https://covjson.org). The spec is supported by documentation in the form of a Cookbook, plus a suite of open source tools for producing and consuming the format. Great interest in the format has already been shown by the user community, who are also encouraged to contribute to its development. CoverageJSON has also been discussed (at the recent INSPIRE conference) as a possible future standard for INSPIRE compliance, as it is perceived to be simpler than existing mechanisms. This strand of work is led by the University of Reading.
ncWMS: Another popular method of building data-driven websites and apps is to use server-side software to turn complex data into images, then serve these images over the web (in a similar manner to Google Maps and other mapping applications). NcWMS is a piece of software, developed by the University of Reading, which does this in a manner specified by the OGC’s Web Map Service specification. NcWMS was originally developed before the MELODIES project, but we have made several improvements and released these to the community (ncWMS is fully open source). These improvements have included support for more data types (including “unstructured” grids common in the coastal-ocean community), a greater choice of parameters when generating images, improved documentation and a refactored code base that is more conducive to wide reuse. More details can be found at https://reading-escience-centre.github.io/ncwms/.
= WP3: Improving emissions inventories with Earth Observation data =
The United Nations Framework Convention on Climate Change (UNFCCC) requires governments to report the greenhouse gas (GHG) emissions from their countries. One way of estimating agricultural GHG emissions is to use information about land cover and land use – different types of land use produce different amounts of GHGs per hectare. GHG emissions are also affected by soil moisture, although there is little available information on the real levels of moisture in the soil. This workpackage encompassed two related “sub-services” that both contribute to helping the UK government to improve its GHG emissions. In both, we worked closely with the Department for the Environment, Food and Rural Affairs (Defra) as well as the Department for Energy and Climate Change (DECC).
New Land Cover Map: Currently the UK uses a particular land cover data product from the Centre for Ecology and Hydrology, which is produced once every few years. Defra has a requirement for more frequent (yearly) updates to the UK land cover map, although they would be satisfied with a lower spatial resolution than the current product. In MELODIES we produced a new yearly Land Cover product using only open data (MODIS and ASTER satellite data, plus the latest CEH land cover map) and open source software (Python, GDAL). Based on user requirements, the new product is focused on accurately estimating land cover *changes* from year to year, in order to generate a highly self-consistent product, which avoids some of the inconsistency problems that are associated with other data products. The data product will soon be available as Open Data from the CEH Environmental Information Platform (http://eip.ceh.ac.uk). The code used to generate the product has been published and will be applied in future projects. The UK government has issued an Invitation to Tender for suppliers to produce this data product (or an equivalent product) on an operational basis, so there is a clear opportunity for this research to have a tangible impact upon the government’s operations. For more information, please see http://www.melodiesproject.eu/node/190.
Soil moisture data portal: CEH has “rescued” a very valuable set of soil moisture data from neutron probe measurements from 1966-2013, much of which was stored on fragile magnetic tape. This dataset (356 tubes across 64 sites) has been checked and corrected, and linked with other data on land cover and soil type. These measurements were expensive and time-consuming to collect but are potentially very valuable for validating other results from cheaper sensors or satellites. The data are published on the CEH Environmental Information Platform, and also through a specially-developed “soil moisture explorer” application, which allows the user to visualise the soil moisture data in concert with other data, including the new MELODIES land cover map (see above). For more information, please see http://www.melodiesproject.eu/node/191.
= WP4: Urban ecosystem accounting service =
The rapid growth in urbanisation across the globe presents many challenges. Policymakers and planners require information support to manage urban growth and all its consequences, especially on the strategic level. They need an understanding of the potential for a city/region potentials to different urban growth strategies (densification, re-development, greening...) and an evidence base to support decisions. A large amount of data are available, but are scattered, heterogeneous and hard to access – hence are not often used in reality.
Therefore this workpackage, led by GISAT, has developed an “Urban Explorer” application, which integrated multiple sources of data in an easy-to-use manner. It provides a means to ingest, integrate, interrogate and visualise data, whilst also supporting the communication of information to stakeholders and collaboration amongst decision makers. Demonstrated use cases include the identification of Potential Developable Areas (PDAs) and “Urban Dynamics”, i.e. changes in the urban environment identified by satellite. The solution has been built on the WP2 cloud platform, using the Linked Data tools to convert and ingest data. For more details, please see http://www.melodiesproject.eu/content/urban-information-exploration-platform.
The Urban Explorer is being further developed in the context of providing data support to multilateral development banks (e.g. World Bank, ADB, IADB). This provides a path to sustainability for the service.
= WP5: Site-specific information service for land management =
The aim of this workpackage, led by VISTA, was to develop methodologies to assess the ecological and economic suitability of the land surface, to support farmers in their decisions regarding crop placement. Under EU “greening” legislation, farmers must set aside a proportion of their land as “ecological focus areas”. This workpackage consists of two sub-services: one focused on economic evaluation of the land surface, and another focused on ecological evaluation.
The economic evaluation sub-service builds up on the previous European Space Agency project TalkingFields. Within the MELODIES project the existing methods were improved, including the extension from field to farm level, a higher resolution and fully automated processing. The objective was to develop a prototype of a service that can be used to assess the economic potential of the land surface, defined as the crop yield potential for crops in t/ha. This service is now available as a prototype and the pre-processing procedure for the necessary input data for the service is fully automated This also includes a fully automated procedure for the pre-processing of the newly available Sentinel 2 data.
The aim of the project concerning the ecological service was to develop a methodology to derive the ecological value of the land surface and to identify areas within the landscape that are suitable for ecological improvement. A methodology was developed to detect grassland turnover and to derive the cutting frequency and cutting dates of grasslands from optical satellite data. Additionally, a methodology was developed to link the information of freely available information on land cover and conservation areas to the economic evaluation in order to identify potential ecological focus areas within the landscape. This sub-service is also now available as a prototype application.
All of this work fully builds on openly available data. For the calculation of the crop yield potential and the detection of cutting events of grasslands freely available satellite data is used, e.g. of Landsat 7 and 8, Sentinel 2 and Sentinel 1. For the ecological service additional open data is used to detect areas of high ecological value and potential, e.g. information about conservation areas and freely available land cover classifications. These data are being provided by the responsible authorities and/or research institutes and are freely available.
