The Ice, Cloud, and land Elevation Satellite-2, or ICESat-2, mission was launched on September 15, 2018. The satellite carries one instrument, ATLAS, the Advanced Topographic Laser Altimeter System, which sends 10,000 laser pulses per second towards Earth and records individual photons reflected back to its telescope. The volume of data produced by the instrument, nearly a TB of data every day, is unprecedented for public Earth remote sensing data. This presents a challenges for both the data center managing the data, as well as the user wishing to exploring and access the data. NASA’s Earth Observing System Data and Information System (EOSDIS), including the National Snow and Ice Data Center (NSIDC) - the data center responsible for distributing the data from the ICESat-2 mission, provides a suite of services to users interested in data from ICESat-2. These services include spatial, temporal and parameter subsetting, and the option to subscribe to data feeds. In addition to these services, OpenAltimetry, a collaborative project between NSIDC and the San Diego Supercomputer Center at the University of California, San Diego, has created a platform for discovery, access, and visualization of data from both ICESat-2 and its predecessor, ICESat, with a focus on ease-of-use and quick response times. OpenAltimetry enables both novice and expert ICESat and ICESat-2 users across different science specializations to explore and download data via an intuitive web-based interactive interface. Users are able view elevation profiles in their region of interest, inspect metadata on a per shot (ICESat) or per segment (ICESat-2) basis, view waveforms (ICESat) and photon clouds (ICESat-2) and download data in the current view. The OpenAltimetry application is highly scalable and employs a service-oriented architecture with optimized tiered databases and object storage. This presentation will highlight the features of OpenAltimetry, with an emphasis on the ability to compare ICESat and ICESat-2 data over selected regions.

Area D - Board 378

Asbestos is carcinogenic to humans and exposure to this substance can cause a wide range of diseases. In Poland 1997, a statutory ban was introduced on the production, use and marketing of products containing asbestos. The National Programme for Asbestos Abatement for 2009-2032 includes scheduled activities considering asbestos exposure assessment and health protection. As there are several data sources for asbestos exposure in Poland, which are not linked, the aim of this study was to gather and order them developing a PostgreSQL database, an open-source, object-relational system. The data gathered combines the following information: the quantity of asbestos-cement products in use, details of asbestos manufacturing plants, the results of the measurements of asbestos fibre concentrations in the air and cases of asbestos-related diseases. The relational database was then used to develop a spatial analysis of asbestos monitoring and exposure in Poland to demonstrate the current state of realisation of the National Asbestos Abatement Programme in the country for 2009–2032 with the use of geoinformation techniques. The use of a database on health aspects of occupational and environmental asbestos exposure was also proposed in “Asbestos, Asbestosis, and Cancer: Helsinki Criteria update 2014”.

