About Us

The Project

The objective of this European Space Agency (ESA) funded project is to investigate the different facets of how crowdsourcing and citizen science impacts upon the validation, use and enhancement of ESA Observations from Satellites (OS) products and services, as well as how the ESA products can be used in crowdsourcing. The project will map the current crowdsourcing communities and initiatives, and survey the technological and community trends, challenges and opportunities with the goal of providing recommendations to ESA on how to maximise the benefits of CS within ESA activities. The project also addresses concrete scientific and societal problems through four use cases demonstration projects, targeted at key scientific and societal issues: pollution in metropolitan areas, land use, water management and snow coverage, and flood management and prevention. The outcome of the project will allow ESA to develop a strategy for exploitation of Crowdsourcing and for evolving Information and Communications Technology (ICT) capabilities in support of data exploitation, support to citizen life and educational activities. It will also allow validation of OS data via comparison and integration with crowd-sourced data. The 14-month project runs from 1 Feb 2015 until 31 March 2016 and is funded by ESA under ITT No. AO/1-8068/14/F/MOS.

The Consortium

The pan-European consortium is led by Dr Suvodeep Mazumdar of The University of Sheffield which has a very strong track record in crowdsourcing and citizen science. We also have two companies specialising in satellite based services (Starlab and e-Geos). In addition, we have two demonstration project partners (Alto Adriatico Water Authority and The Floow) and a dissemination and exploitation partner (AizoOn).

Contact Info

Case Studies

Case Study 1

Snow covered area in the Pyrenees involving participatory crowdsourcing of snow coverage data from hikers and its integration with Sentinel-1 and MODIS satellite data. Lead: Starlab

Case Study 2

Pollution in large metropolitan areas using satellite data and data opportunistically crowdsourced from tens of thousands of car black boxes, OBD devices, white boxes and telematics insurance mobile apps. Lead: The Floow

Case Study 3

Integration of opportunistically crowdsourced data from social media (Twitter, Facebook, etc.) with satellite data from Sentinel-1, Landsat-8 and MODIS for floods. Lead: e-GEOS

Case Study 4

Land use in the Bacchiglione river catchment area (North East Italy) using photo data from citizens, social media and CORINE land cover data from IRS P6 LISS III and RapidEye. Lead: Starlab

Case Study 1

This case study is concerned with the precise identification of snow coverage (SCA) in the Pyrenees. It involves participatory crowdsourcing of snow coverage data from hikers and its integration with Sentinel-1 and Modis. The objective is to take advantage of social crowdsourcing data for validation of SCA maps calculated with OS products (Sentinel-1 and MODIS) to obtain up-to-date SCA diagnosis.

The monitoring of snow melt is a key parameter for the management of water resources and runoff modelling. Remote sensing techniques are very useful in this context and have reached operational maturity. With the launch of new satellites such as ESA’s Sentinel-1A with improved performances and revisit time, it will be possible to monitor snow cover area with high accuracy.

However, one of the intrinsic problems of measuring SCA in mountain terrains with Synthetic Aperture Radar (SAR) satellite image processing is the slant-range distortion effects, like foreshortening, layover and shadowing. Moreover, optical data suffers from cloud coverage especially during snow season in mountains area. The idea of this demonstration is to work around these effects with crowdsourcing tools to better improve the accuracy and precision of the SCA measurements.

The study is taking place in the Catalan Pyrenees with local hiker(s) association(s) acting as the crowdsourcing participant. This area exhibits snow coverage which changes along the seasons.

dp1_map

Case Study 2

This case study deals with pollution in large metropolitan areas using OS and data opportunistically crowdsourced from tens of thousands of car black boxes, OBD devices, white boxes and telematics insurance mobile apps. This demonstration project is led by The Floow Ltd and uses its in-house data collected as part of its provision of telematics insurance technologies for the motor insurance market.

Sheffield is the fourth largest city in the UK with an urban population in excess of 550,000 people and over 1.5M people in the wider metropolitan area. The city region encompasses a large variety of environments:

  • Dense manufacturing centres
  • Business and development parks
  • Industrial regions
  • Residential areas
  • Rural areas
  • National Parkland
  • Commercial and retail districts

These areas include roads ranging from remote rural unclassified roads to the west to the M1 motorway in close connection to the east of the city. Like most major cities it operates a ‘hub and spoke’ road design with an inner ring road and a partial outer ring road. Sheffield town centre sits within a series of hills, which concentrate traffic, mobility, development and pollution focusing towards the busy heart of the city.

It is estimated that poor air quality in Sheffield adversely affects human health and is estimated to account for approximately 500 premature deaths each year. This also impacts health costs with around £160 million per year and an estimated reduction of life expectancy of as much as nine years in those affected. For these reasons to address these unbalanced societal and health problems (disproportionately affecting the poor, elderly and vulnerable) the city took an unprecedented step of designating the entire city as an Air Quality Management Area in 2003.