For more information, please see http://www.melodiesproject.eu/content/site-specific-information-service-land-management-1.
= WP6: Ocean Status Assessment Service =
The Marine Strategy Framework Directive (MSFD), adopted in 2008, has the objective of protecting the marine environment across Europe. It sets a target for 2020 of Good Environmental Status (GES) based on 11 GES Descriptors, ranging from Biodiversity to Eutrophication and from contaminants to noise. In this service (led by ACS), a service was developed to contribute to a wider and easier monitoring of “Descriptor 7”, which pertains to hydrographical conditions. These are the physical characteristics of seawater (including salinity, temperature, pH and hydrodynamics). Monitoring of these is key to detecting climatic and long-term cyclical processes but also to evaluate the impacts of anthropogenic activities, including offshore wind farms, tidal barrages, and infrastructures in coastal areas.
This Ocean Status Assessment Service, developed in collaboration with the Ligurian Environment Agency, provides users with information on these descriptors, without requiring the user to deal directly with large volumes of data in complex formats. It uses data from the Copernicus Marine Environment Monitoring Service (CMEMS) and offers a portal for the visualisation of the computed physical variables and a facility for launching on-demand re-analysis algorithms on historical data. It offers two modes: a “basic” mode for regular production and dissemination of a set of selected indicators over a configurable spatial and temporal window, and an “interactive” mode to allow the advanced user to customize the processing chains and tailor the GES Characteristics and indicators. Products can be generated in near real time (within 2 hours of the input data becoming available) and are available through a user-friendly catalogue.
Further information may be found at http://www.melodiesproject.eu/content/assessment-good-environmental-status-oceans-and-seas.
= WP7: Desertification indicators service =
Desertification is a worldwide growing problem that directly affects millions of people. Desertification means the degradation of dry land – i.e. the loss of biological or economic productivity. Dry lands cover approximately 40% of the Earth's land surface. The United Nations Convention to Combat Desertificaion (UNCCD) is the sole legally binding international agreement which links the environment and development to sustainable land management. UNCCD records several environmental and humanitarian variables which are valuable indicators of actions taken or actions needed. The results of this monitoring must be reported annually by each country's representative.
This service responds directly to UNCCD strategic objective 2 (“To improve the condition of affected ecosystems”), by providing information on trends in land cover and productivity. It also responds to the Sustainable Development Goals “Target 15.3: Land Degradation Neutrality by 2030”. It processes numerous sources of open Earth Observation, climate and biophysical data (Sentinel data will be used in future iterations of the service). Three sub-services are offered: (1) Land Condition; (2) Land Use / Land Cover and LULC changes; and (3) Integrated Biophysical and Climatic Susceptibility. Geographically, there is a focus on the Iberian peninsula and Turkey and the service developers (led by Critical Software) collaborated closely with the relevant UNCCD national focal points. Expansion to more geographic areas (Maghred and Mozambique) is planned for future activities.
All data from the service is presented through a user-friendly interactive web application, which also provides the possibility for advanced users to create their own data processing chains and execute them on the cloud computing back-end (see WP2), which enables many EO scenes to be processed in parallel for speed. More information can be found at http://www.melodiesproject.eu/content/desertification-indicators-service.
= WP8: Hazard information services =
The activities in this workpackage are organised in two sub-services, both addressing the area of providing information that is useful in case of a crisis.
Hotspot Mapping Service: When a crisis occurs, access to information is paramount for responders “on the ground”. Open Data creates the possibility to provide relevant information through maps at short notice and at low cost. However, not all agencies have the capability to access the latest digital information, or lack the necessary broadband connections. The Hotspot Mapping Service brings together many sources of data to enable users to very easily generate a map for their location, which is designed for printing and distributing to responders on the ground. It is aimed at three main kinds of customer: (1) Local or regional disaster manager and urban planners supporting planning decisions for disaster risk reduction; (2) NGOs and International Organisations active in humanitarian relief and development; and (3) Global employee safety and risk management solutions for business activities. This service is operational at the technical level and will be operated for at least one year under internal funding while the customer base is grown.
European Historical Flood Database: Flooding is the world’s most expensive natural disaster. The European summer floods of 2013 caused around €15 billion in overall losses. There is a need for the systematic collection of data on hazard events, disasters and their impacts – but so far, there has been no such effort at the European level. The European Historical Flood Database collects information on over 1200 flood events in Europe. The data can be explored on a web interface and, through the use of Linked Data technology, can be combined and cross-referenced with other data sources (e.g. Open Street Map, CORINE Land Cover). This sub-service was co-designed with users in the insurance industry and is being sustained after the MELODIES project through further European funding (H2020 SME instrument).
For more information, please see http://www.melodiesproject.eu/node/37.
= WP9: MetOcean services to the marine transport sector =
Knowledge of present and forecast ocean currents is potentially very important to the marine transport sector, for optimising routes and for predicting the length of a ship’s journey. However there is a current lack of information: observations of ocean currents are scarce and computer simulations typically do not capture every feature of interest. The Automated Identification System (AIS) provides a potential solution to this. Originally designed to help ships avoid collisions, AIS data provides information on the position, speed and heading of all large ships. From these data, this workpackage developed two sub-services:
Ship arrival time service: this uses forecasts of wind, wave, and currents (from computer models provided through Copernicus services) to forecast a ship’s speed between its present position (given by AIS) and the destination harbour, giving an arrival time estimate. This forecast is updated every 2 hours and can be issued more frequently as required (e.g. as a ship approaches port). A full launch of this service with a commercial partner is expected soon, aimed at fleet owners, harbour authorities, terminal owners and logistics service providers.
Ocean current assessment service: The AIS data contain information about the ocean currents, effectively turning the fleet of ships into an array of moving buoys. This can be used to compare with the computer simulations and investigate features such as tidal currents, internal waves and large eddies, yielding improved estimates of the real currents. This sub-service is still under research and development but has many potential applications in areas such as safety at offshore platforms, scientific research and of course predicting ship arrival times (as above). When this service achieves operations, it will produce open data as output.
For more information, please see http://www.melodiesproject.eu/content/marine-transport-services.
= WP10 Groundwater modelling service =
The Water Framework Directive demands public participation, intense monitoring and active management of water bodies. Numerical models (i.e. computer simulations) are the natural tool to integrate diverse data, to assess the natural evolution of water bodies and their sensitivity to human actions, and to communicate the actual status of water bodies. However, models need to be validated and driven by real data.