Area D - Board 379

Landscape quality assessment is an increasingly important part of environmental assessment and monitoring. Unlike widely recognized values and thresholds for air, water and land quality, assessable with remote sensing, visual landscape values are still hardly quantifiable since they originate in objective conditions of Earth environment on the one hand and human perception and appreciation on the other hand. Whereas authors already proposed several measurable visual GIS-based indicators of landscape preferences linking psychological and biophysical concepts, the part of studies utilizing remote sensing data in combination with volunteered geographic information for visual landscape quality assessment remains extremely low.
We started with mapping of the colour harmony of land cover as a subcategory of landscape harmony, serving as the one of key concepts in the frameworks of visual landscape quality assessment and it is widely studied in the psychology of perception. Principles of high colour harmony are homogeneous enough to be utilized with remote sensing data to indicate the extent of the colour harmony of land cover and could be borrowed from psychological literature. We hypothesized that the Landsat-derived map of colour harmony extent corresponds to both specific environmental settings within the study area and crowdsourced imagery collected from Mapillary service.
The colour harmony of land cover was examined within a study area in Eastern Estonia. We utilised a cloud-free part of the Landsat-8 scene dated 17-06-2017 with original spatial resolution. The RGB composite of Landsat 8 OLI image satellite bands corresponds to visible light wavelengths (0452–0673 µm), being applicable for studies aiming to examine the visually perceived Earth environment. The satellite image was radiometrically and atmospherically corrected. The red, green and blue bands of the image were transformed into HSV colour space to fulfil the requirements of colour harmony assessment. Hue corresponds to the colour itself, for instance, red, green, or blue. Saturation (or Chroma, depending on colour space) refers to the amount of grey, and Value (or Lightness, Brightness, depending on colour space) is usually associated with the amount of white in the colour. The CORINE 2012 land cover data and the EU-DEM 1.1 digital elevation model were obtained through the Copernicus Land Monitoring Service. Crowdsourced photographs were obtained via the Mapillary API and viewshed analysis was done for the sample of Mapillary imagery to derive the colour harmony score per each viewshed and link satellite-derived colour harmony scores to the in-situ data.
As the colour harmony principles are the most shared for two-colour combinations, we applied those principles to satellite image with Haralick’s textural metrics based on Grey-Level Co-Occurrence Matrix. As a result, maps of colour harmony extent after different authors and summarizing map of Colour Harmony Index were designed. To identify the drivers of the high colour harmony of land cover, the following geographic variables were used in regression modelling as predictors: Albedo index, cost distance from the road network (Terrain Ruggedness Index as a cost surface), SAGA Wetness index and brightness temperature as a rough approximation of the temperature of land cover.
We found that the extent of the colour harmony of land cover varies within CORINE land cover classes, indicating the naturalness of land cover as a positive factor of high colour harmony: the highest scores are for water bodies, wetlands and forests, while the lowest ones are for different kinds of urban fabric. Increasing distance from the road infrastructure supports higher colour harmony, indicating anthropic modification of Earth environment (human disturbance) as a negative factor. Topographic SAGA Wetness index has a positive relationship with colour harmony score, while albedo has a negative association with colour harmony. The relationship between brightness temperature and colour harmony index is U-shaped, demonstrating a non-linear character. The least brightness temperature values (approx. 21–23 °C) are typical for crop fields, and the maximum values are for settlements and bogs. Altogether, the mentioned factors explain up to 54% of the colour harmony variability, but some factors still remain unclear. Comparison of Colour Harmony Index with the sample of the Mapillary content supports the idea that the presence of signs of human disturbance decreases colour harmony of landscape, while weather and lighting conditions, as well as coverage pattern, viewing angles of Mapillary imagery are the limiting factors, thought-provoking to use Flickr and VK.com geotagged content instead.
To sum up, there is a regular relationship between the colour harmony ratings and environmental geographic attributes revealing the objective drivers of colour harmony: regimes of moisturisation and heat supply as well as cultural modification. Volunteered geographic information originating from different sources supports the application of satellite-based Earth observation data to the visual landscape quality mapping, serving for verification purposes. What is more, a prominent direction of the further work is a step towards the linking visitation rate and aesthetic response of landscape observers to the objective estimations of land cover colour harmony from space.

Area D - Board 380

There is a global trend towards urbanization, and the data regarding urban areas is big. However, the available big urban data is not fully used while making decisions regarding cities. As a result, city management policies cannot be optimally effective. Moreover, currently, the scientific knowledge is not effectively communicated to the public. This results in a lack of engagement of citizens in the decision-making processes. We developed a geoportal to address these issues. For development we used free and open source software and open data to facilitate adoption throughout the world. Furthermore, we used Web technologies for the development so that it can run on both mobile and desktop devices regardless of the operating system to improve accessibility. The development is supported by URBAN GEOmatics for Bulk Information Generation, Data Assessment and Technology Awareness (URBAN GEO BIG DATA), a project of national interest (PRIN) funded by the Italian Ministry of Education, University and Research (MIUR)—id. 20159CNLW8.