The air quality management area is backed up with an Air Quality Action Plan seeking to address exposure levels especially those frequently beyond annual limits of 40µg.m-3 for NO2 and the daily limit of 50µg.m-3 for PM10 (which is not to be exceeded more than 35 times a year). These EU regulation figures continue to be breached.

Although of interest to explore all of Sheffield this investigation has focused attention upon particular exemplar cases and problem areas within the Sheffield region. Each of these regions has on-going issues with air pollution. These regions were chosen in conjunction with the Local Authority stakeholders following the initial set of stakeholder interviews, to provide maximum value from the investigation by informing at the same time operations of Sheffield County Council.

dooley

Derek Dooley Way

park_square

Park Square Roundabout

uni

University roundabout

This demonstration project will show how opportunistically crowdsourced data can be used to improve the accuracy of ESA air pollution products. Satellite data can inform to a degree air quality (in a vertical column) but crowdsourced vehicular movement and acceleration data indicate vehicular emissions and impact at the ground level. Currently it is not possible to use satellite data to inform local government ground readings or models for pollution (which are required to be collated by law across Europe). Using crowdsourced vehicle trajectories and ground sensor data with satellite data will give an improved ability to enrich data to provide improved predicted values at the ground level. This allows estimation and understanding of ground based pollution which is essential to on the ground operations due to extreme costs to install and maintain ground based sensor networks across the environment, a satellite and crowdsourced approach is essential to overcome current limitations and provide a complete model which thus far has not been possible.

Case Study 3

This case study, led by e-GEOS, concerns the integration of opportunistically crowdsourced data from social media (Twitter, Youtube, Instagram, etc.) with OS data from Sentinel-1, Landsat-8 and MODIS in order to improve the overall quality and timeliness of pure satellite based flood mapping services. The Demonstration Project will tentatively run over a future and real flood event. If no flood occurs before Summer 2015, a past flood event will be selected. The selection will be made to ensure the availability of OS and CS data.

Floods are the most recurrent natural disaster and every year they threaten both urban settlements and agricultural areas causing significant damages and losses both in terms of property, arable land and human lives. In case of such events, the rapid and timely identification of the most heavily affected areas is a critical piece of information for organizations involved in emergency response operations (e.g. civil protection agencies, red cross, etc) in order to plan their in field intervention and direct the available resources where it is most needed.

The use of satellite observations has been increasingly adopted by civil protection organizations in order to rapidly gather a synoptic picture over a large area during the on-going event, since generally most affected areas are difficult to reach and in in-field surveys result as high time and resources consuming activities.

dp3

However, the use of satellite data in case of natural disasters, and specifically flood mapping, is still limited by the delay between the request of a new satellite image collection and its acquisition and delivery for further processing and analysis. Typically such delay is around 24 hours, therefore not always compatible with the immediate information needs of civil protection operators. On the other hand, CS data can be extremely relevant to bridge this gap of information during the first 24 hours and between two consecutive satellite observations, in a complementary perspective.

The scope of this case study is to prove that the combined use of satellite images (including OS data such as ESA Sentinels, MODIS and Landsat-8) with other OS and CS data can improve the overall quality and timeliness of pure satellite based flood mapping services, thus increasing the overall perceived usefulness of the service by both its historic (civil protection authorities) and new potential users (e.g. the single citizen).

The case study will deliver highly accurate post event digital maps showing the observed flood extent monitored from satellites in combination with continuous 'in-field' information gathered through social media.

Case Study 4

This case study, lead by Starlab and the University of Sheffield in collaboration with the Alto Adriatico Water Authority, is centred on land use in the Bacchiglione river catchment area of Italy, using data from citizens, social media and CORINE land cover data from IRS P6 LISS III and RapidEye. The main objective of this demo is to work with crowdsourced data to validate land cover information.

The high plain area of the Bacchiglione river network draining to the cities of Vicenza and Padua supports industrial, commercial and agricultural activities and is intensively used for settlement, production systems and infrastructure. The increasing demand for living space per person and the link between economic activity, increased mobility and growth of transport infrastructure usually result in land uptake.

The definition of land use is a key parameter in the management of water resources and the wider environment. Land cover and land cover change is an important element to know if our environment and natural heritage are to be properly managed and one of the key tools needed by decision-makers in the implementation of Water Management plans (2000/60/EC) and Flood Risk Management plans (2007/60/EC) which indicate the measures necessary to protect quality and quantity of surface waters and groundwater and to improve the definition of a proper water policy and the measures aimed for prevention, protection, preparedness, including flood risk maps.

dp4

Vicenza area - CORINE Land Cover geographical coverage

The main objective of this case study is to work with crowdsourced data to improve the accuracy and the frequency of the land cover information. The aim of crowdsourcing is to provide up-to-date information that will help to monitor and track land cover changes, in accordance with the speed of environmental information delivery needs to keep pace with decision-makers.

Contact

We welcome your input and questions. Use the form below to leave a message.