In this service, initially focused on the Barcelona aquifer, computer simulations are continually improved using observations of water depth provided by citizens (e.g. landowners and farmers with wells or other means to observe groundwater levels). These observations feed into the model to provide improved information of the flow of water in the aquifer, enabling key parameters (such as chloride infiltration) to be estimated. A key innovation of the service is the mechanism for making the best use of citizen data (which may be unreliable in certain aspects). The key to this is to focus on the changes observed by the citizens, not the absolute head level. Other data are also incorporated, including meteorological data and water withdrawal data (from the Catalan Water Agency). The service is provided both through the Web and as a mobile app.
The service is now expanding beyond its initial geographic area, with potential partnerships in Europe and South America. For more information, please see http://www.melodiesproject.eu/content/groundwater-model-service-driven-open-data-and-social-networks-0.
Potential Impact:
We express the potential impact of the project through a number of different categories, from the impact upon specific users and sectors to the wider societal and policy impact. Many of these categories are taken from the requirements of the original call to which the MELODIES consortium responded. From the proposal stage, each workpackage within the project identified named users, who shaped the development of the service and participated in the project’s Advisory Board.
= Impacts upon economic growth =
The participation of SMEs in MELODIES was very strong: nine of the sixteen project partners were SMEs, ranging from small start-ups (e.g. EOXPLORE, Terranea) to more established entities (e.g. Terradue, VISTA, GISAT). We estimate that MELODIES created eight new jobs that will outlast the project itself, primarily within these SME partners. (More jobs were created by the project, but not all of these were sustained.)
Three MELODIES services were directly aimed at commercial customers:
* VISTA built a service to offer advice to farmers on optimising their ecological and economic returns, based on open satellite data. This responds directly to European “greening” legislation (2015). The service quantifies the crop yield potential (economic service) and identifies areas with high potential for ecological development (ecological service) within the landscape. Yield loss can be reduced by 5% by optimal positioning of “greening” areas.
* HERMESS developed technology to better estimate the arrival time of ships into port. This has potential impacts upon both shipping companies and port operators. A terminal owner could avoid losing €18-30k on a single incident if they know a ship will be 6 hours late and can unload another ship in its place.
* EOXPLORE and Terranea collated information on historic European floods, aimed at risk modellers in the insurance industry. More accurate knowledge of natural hazard risks is a potentially high-impact activity (e.g. the European summer floods of 2013 caused around €15 billion in overall losses).
In addition to the impacts of the specific services listed above, we are pleased to note the expansion of some of the MELODIES partners and the formation of significant new initiatives. In particular, GISAT and Hydromodel Host are expanding their activities into new geographic regions, Assimila have employed new staff to take forward the MELODIES research on a commercial basis, and the University of Reading has undertaken a major new initiative in the formation of the Institute for Environmental Analytics (IEA, www.the-iea.org). The IEA is a partnership of UK universities, research institutions (including MELODIES partner CEH) and industry, collaborating together to innovate and develop new applications based upon environmental data. The MELODIES project was instrumental in equipping staff at the University of Reading with the technical skills and contacts to build this new initiative (e.g. the MELODIES coordinator has become the Chief Technology Officer of the IEA). We expect this to be a key driver for innovation and growth in the UK’s environmental sector.
We also anticipate further release of data and development of innovative services in future projects, building upon (or inspired by) the MELODIES services. We hope that the MELODIES project has provided evidence that Open Data can stimulate the development of new services that would not otherwise be viable. MELODIES partners report savings of tens or hundreds of thousands of euros compared with commercial data in the development of their services.
= Impacts upon public policy and services =
Several of the MELODIES services were designed to address the needs of public-sector policymakers and service developers:
* GISAT integrated several sources of urban open data into an “Urban Explorer” platform, designed to improve urban planning and management. This allows the visualisation of multiple data sources (e.g. developable regions, flood risk zones) and the detection of changes, initially targeted at the Prague municipal area but now expanding to new regions.
* The University of Reading and the Centre for Ecology and Hydrology developed a new methodology for monitoring land cover using open Earth Observation data. This supports the UK’s Department for the Environment, Food and Rural Affairs (Defra), who use land cover information to calculate agricultural greenhouse gas emissions as part of the United Nations Framework Convention on Climate Change.
* ACS and INGV developed an Ocean Status Assessment Service, which enables data about the oceans (from the Copernicus Marine Environment Monitoring Service) to be processed to calculate indicators of “Good Environmental Status”. This responds to the needs of the Marine Strategy Framework Directive. This was developed in close collaboration with the Ligural Environment Agency.
* Critical Software, CSIC and IST collaborated to develop a service that calculates several indicators of desertification that are used within the United Nations Framework to Combat Desertification (UNCCD). This information is also useful for biodiversity assessment. In this work, the team collaborated closely with the UNCCD National Focal points for Portugal and Turkey.
* The Water Framework Directive requires public participation and active monitoring of water resources. Hydromodel Host developed a service that combines data from computer simulations of groundwater flow with citizen observations to provide accurate information on aquifers, initially focusing on Barcelona. This has been developed in collaboration with the Catalan Water Agency and the service is being expanded internationally.
Additionally, the technical work within the MELODIES project is expected to influence European policy in cloud computing (e.g. the European Open Science Cloud) and spatial data infrastructures. MELODIES work has been presented to the INSPIRE community and there appears to be great promise in applying the technologies and data standards developed within MELODIES (e.g. CoverageJSON) to data-sharing within Europe.
= Technological impacts =
The cloud computing platform (developed by Terradue) is now being applied in other European projects (notably NextGEOSS and CoReSyf). It demonstrates that cloud computing can reduce the costs of data processing, particularly for organisations that lack the budget or capability to operate their own high-end computing resources. MELODIES partners report savings of “tens to hundreds of €k” compared with in-house resources and a “10-20x reduction in data production time”. Platforms such as these can enable the community as a whole to benefit from economies of scale and the potential to share data more effectively.
By working with standardisation bodies (notably the Open Geospatial Consortium and the World Wide Web Consortium), we believe that the MELODIES project will impact upon future standards for describing and sharing data. One example of this is the CoverageJSON standard for publishing complex data in a web-friendly format, which is in the process of being described as a joint OGC/W3C specification.