The geoportal visualizes and processes big, open, multidimensional geospatial data. It encompasses various use cases related to urban areas. The visualization is performed by means of a 3D Web GIS, using mainly NASA WebWorld Wind and CesiumJS APIs. Although NASA Web WorldWind is found to have better performance, for the use cases intended, whenever it did not have the required functionalities CesiumJS was chosen. However, NASA WorldWind has proved to be very effective for the visualization of geographic data in the past (Pirotti et al., 2017; Brovelli et al., 2017). Through effective visualization, the geoportal aims to communicate massive data in a clear way. Besides visualization, data used in some of the use cases can be queried and processed.

The first use case is OpenStreetMap buildings visualization in three dimensions for the cities that are focused on in URBAN GEO BIG DATA project, which are Milan, Turin, Padua, Naples, and Rome. This has been achieved using NASA Web WorldWind and 3dosm (https://github.com/kilsedar/3dosm) plugin API that we developed. The second use case is to visualize CityGML data on a virtual globe developed with CesiumJS using 3DCityDB-Web-Map-Client. The CityGML data in this use case can be queried. The data is generated by one of the institutions in the project, University of Padua (Fissore and Pirotti, 2018).

The third use case is deformation visualization and query. The deformation maps are produced by National Research Council of Italy (Consiglio Nazionale delle Ricerche, CNR)-Institute for Electromagnetic Sensing of the Environment (Istituto per il Rilevamento Elettromagnetico dell’Ambiente, IREA) (Sansosti et al., 2002; Sansosti et al., 2004). The data is stored on GeoServer and used as a Web Map Service (WMS) image layer on a virtual globe developed with CesiumJS. Each target on the ground can be queried to display the deformation time series plot. Within this use case, ImageMosaic Web Map Tile Service (WMTS) of GeoServer is used to animate the land deformation of 16 years (1992-2010).

The last use case is on land use and land cover (LULC) and soil consumption maps. We developed an application called Land Cover Collector (https://github.com/kilsedar/land-cover-collector) to collect land cover data using the nomenclature of GlobeLand30. The data collected through this application can be classified as volunteered geographic information (VGI). The VGI land cover data is overlayed on GlobeLand30 raster imagery for visual inspection. The GlobeLand30 raster imagery visualization on the virtual globe developed with CesiumJS is achieved using WMTS available in GeoServer. Moreover, within the same use case, the land cover and soil consumption imagery given by the Italian Institute for Environmental Protection and Research (Istituto Superiore per la Protezione e la Ricerca Ambientale, ISPRA) is visualized. As these data are available for various times, a time slider is used to visually identify the change. Big raster data with time dimension is stored on GeoServer as ImageMosaic WMTS. Finally, using Web Coverage Processing Service (WCPS) of rasdaman (http://rasdaman.org/), processing big multidimensional LULC and soil consumption raster data is integrated to the geoportal. The processing enables to determine the change of a LULC class or soil consumption between two different years for a bounding box.

The uses cases approach various challenges for monitoring cities, which are: 3D city visualization and query, observing land deformation through time, enabling the comprehension of change in LULC and soil consumption through time via visualization and processing. As the nature of the data used is big and multidimensional, in all the use cases we faced technical difficulties. The choices we made for solving them are summarized in this abstract.


REFERENCES
Fissore, F., and Pirotti, F. (2018). Migration of digital cartography to CityGML; A Web-based tool for supporting simple ETL procedures. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLII-4, 193-200.

Berardino P., Fornaro G., Lanari R., Sansosti E. (2002) A new algorithm for surface deformation monitoring based on small baseline differential SAR interferograms. IEEE Transactions on Geoscience and Remote Sensing, 40(11), 2375-2383.

Lanari R., Mora O., Manunta M., Mallorqui J. J., Berardino P., Sansosti E. (2004). A small-baseline approach for investigating deformations on full-resolution differential SAR interferograms. IEEE Transactions on Geoscience and Remote Sensing, 42(7), 1377-1386.

