Citizen Science 2

Crowdsourcing and citizen science are creating an increasingly vast amount of data related to the environment. One of the major emerging challenges in working with such diverse and often highly uncertain datasets is how to best make use of them and provide citizens as well as other end users with a relevant value-added product. Urban air quality is one of the major current and future environmental concerns. Crowdsourcing has significant potential for obtaining detailed spatial and temporal information about air quality at unprecedented detail. However, air quality information from crowdsourcing is subject to substantial uncertainties and data gaps in both space and time. One way to overcome this limitation is to apply data assimilation and data fusion techniques, which allow for combining models with observations in an objective and meaningful way. In this process, value is added to the resulting analysis by filling in spatio-temporal gaps in the observations and at the same time constraining the model to realistic values. In addition, the model can further provide information in areas where no observations are available.

We show that data fusion of air quality observations from crowdsourcing with dispersion model information offers a novel way of generating spatially detailed maps of air quality in the urban environment. The resulting concentration fields have significant potential for providing citizens with personalized information about air quality and exposure. The EU-funded CITI-SENSE project (www.citi-sense.eu) is deploying a dense network of low-cost sensors measuring air quality in eight cities around Europe. These crowdsourced observations are used for mapping urban air quality by fusing them with data obtained from statistical and deterministic air quality models. The data fusion approach used here is based on geostatistics and has been implemented as a fully automated system running in near-real-time. Initial testing of the methodology has been performed in the city of Oslo, Norway. The results indicate that the system is capable of producing detailed urban air quality maps based on adjusting the model output with new information from the crowdsourced observations. The system limits the spatial impact of the observations to their immediate surroundings as specified by the covariance model. Evaluation of the methodology is being carried out using the leave-one-out cross validation technique and simulated datasets. The results obtained with observations simulated from a model-based “true” concentration field indicate that the concentration field provided by the data fusion technique is able to replicate the original “true” concentration field in terms of both spatial patterns and absolute values. The technique accomplishes a global bias correction of the model-based proxy dataset while at the same time performing a local level-shift around observation sites, with the spatial influence of the site characterised by the fitted semivariogram model. While there is a strong dependence of the achievable mapping accuracy on the total number of available observation sites, the mapping accuracy for, e.g., NO₂ was found to reach root mean squared error (RMSE) values of less than 5 µg/m³ when a total number of 50 or more simulated stations were used throughout the mapping domain.

Aside from obvious applications for air quality monitoring and personal exposure assessment, the high-resolution up-to-date air quality maps resulting from the presented methodology have significant potential for upcoming ESA satellite missions related to atmospheric composition. In the future, crowdsourcing of air quality data and related mapping techniques can provide important information on sub-pixel variability of air pollutants measured from space and thus can be used for comprehensive validation studies or downscaling of air quality-related satellite products, such as those acquired by the TROPOMI instrument on the upcoming Sentinel-5P platform, but also later on by the Sentinel-5 and the geostationary Sentinel-4 mission.

The rapid changes in climate over the last decades, together with the explosion of human population, have shaped the context for a fragile biosphere, prone to natural and manmade disasters that result in massive flows of environmental immigrants and great disturbances of ecosystems. Nowadays, the great disasters (e.g. the Indonesian tsunami in 2004, the Fukushima nuclear disaster in 2011, the European Heat Wave in 2003, the Greece's wildfires in 2007, the Earthquake of Nepal) have shown great evidence for high quality Earth Observation (EO) services as it regards disaster and emergency management, and risk reduction (DRR & EMS). The EO community has initiated large scale initiatives in order to: a) generate operational EO services with direct impact in the biosphere and useful to the societies, b) stimulate wider participation of the society including local authorities, volunteers, NGOs, and after all the citizens, enabling the Openness effect and promoting the Open Innovation paradigm, and c) utilize the rapidly growing technologies associated with web, social media, mobile, Crowdsourcing and Participatory Sensing.

The Institute for Astronomy, Astrophysics, Space Applications and Remote Sensing of the National Observatory of Athens (IAASARS/NOA) has developed, in the framework of the BEYOND Centre of Excellence for EO-based monitoring of Natural Disasters (http://www.beyond-eocenter.eu), and operates a rich EO ecosystem of Copernicus compliant services addressing diverse hazardous phenomena caused from climate and weather extremes (fires, floods, windstorms, heat waves), atmospheric disturbances (smoke, dust, ozone, UV), and geo-hazards (earthquakes, landslides, volcanoes).However, there is a communication gap between the BEYOND EO services ecosystem and those either directly concerned by natural disasters, i.e. the citizens or are responsible for managing them. This disruption of information flow – a dissemination break - between interested parties is addressed by DisasterHub.

DisasterHub will fill in this gap by introducing a mobile application that will act as a middleware between a mobile user and the rich suite of the BEYOND EO services, building on the concept of citizen observatories in support of Copernicus, GEO, GEOSS, and UN-SPIDER. In this context the roadmap for generating beneficial EO services through DisasterHub is sketched in two main branches: (i) ingestion, processing and proper fusion of multimodal big EO data (space, in-situ, airborne, and crowd) with additional spatiotemporal evidences (originated from Core Copernicus, GEO, GEOSS) for deriving higher value DRR and EMS decision support information products, and (ii) interlinking the web and mobile platforms for the exchange and ease access of the societies to open EO/crowd generated information products and services.