= Main dissemination activities =
The MELODIES project has maintained a very active programme of dissemination activities. In total we recorded 160 such activities (full list is elsewhere in this report), ranging from one-on-one meetings to large conferences, and also including articles in the press. These activities addressed a number of different communities:
- Technical community (e.g. Open Geospatial Consortium, data standards working groups)
- Scientific community (e.g. ESA Living Planet Symposium)
- Industry (e.g. Oceanology conference, small workshops)
- Government and policy (e.g. Defra, Department for the Environment and Climate Change – also see above)
- International initiatives (e.g. Plenary meetings of the Group on Earth Observations)
In addition to these meetings, we maintained a website, the most popular aspect of which was the blog. The blog published 24 overview articles on key activities within the project, which accumulated thousands of views in total, with the most popular articles receiving many hundreds of “hits”. We also maintained a Twitter feed (@MelodiesProject).
The culmination of our dissemination activities was our public event (“Exploiting Open Data”), held in Brussels in October 2016. This event was held over three days: on the first day we invited external keynote speakers to give talks on the subject of Open Data; on the second day the MELODIES project participants gave summaries of their results; and the third day was an opportunity for the software developer community to receive some hands-on training on the tools we have developed. The meeting was also attended by several representatives of different programmes at the European Commission. We published a set of fact sheets and posters summarising the project results, which may be accessed online or requested in hardcopy from the coordinator.
= Exploitation of results =
The mechanisms for exploiting the results of the project are described in the section above. In summary, all the services developed within the MELODIES project have a concrete plan for sustainability, by which we mean the existence of a “living outcome” that lasts after the end of the project. Six of the services and sub-services are operational (at some level) at the end of the project, and a further seven have concrete plans for further technical or business development. The underlying cloud computing platform will be sustained through a mixed model of research grant support and commercial subscriptions.
List of Websites:
Web: http://www.melodiesproject.eu
Email: info@melodiesproject.eu or j.d.blower@reading.ac.uk
Twitter: @MelodiesProject
The MELODIES project was conceived as a direct response to the European Open Data Strategy. Its principal objective is to demonstrate that the release of Open Data simulates innovation and growth, by enabling researchers, industry and government to develop new and improved services that benefit society in a variety of different ways. The project has achieved this by developing eight new real-world services, based upon Open Data and aimed at different users within the environmental sector. The services are underpinned by a cloud computing platform, which is designed specifically for the purpose of managing and processing large amounts of environmental data (e.g. from satellites). These services were:
* “Emissions Inventory Service”: produced information on land cover and soil moisture for use in estimating greenhouse gas emissions in the UK.
* “Urban Accounting Service”: combined, visualised and analysed open data about urban environments to support urban planning and management.
* “Land Management Service”: used satellite data to optimise the economic and ecological benefits of farming.
* “Ocean Status Assessment Service”: produced information on “Good Environmental Status” of oceans and seas, for compliance with the Water Framework Directive.
* “Desertification Indicators Service”: used satellite data to calculate indicators of desertification and land condition, to support monitoring required by the United Nations Convention to Combat Desertification.
* “Crisis Mapping Service”: developed a rapid mapping application for disaster response, and a European Historic Floods database for users in the insurance industry.
* “Marine Transport Service”: used data and modelling to improve predictions of when ships will arrive into port, avoiding costly and inefficient delays
* “Groundwater Modelling Service”: use models and citizen observations to understand and manage groundwater bodies
Many of these services have become operational at some level during the project. All the remaining services have concrete plans for further technical and business development, ensuring their continuation beyond the timeframe of the project.
During the course of the project, the Sentinel satellites started delivering open data, and we were able to take advantage of this and become early adopters of this extremely valuable source of data.
A key research theme throughout the project was the use of Linked Data. “Linked Data” describes a set of best practices for describing data and publishing it on the Web, in a manner that permits different data sources to be explicitly linked with each other. In this way, it is possible to break out of “data silos”, in which different datasets are managed and accessed completely independently – these “silos” mean that it can be very difficult to merge data from different sources.
The project was very active in disseminating its results to a wide audience, through over 150 meetings and workshops, the project website, a popular blog and social media. At the end of the project we held a three-day event in Brussels, in which we presented our results to a wide audience through talks, demonstrations, posters and fact sheets.
Project Context and Objectives:
The European Open Data Strategy establishes important new principles that ensure that European public sector data will be released at no cost (or marginal cost), in machine-readable, commonly-understood formats, and with liberal licenses enabling wide reuse. These data encompass both scientific data about the environment (from Earth Observation and other fields) and other public sector information, including diverse topics such as demographics, health and crime. Many open geospatial datasets (e.g. land use) are already available through the INSPIRE directive and made available through infrastructures such as the Global Earth Observation System of Systems (GEOSS). The intention of the Open Data Strategy is to stimulate the growth of research and value-adding services that build upon these data streams; however, the potential value inherent in open data, and the benefits that can be gained by combining previously-disparate sources of information are only just starting to become understood.
The MELODIES project (Maximising the Exploitation of Linked Open Data In Enterprise and Science) is a direct response to this Strategy. Its principal objective is to demonstrate that the release of Open Data simulates innovation through the provision of new services for users in research, government, industry and the general public in a broad range of societal and environmental benefit areas.
Open data are highly diverse and can be very large (scientific datasets can reach the scale of petabytes) and therefore require significant computing capacity and informatics expertise to exploit them to their full potential. This capacity is often beyond the reach of businesses (particularly SMEs), requiring large capital investments as well as highly specialist skills in several aspects of computing. Therefore, another key project objective was to build a computing platform that provides the service developers with the high-level computing and data management capabilities that they required. This allowed them to focus on the design and evolution of their service, enable collaboration on common tasks and contributes towards the long-term sustainability of the services.
Linked Data is an emerging paradigm for data publication that is intended to ensure that diverse datasets are joined up, and that data can be reused by the widest possible audience, beyond the specialist user community. The MELODIES project applied Linked Data methodologies to test their efficacy in developing real-world services, building heavily on previous European investments in the field.