Pirotti F., Brovelli M. A., Prestifilippo G., Zamboni G., Kilsedar C. E., Piragnolo M., Hogan P. (2017). An open source virtual globe rendering engine for 3D applications: NASA WorldWind. Open Geospatial Data, Software and Standards, 2(4). doi:10.1186/s40965-017-0016-5

Brovelli M. A., Kilsedar C. E., Hogan P., Prestifilippo G., Zamboni G. (2017). NASA
WorldWind: virtual globe for an open smart city. PeerJ Preprints. doi:10.7287/peerj.preprints.2139v5

Area D - Board 381

The Horizon 2020 project, LandSense, is building a modern citizen observatory for Land Use & Land Cover (LULC) monitoring, by connecting citizens with Earth Observation (EO) data to transform current approaches to environmental decision making. Citizen Observatories are community-driven mechanisms to complement existing environmental monitoring systems and can be fostered through EO-based mobile and web applications, allowing citizens to not only play a key role in LULC monitoring, but also to be directly involved in the co-creation of such solutions. A critical component within the project is the LandSense Engagement Platform, a service platform comprised of highly marketable EO-based solutions that contribute to the transfer, assessment, valuation, uptake and exploitation of LULC data and related results. The platform engages citizens to view, analyze and share data collected from different citizen science campaigns and create their own maps, individually and collaboratively. In addition, citizens can participate in ongoing demonstration pilots using their own devices (e.g. mobile phones and tablets), through interactive reporting and gaming applications, as well as launching their own campaigns. This interaction is achieved by bringing together and extending various key pieces of technology like Geo-Wiki, LACO-Wiki, Geopedia, SentinelHub and the Earth Observation Data Centre. Furthermore, a key pillar of the platform is the LandSense Federation which supports users to authenticate from a variety of login providers using social media (i.e. Facebook and Google) and some 2500 academic institutions globally (eduGAIN). Such a federated approach will promote the awareness, outreach, uptake and ultimately the science of citizen science. Services and solutions from the LandSense Engagement Platform are currrently deployed through a series of citizen science campaigns in Vienna, Toulouse, Amsterdam, Serbia, and Spain covering topics such as urban greenspaces, agricultural management and bird habitat/biodiversity monitoring. The presentation will not only showcase the results from these campaigns, but also highlight how one can link to the platform to exploit its EO services and launch your own citizen science campaigns. Such initiatives present clear opportunities to integrate citizen-driven observations with established authoritative data sources to further extend GEOSS and Copernicus capacities, and support comprehensive environmental monitoring systems. In addition, these applications have considerable potential in lowering expenditure costs on in-situ data collection and current calibration/validation approaches within the processing chain of environmental monitoring activities both within and beyond Europe.

Area D - Board 382

The full potential of deriving information for phenological monitoring based on EO data has not been used so far. Reasons are lacking knowledge of remote sensing data and techniques of users with no remote sensing background, lacking knowledge of data portals and download opportunities for big datasats, lacking knowledge of data preparation, data fusion and processing and limitation of the storage volumes. A data cube based on the open source toolkit of the CEOS Open Data Cube initiative (ODC) has been designed to foster and support the integration of multisource (Sentinel-2, MODIS, Sentinel-1) time series and historical data of the Landsat archive in local scale phenological monitoring activities. Analyses-ready data (ARD) are integrated in a user-friendly web-based environment and linked with easy to use analyses tools that can be applied by non-remote sensing analysts. Most important is the fact that users are able to analyse EO data with the needed flexibility regarding the study area, analyses tool and observation time. PhenoCube provides simple time series tools based on NDVI and Leaf Area Index products for two different environmental monitoring scenarios. The first scenario covers the Durable Environmental Multidisciplinary Monitoring Information Network (DEMMIN), which is part of the GEOGLAM / JECAM calibration and validation activities. The data cube provides tools for agricultural monitoring purposes. The second area covers the forest ecosystems of the Steigerwald, Germany. The Phenocube is embedded in a network of local scientists with the goal to test the applicability and interoperability in their operational working environment. We show the first results of th Summer 2018 drought monitoring in Germany.