Through the DisasterHub mobile application the benefited communities will be effectively enlarged. Mutually the BEYOND ecosystem will profit from the large amount of tagged information returned from the field, forming a unique input to the production chains and assimilation of predictive modelling. In summary DisasterHub will showcase in the EO community an upgraded version of the BEYOND ecosystem, the DisasterHub EO services ecosystem which will enable enhanced capabilities such as: software infrastructure for easy access of a mobile user to the EO services and products of BEYOND, tools for practical communication with the EO services in real-time, integration of open geospatial and socioeconomic data via open/linked data ingestion mechanisms (APIs), retrieved from the GEOSS Data-CORE, Copernicus and other EU portals, as well as site specific data sources as cadastral data, asset maps, etc., new validation and fusion techniques relevant to ingesting Crowdsourcing and Participatory Sensing data in EO services.

The objective of the Crowd4Sat 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 ESA products can be used in crowdsourcing. The project addresses concrete scientific and societal problems through four use cases; in particular a snow coverage demonstration project is presented here.

The monitoring of snowmelt 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 and Sentinel-3A (and their brother satellites 1B and 3B) with improved performances and revisit time, it will be possible to monitor snow cover area (SCA) with high accuracy.

However, still exist some problems of measuring SCA using remote sensing techniques, in particular, in mountain terrains, due to atmospheric and relief considerations (SAR suffers strong geometric distortion in mountain, while cloud coverage can affect Optical data).

This demonstration project aims at taking advantage of crowdsourcing activities to validate snow coverage OS products. Through local associations (Catalonia), hikers will participate in providing with in situ snow coverage information. A dedicated mobile application has been developed allowing them to send geolocated reports about the presence/absence of snow at their specific location. In the same time, we provide them with relevant information such as snow coverage over specific hiking routes.

Campaign details will be presented, including a description of the crowdsourcing tool developed and an analysis of the citizens engagement through the activities. Moreover, the validation results of the EO products using citizens observations will be analysed and presented.

Citizen science is an efficient way to collect data about plant phenology, which is one important vegetation functional trait in climate change impact studies, across large areas such as Canada. However, such time series are often fragmental and observations are merely available away from inhabited areas. On the other hand, as it is synoptic and repetitive, remote sensing compensate such drawbacks but gives only one indicator which is the date of the beginning of the growing season, called green-up date, and thus does not inform about inter-species differences in spring phenological response to climate variability. In this study, annual maps of the remote sensing green-up date derived from SPOT-VEGETATION data were compared to the phenological observations collected by the PlantWatch citizen science project across Canada between 1998 and 2012. First, green-up dates were found to relate to the leaf-out dates for four woody species (Populus tremuloides, Acer rubrum, Syringa vulgaris, Larix laricina) for all landcover types except in pixels where agriculture or water bodies were dominant. Second, when spatially aggregated, the remote sensing green-up date matched well the interannual variations in leafing and also in flowering of most of the recorded species. The single date recorded by our remote sensing method is thus clearly an indicator of the whole plant community phenology. Data from this volunteer PlantWatch network proved consistent with independent satellite data, suggesting that combining two will strengthen future capacity to monitor vegetation changes, including interspecies, spatial and interannual variabilities.  

Urbanization induces health and environmental risk-related challenges. One of the key challenges in urbanization refers to the matching between expected population growth and space availability. City planners, encouraged by the development of the Smart City concept, are proposing innovative sustainable, dynamic and participative mitigation strategies. Based on cutting edge information and communications technologies (ICT) including remote sensing, WebGIS tools and participatory mobilization, new solutions are designed for improving the living, working and natural environments of the citizens. However, city authorities always need comprehensive, user-driven and holistic visions of their fast changing urban territory to address the population growth challenge. In Wallonia, there is no reliable and frequently updated Land Cover and Land Use map (LULC). The SmartPop project presented in this paper aims at developing a methodology for updating the 2008 LULC map of Wallonia with very high resolution satellite PLEIADES, LiDAR and various geospatial datasets which are available over the Walloon region. Population density map disaggregating demographic figures on this LULC map by dasymetric methods will feed risk models. Prospective analysis of LULC till 2050 using Cellular Automata model will provide spatially explicit scenarios of LULC and population density which are some important inputs in order to adapt current planning policies to future trends.

This paper describes the participatory spatial planning approach of SmartPop which is seen as a preliminary and on-going process in LULC map updating. The user-needs analysis conducted in 2007, at the creation of the LULC map, is currently updated by contacting all the former map users and identifying some new ones. These users should be involved in the research process in three aspects: (A) identification of issues in the existing dataset. This includes reviewing map legend (i.e. presence of heterogeneous classes, confusion between LU and LC, homogenization with neighboring regions maps), scale/minimum mapping unit, format, quality … (B) Precision of evolving challenges in decision making needs (diffusion format, new purposes, software changes) and (C) validation of the interest of the LULC map for their use. This paper presents the user participatory approach currently conducted in four steps: [1] review of past surveys, literature and key environmental reports; [2] distribution of questionnaires related to the exiting LULC map to the identified former and new users by making use of existing geomatics consultative committee of Wallonia; [3] face to face interviews to better understand their thematic challenges and [4] inclusion of these results in the LULC map methodology and proposition of a user oriented validation procedure (fitness to use).

User participation, from policy makers to citizens, is a mandatory step to ensure smart city initiative success. This participation is also mandatory in the SmartPop project. It helps gathering data, choosing adapted methods, validating and valorizing results and generate craze around the research. Engaging citizens in environmental monitoring should be a key focus in any applied scientific research. The GEO-Wiki (http://www.geo-wiki.org/) platform success proves that participatory planning is possible, especially in LULC mapping projects combining voluntary information and remote sensing data. The operational future of this research project required the involvement of a panel of Walloon decision makers at the first stage of the project with a clear willingness of follow up in the validation stage.