In summary, the main objectives of the project were:
1. To combine multiple open data streams (including Earth Observation and government data) to develop eight new innovative services covering a range of domains including agriculture, urban ecosystems, land use management, marine information, desertification, crisis management and hydrology. These services combine Earth Observation data with other open data sources to produce new information for the benefit of scientists, industry, government decision-makers, public service providers and citizens.
2. To apply Linked Data and Semantic Web approaches to enhance the services and enable multiple sources of data to be discovered, interpreted and cross-referenced, thereby generating new knowledge;
3. To ensure that the services are sustainable beyond the lifetime of the project by considering cost-effectiveness, improving data quality and maximizing their user base, whilst also considering the direction of travel of wider initiatives;
4. To deliver reusable technology, building upon previous e-Infrastructure investments, that facilitates the exploitation of open data by both the MELODIES services and by future services outside of the project;
5. To implement a comprehensive and coherent strategy of user engagement and dissemination ensured that the project develops in response to real user needs;
6. To demonstrate the value of open data and stimulate the release of more open data by data providers.
Project Results:
= Overview =
The overall structure of the project is as follows. The eight services (applications) are developed in workpackages 3 to 10 respectively. Some of the eight services are be further divided into “sub-services”, reflecting distinct streams of activity within their workpackage. Workpackage 2 encompasses the development of the technical platform, which underpins the services. Workpackage 11 includes activities relating to ensuring the sustainability of the services, and workpackage 12 includes dissemination activities; both of these are cross-cutting workpackages in which all partners have participated. Workpackage 1 encompasses project management activities; we do not describe these further here.
The main results of the project are given here, organised by workpackage and focusing on the technical workpackages (2 to 10 inclusive). This information is suitable for wide publication. In some cases, for commercial reasons, not all information about the results can be made public. In these cases, further information can be found in Section B.
The text below contains many links to the MELODIES website, where presentations, videos and other dissemination material can be found. The reader is also referred to the introductory video, poster and factsheets that are attached to this Report.
= WP2: Technical Platform =
The primary purpose of the project’s Technical Platform was to provide underpinning technology to support the development and deployment of the services. However, this workpackage yielded a number of technologies of its own, which will be exploited further in future projects.
Cloud computing platform: The IT infrastructure that forms the basis of the Platform is a cloud computing platform, built and operated by Terradue. One key feature of this is the ability for service developers to start their operations at a small scale (using modest computing resources), and then easily scale as their requirements grow. In return, the service developers in the MELODIES project have provided very valuable requirements (e.g. for access to particular software packages and/or key datasets) that have helped Terradue to improve their platform. As a result, the cloud platform has been significantly improved within MELODIES and is now being made available for much wider use, both within the context of research projects (e.g. NextGEOSS and CoReSyf) and commercial use. The platform will therefore be sustained through a mixed model of R&D and commercial revenue. More information can be found at http://www.melodiesproject.eu/content/melodies-shared-platform.
Linked Data tools: The use of Linked Data techniques was a key technical research element of the MELODIES project. (“Linked Data” describes a set of best-practice approaches for publishing and processing data, using widely-adopted open standards, allowing previously-separate datasets to be brought together more easily.) This approach is not specific to geospatial or scientific data: many kinds of data (e.g. library catalogues, socioeconomic data) are published in this way. Therefore one challenge of the MELODIES project was to develop technology that applies Linked Data techniques to complex and high-volume Earth Observation data. The University of Athens led this part of the project, and have developed a suite of software tools for handling geospatial Linked Data. These range from tools to convert data from existing formats into Linked Data (GeoTriples), through database tools (Strabon and Ontop-spatial) to Linked Data visualisation tools (Sextant). All are available as open source software and can be exploited by anyone. Development of these tools is ongoing through other research projects, including BigDataEurope. More information can be found at http://www.melodiesproject.eu/software-tools.
CoverageJSON data format: During the project we observed that there was a gap in the capabilities available to the developers of websites and apps who wish to consume scientific data. Existing data formats (e.g. NetCDF, HDF, GRIB, XML) were either highly complex or too inefficient to be used in these environments. Therefore we developed a new data format, known as CoverageJSON, which is designed as a format for encoding many kinds of scientific and Earth Observation data in a manner that is friendly for web and app developers. (In web development, the pre-eminent data format is JSON, the JavaScript Object Notation.) The format is based on the concepts and standards from ISO and OGC and its specification is published openly on the web (https://covjson.org). The spec is supported by documentation in the form of a Cookbook, plus a suite of open source tools for producing and consuming the format. Great interest in the format has already been shown by the user community, who are also encouraged to contribute to its development. CoverageJSON has also been discussed (at the recent INSPIRE conference) as a possible future standard for INSPIRE compliance, as it is perceived to be simpler than existing mechanisms. This strand of work is led by the University of Reading.
ncWMS: Another popular method of building data-driven websites and apps is to use server-side software to turn complex data into images, then serve these images over the web (in a similar manner to Google Maps and other mapping applications). NcWMS is a piece of software, developed by the University of Reading, which does this in a manner specified by the OGC’s Web Map Service specification. NcWMS was originally developed before the MELODIES project, but we have made several improvements and released these to the community (ncWMS is fully open source). These improvements have included support for more data types (including “unstructured” grids common in the coastal-ocean community), a greater choice of parameters when generating images, improved documentation and a refactored code base that is more conducive to wide reuse. More details can be found at https://reading-escience-centre.github.io/ncwms/.
= WP3: Improving emissions inventories with Earth Observation data =
The United Nations Framework Convention on Climate Change (UNFCCC) requires governments to report the greenhouse gas (GHG) emissions from their countries. One way of estimating agricultural GHG emissions is to use information about land cover and land use – different types of land use produce different amounts of GHGs per hectare. GHG emissions are also affected by soil moisture, although there is little available information on the real levels of moisture in the soil. This workpackage encompassed two related “sub-services” that both contribute to helping the UK government to improve its GHG emissions. In both, we worked closely with the Department for the Environment, Food and Rural Affairs (Defra) as well as the Department for Energy and Climate Change (DECC).
New Land Cover Map: Currently the UK uses a particular land cover data product from the Centre for Ecology and Hydrology, which is produced once every few years. Defra has a requirement for more frequent (yearly) updates to the UK land cover map, although they would be satisfied with a lower spatial resolution than the current product. In MELODIES we produced a new yearly Land Cover product using only open data (MODIS and ASTER satellite data, plus the latest CEH land cover map) and open source software (Python, GDAL). Based on user requirements, the new product is focused on accurately estimating land cover *changes* from year to year, in order to generate a highly self-consistent product, which avoids some of the inconsistency problems that are associated with other data products. The data product will soon be available as Open Data from the CEH Environmental Information Platform (http://eip.ceh.ac.uk). The code used to generate the product has been published and will be applied in future projects. The UK government has issued an Invitation to Tender for suppliers to produce this data product (or an equivalent product) on an operational basis, so there is a clear opportunity for this research to have a tangible impact upon the government’s operations. For more information, please see http://www.melodiesproject.eu/node/190.
Soil moisture data portal: CEH has “rescued” a very valuable set of soil moisture data from neutron probe measurements from 1966-2013, much of which was stored on fragile magnetic tape. This dataset (356 tubes across 64 sites) has been checked and corrected, and linked with other data on land cover and soil type. These measurements were expensive and time-consuming to collect but are potentially very valuable for validating other results from cheaper sensors or satellites. The data are published on the CEH Environmental Information Platform, and also through a specially-developed “soil moisture explorer” application, which allows the user to visualise the soil moisture data in concert with other data, including the new MELODIES land cover map (see above). For more information, please see http://www.melodiesproject.eu/node/191.
= WP4: Urban ecosystem accounting service =
The rapid growth in urbanisation across the globe presents many challenges. Policymakers and planners require information support to manage urban growth and all its consequences, especially on the strategic level. They need an understanding of the potential for a city/region potentials to different urban growth strategies (densification, re-development, greening...) and an evidence base to support decisions. A large amount of data are available, but are scattered, heterogeneous and hard to access – hence are not often used in reality.
Therefore this workpackage, led by GISAT, has developed an “Urban Explorer” application, which integrated multiple sources of data in an easy-to-use manner. It provides a means to ingest, integrate, interrogate and visualise data, whilst also supporting the communication of information to stakeholders and collaboration amongst decision makers. Demonstrated use cases include the identification of Potential Developable Areas (PDAs) and “Urban Dynamics”, i.e. changes in the urban environment identified by satellite. The solution has been built on the WP2 cloud platform, using the Linked Data tools to convert and ingest data. For more details, please see http://www.melodiesproject.eu/content/urban-information-exploration-platform.
The Urban Explorer is being further developed in the context of providing data support to multilateral development banks (e.g. World Bank, ADB, IADB). This provides a path to sustainability for the service.
= WP5: Site-specific information service for land management =
The aim of this workpackage, led by VISTA, was to develop methodologies to assess the ecological and economic suitability of the land surface, to support farmers in their decisions regarding crop placement. Under EU “greening” legislation, farmers must set aside a proportion of their land as “ecological focus areas”. This workpackage consists of two sub-services: one focused on economic evaluation of the land surface, and another focused on ecological evaluation.
The economic evaluation sub-service builds up on the previous European Space Agency project TalkingFields. Within the MELODIES project the existing methods were improved, including the extension from field to farm level, a higher resolution and fully automated processing. The objective was to develop a prototype of a service that can be used to assess the economic potential of the land surface, defined as the crop yield potential for crops in t/ha. This service is now available as a prototype and the pre-processing procedure for the necessary input data for the service is fully automated This also includes a fully automated procedure for the pre-processing of the newly available Sentinel 2 data.
The aim of the project concerning the ecological service was to develop a methodology to derive the ecological value of the land surface and to identify areas within the landscape that are suitable for ecological improvement. A methodology was developed to detect grassland turnover and to derive the cutting frequency and cutting dates of grasslands from optical satellite data. Additionally, a methodology was developed to link the information of freely available information on land cover and conservation areas to the economic evaluation in order to identify potential ecological focus areas within the landscape. This sub-service is also now available as a prototype application.
All of this work fully builds on openly available data. For the calculation of the crop yield potential and the detection of cutting events of grasslands freely available satellite data is used, e.g. of Landsat 7 and 8, Sentinel 2 and Sentinel 1. For the ecological service additional open data is used to detect areas of high ecological value and potential, e.g. information about conservation areas and freely available land cover classifications. These data are being provided by the responsible authorities and/or research institutes and are freely available.
For more information, please see http://www.melodiesproject.eu/content/site-specific-information-service-land-management-1.
= WP6: Ocean Status Assessment Service =
The Marine Strategy Framework Directive (MSFD), adopted in 2008, has the objective of protecting the marine environment across Europe. It sets a target for 2020 of Good Environmental Status (GES) based on 11 GES Descriptors, ranging from Biodiversity to Eutrophication and from contaminants to noise. In this service (led by ACS), a service was developed to contribute to a wider and easier monitoring of “Descriptor 7”, which pertains to hydrographical conditions. These are the physical characteristics of seawater (including salinity, temperature, pH and hydrodynamics). Monitoring of these is key to detecting climatic and long-term cyclical processes but also to evaluate the impacts of anthropogenic activities, including offshore wind farms, tidal barrages, and infrastructures in coastal areas.
This Ocean Status Assessment Service, developed in collaboration with the Ligurian Environment Agency, provides users with information on these descriptors, without requiring the user to deal directly with large volumes of data in complex formats. It uses data from the Copernicus Marine Environment Monitoring Service (CMEMS) and offers a portal for the visualisation of the computed physical variables and a facility for launching on-demand re-analysis algorithms on historical data. It offers two modes: a “basic” mode for regular production and dissemination of a set of selected indicators over a configurable spatial and temporal window, and an “interactive” mode to allow the advanced user to customize the processing chains and tailor the GES Characteristics and indicators. Products can be generated in near real time (within 2 hours of the input data becoming available) and are available through a user-friendly catalogue.
Further information may be found at http://www.melodiesproject.eu/content/assessment-good-environmental-status-oceans-and-seas.
= WP7: Desertification indicators service =
Desertification is a worldwide growing problem that directly affects millions of people. Desertification means the degradation of dry land – i.e. the loss of biological or economic productivity. Dry lands cover approximately 40% of the Earth's land surface. The United Nations Convention to Combat Desertificaion (UNCCD) is the sole legally binding international agreement which links the environment and development to sustainable land management. UNCCD records several environmental and humanitarian variables which are valuable indicators of actions taken or actions needed. The results of this monitoring must be reported annually by each country's representative.
This service responds directly to UNCCD strategic objective 2 (“To improve the condition of affected ecosystems”), by providing information on trends in land cover and productivity. It also responds to the Sustainable Development Goals “Target 15.3: Land Degradation Neutrality by 2030”. It processes numerous sources of open Earth Observation, climate and biophysical data (Sentinel data will be used in future iterations of the service). Three sub-services are offered: (1) Land Condition; (2) Land Use / Land Cover and LULC changes; and (3) Integrated Biophysical and Climatic Susceptibility. Geographically, there is a focus on the Iberian peninsula and Turkey and the service developers (led by Critical Software) collaborated closely with the relevant UNCCD national focal points. Expansion to more geographic areas (Maghred and Mozambique) is planned for future activities.
All data from the service is presented through a user-friendly interactive web application, which also provides the possibility for advanced users to create their own data processing chains and execute them on the cloud computing back-end (see WP2), which enables many EO scenes to be processed in parallel for speed. More information can be found at http://www.melodiesproject.eu/content/desertification-indicators-service.
= WP8: Hazard information services =
The activities in this workpackage are organised in two sub-services, both addressing the area of providing information that is useful in case of a crisis.
Hotspot Mapping Service: When a crisis occurs, access to information is paramount for responders “on the ground”. Open Data creates the possibility to provide relevant information through maps at short notice and at low cost. However, not all agencies have the capability to access the latest digital information, or lack the necessary broadband connections. The Hotspot Mapping Service brings together many sources of data to enable users to very easily generate a map for their location, which is designed for printing and distributing to responders on the ground. It is aimed at three main kinds of customer: (1) Local or regional disaster manager and urban planners supporting planning decisions for disaster risk reduction; (2) NGOs and International Organisations active in humanitarian relief and development; and (3) Global employee safety and risk management solutions for business activities. This service is operational at the technical level and will be operated for at least one year under internal funding while the customer base is grown.
European Historical Flood Database: Flooding is the world’s most expensive natural disaster. The European summer floods of 2013 caused around €15 billion in overall losses. There is a need for the systematic collection of data on hazard events, disasters and their impacts – but so far, there has been no such effort at the European level. The European Historical Flood Database collects information on over 1200 flood events in Europe. The data can be explored on a web interface and, through the use of Linked Data technology, can be combined and cross-referenced with other data sources (e.g. Open Street Map, CORINE Land Cover). This sub-service was co-designed with users in the insurance industry and is being sustained after the MELODIES project through further European funding (H2020 SME instrument).
For more information, please see http://www.melodiesproject.eu/node/37.
= WP9: MetOcean services to the marine transport sector =
Knowledge of present and forecast ocean currents is potentially very important to the marine transport sector, for optimising routes and for predicting the length of a ship’s journey. However there is a current lack of information: observations of ocean currents are scarce and computer simulations typically do not capture every feature of interest. The Automated Identification System (AIS) provides a potential solution to this. Originally designed to help ships avoid collisions, AIS data provides information on the position, speed and heading of all large ships. From these data, this workpackage developed two sub-services:
Ship arrival time service: this uses forecasts of wind, wave, and currents (from computer models provided through Copernicus services) to forecast a ship’s speed between its present position (given by AIS) and the destination harbour, giving an arrival time estimate. This forecast is updated every 2 hours and can be issued more frequently as required (e.g. as a ship approaches port). A full launch of this service with a commercial partner is expected soon, aimed at fleet owners, harbour authorities, terminal owners and logistics service providers.
Ocean current assessment service: The AIS data contain information about the ocean currents, effectively turning the fleet of ships into an array of moving buoys. This can be used to compare with the computer simulations and investigate features such as tidal currents, internal waves and large eddies, yielding improved estimates of the real currents. This sub-service is still under research and development but has many potential applications in areas such as safety at offshore platforms, scientific research and of course predicting ship arrival times (as above). When this service achieves operations, it will produce open data as output.
For more information, please see http://www.melodiesproject.eu/content/marine-transport-services.
= WP10 Groundwater modelling service =
The Water Framework Directive demands public participation, intense monitoring and active management of water bodies. Numerical models (i.e. computer simulations) are the natural tool to integrate diverse data, to assess the natural evolution of water bodies and their sensitivity to human actions, and to communicate the actual status of water bodies. However, models need to be validated and driven by real data.
In this service, initially focused on the Barcelona aquifer, computer simulations are continually improved using observations of water depth provided by citizens (e.g. landowners and farmers with wells or other means to observe groundwater levels). These observations feed into the model to provide improved information of the flow of water in the aquifer, enabling key parameters (such as chloride infiltration) to be estimated. A key innovation of the service is the mechanism for making the best use of citizen data (which may be unreliable in certain aspects). The key to this is to focus on the changes observed by the citizens, not the absolute head level. Other data are also incorporated, including meteorological data and water withdrawal data (from the Catalan Water Agency). The service is provided both through the Web and as a mobile app.
The service is now expanding beyond its initial geographic area, with potential partnerships in Europe and South America. For more information, please see http://www.melodiesproject.eu/content/groundwater-model-service-driven-open-data-and-social-networks-0.
Potential Impact:
We express the potential impact of the project through a number of different categories, from the impact upon specific users and sectors to the wider societal and policy impact. Many of these categories are taken from the requirements of the original call to which the MELODIES consortium responded. From the proposal stage, each workpackage within the project identified named users, who shaped the development of the service and participated in the project’s Advisory Board.
= Impacts upon economic growth =
The participation of SMEs in MELODIES was very strong: nine of the sixteen project partners were SMEs, ranging from small start-ups (e.g. EOXPLORE, Terranea) to more established entities (e.g. Terradue, VISTA, GISAT). We estimate that MELODIES created eight new jobs that will outlast the project itself, primarily within these SME partners. (More jobs were created by the project, but not all of these were sustained.)
Three MELODIES services were directly aimed at commercial customers:
* VISTA built a service to offer advice to farmers on optimising their ecological and economic returns, based on open satellite data. This responds directly to European “greening” legislation (2015). The service quantifies the crop yield potential (economic service) and identifies areas with high potential for ecological development (ecological service) within the landscape. Yield loss can be reduced by 5% by optimal positioning of “greening” areas.
* HERMESS developed technology to better estimate the arrival time of ships into port. This has potential impacts upon both shipping companies and port operators. A terminal owner could avoid losing €18-30k on a single incident if they know a ship will be 6 hours late and can unload another ship in its place.
* EOXPLORE and Terranea collated information on historic European floods, aimed at risk modellers in the insurance industry. More accurate knowledge of natural hazard risks is a potentially high-impact activity (e.g. the European summer floods of 2013 caused around €15 billion in overall losses).
In addition to the impacts of the specific services listed above, we are pleased to note the expansion of some of the MELODIES partners and the formation of significant new initiatives. In particular, GISAT and Hydromodel Host are expanding their activities into new geographic regions, Assimila have employed new staff to take forward the MELODIES research on a commercial basis, and the University of Reading has undertaken a major new initiative in the formation of the Institute for Environmental Analytics (IEA, www.the-iea.org). The IEA is a partnership of UK universities, research institutions (including MELODIES partner CEH) and industry, collaborating together to innovate and develop new applications based upon environmental data. The MELODIES project was instrumental in equipping staff at the University of Reading with the technical skills and contacts to build this new initiative (e.g. the MELODIES coordinator has become the Chief Technology Officer of the IEA). We expect this to be a key driver for innovation and growth in the UK’s environmental sector.
We also anticipate further release of data and development of innovative services in future projects, building upon (or inspired by) the MELODIES services. We hope that the MELODIES project has provided evidence that Open Data can stimulate the development of new services that would not otherwise be viable. MELODIES partners report savings of tens or hundreds of thousands of euros compared with commercial data in the development of their services.
= Impacts upon public policy and services =
Several of the MELODIES services were designed to address the needs of public-sector policymakers and service developers:
* GISAT integrated several sources of urban open data into an “Urban Explorer” platform, designed to improve urban planning and management. This allows the visualisation of multiple data sources (e.g. developable regions, flood risk zones) and the detection of changes, initially targeted at the Prague municipal area but now expanding to new regions.
* The University of Reading and the Centre for Ecology and Hydrology developed a new methodology for monitoring land cover using open Earth Observation data. This supports the UK’s Department for the Environment, Food and Rural Affairs (Defra), who use land cover information to calculate agricultural greenhouse gas emissions as part of the United Nations Framework Convention on Climate Change.
* ACS and INGV developed an Ocean Status Assessment Service, which enables data about the oceans (from the Copernicus Marine Environment Monitoring Service) to be processed to calculate indicators of “Good Environmental Status”. This responds to the needs of the Marine Strategy Framework Directive. This was developed in close collaboration with the Ligural Environment Agency.
* Critical Software, CSIC and IST collaborated to develop a service that calculates several indicators of desertification that are used within the United Nations Framework to Combat Desertification (UNCCD). This information is also useful for biodiversity assessment. In this work, the team collaborated closely with the UNCCD National Focal points for Portugal and Turkey.
* The Water Framework Directive requires public participation and active monitoring of water resources. Hydromodel Host developed a service that combines data from computer simulations of groundwater flow with citizen observations to provide accurate information on aquifers, initially focusing on Barcelona. This has been developed in collaboration with the Catalan Water Agency and the service is being expanded internationally.
Additionally, the technical work within the MELODIES project is expected to influence European policy in cloud computing (e.g. the European Open Science Cloud) and spatial data infrastructures. MELODIES work has been presented to the INSPIRE community and there appears to be great promise in applying the technologies and data standards developed within MELODIES (e.g. CoverageJSON) to data-sharing within Europe.
= Technological impacts =
The cloud computing platform (developed by Terradue) is now being applied in other European projects (notably NextGEOSS and CoReSyf). It demonstrates that cloud computing can reduce the costs of data processing, particularly for organisations that lack the budget or capability to operate their own high-end computing resources. MELODIES partners report savings of “tens to hundreds of €k” compared with in-house resources and a “10-20x reduction in data production time”. Platforms such as these can enable the community as a whole to benefit from economies of scale and the potential to share data more effectively.
By working with standardisation bodies (notably the Open Geospatial Consortium and the World Wide Web Consortium), we believe that the MELODIES project will impact upon future standards for describing and sharing data. One example of this is the CoverageJSON standard for publishing complex data in a web-friendly format, which is in the process of being described as a joint OGC/W3C specification.
= Main dissemination activities =
The MELODIES project has maintained a very active programme of dissemination activities. In total we recorded 160 such activities (full list is elsewhere in this report), ranging from one-on-one meetings to large conferences, and also including articles in the press. These activities addressed a number of different communities:
- Technical community (e.g. Open Geospatial Consortium, data standards working groups)
- Scientific community (e.g. ESA Living Planet Symposium)
- Industry (e.g. Oceanology conference, small workshops)
- Government and policy (e.g. Defra, Department for the Environment and Climate Change – also see above)
- International initiatives (e.g. Plenary meetings of the Group on Earth Observations)
In addition to these meetings, we maintained a website, the most popular aspect of which was the blog. The blog published 24 overview articles on key activities within the project, which accumulated thousands of views in total, with the most popular articles receiving many hundreds of “hits”. We also maintained a Twitter feed (@MelodiesProject).
The culmination of our dissemination activities was our public event (“Exploiting Open Data”), held in Brussels in October 2016. This event was held over three days: on the first day we invited external keynote speakers to give talks on the subject of Open Data; on the second day the MELODIES project participants gave summaries of their results; and the third day was an opportunity for the software developer community to receive some hands-on training on the tools we have developed. The meeting was also attended by several representatives of different programmes at the European Commission. We published a set of fact sheets and posters summarising the project results, which may be accessed online or requested in hardcopy from the coordinator.
= Exploitation of results =
The mechanisms for exploiting the results of the project are described in the section above. In summary, all the services developed within the MELODIES project have a concrete plan for sustainability, by which we mean the existence of a “living outcome” that lasts after the end of the project. Six of the services and sub-services are operational (at some level) at the end of the project, and a further seven have concrete plans for further technical or business development. The underlying cloud computing platform will be sustained through a mixed model of research grant support and commercial subscriptions.
List of Websites:
Web: http://www.melodiesproject.eu
Email: info@melodiesproject.eu or j.d.blower@reading.ac.uk
Twitter: @MelodiesProject