Area D - Board 383

INTRO.

The goal of E2mC H2020 project is to demonstrate feasibility and usefulness of integration of social media analysis and crowdsourced information within both Rapid Mapping and Early Warning Components of Copernicus Emergency Management Service (EMS). Recently, several operational experiences (large earthquakes such as the one in Central Italy 2016/2017 or large hurricanes and floods such as Harvey Texas 2017) - have shown the high potential contribution of social media and crowdsourcing in improving the overall quality and timeliness of satellite-based Rapid Mapping services. E2mC project has accepted these challenges and designed an innovative approach.
E2mC has succeeded in implementing a prototype platform (“Social&Crowd” platform) implementing the necessary modules to demonstrate under pre-operational conditions the added value of social media and crowdsourcing in Rapid Mapping context. In particular, “Social&Crowd” platform specific characteristics are: a) multi-source social media and news crawling engine, b) customized geocoding engine based on semantic analysis coupled with open source gazetteers, c) deep learning engine to automatically tag media contents and filter out irrelevant contents, d) multi-purpose crowdsourcing platform managing simple micro-tasks to be assigned to the crowd such as keywords translation, media relevance assessment, content geolocation improvement, simple mapping tasks, etc. e) web interface to interact with the platform, trigger ad hoc activations of the “Social&Crowd” platform, inspect and download the results, further integrate them into other generic GIS environments. In this way, the “Social&Crowd” platform demonstrates how crowdsourcing, data mining and Artificial Intelligence can be combined to deliver higher quality data driven services (e.g. crowdsourcing data are used for feeding AI algorithm for image recognition, while AI is used for removing duplicated images automatically or for detecting false positive from images coming from previous disasters).
E2mC has also made significant progresses in the crowdsourcing component and it is now actively managing a hybrid crowdsourcing community composed by heterogeneous groups such as general purpose (e.g. BOINC, through CERN) and emergency specific ones (e.g. HOT, SBTF). E2mC project has started a process to establish links with other relevant crowdsourcing initiatives active in the emergency response domain to join forces and efforts in providing effective and timely answers to Copernicus EMS Rapid Mapping needs.
This paper presents the technological achievements of E2mC as well as the results of the testing and demonstration of the “Social&Crowd” platform during both past events (cold cases) as well as during real and time critical Copernicus EMS Rapid Mapping activations. The results of the demonstrations have been used for qualitative and quantitative assessment of benefits and added value brought by E2mC to satellite-based mapping activities, alone or in combination with complementary data analysis techniques such as, for example, hydraulic modelling of large floods where the data generated by the Social&Crowd platform are integrated as ground truth for model calibration in time critical operational conditions.

METHODS.

Geolocating social media posts for emergency mapping.
A geolocation framework focusing on image extraction, called CIME, was developed with the goal of improving the geolocation of posts consistent with the mapping purposes of E2mC. CIME uses both the local context of tweets (text and associated metadata) and their global context (relations to other tweets, common hashtags, retweets, mentions, etc.) to geolocate the media (images and videos) extracted from tweets. The algorithm has been extended to other social media, Flickr and Youtube, and can use tweets linking these sources to retrieve additional media content, as well as improve the geolocation of Twitter media based on geolocation information available from these additional sources.
Based on an example of the 2014 UK floods, the platform showed information associated with the tweet (text and media) with a link to the original post, and information derived from its analysis, such as validation information from the crowd and tags obtained by analysing the image with deep learning techniques.

Identifying the Footprint of Natural Disasters.
Georeferenced social media posts do not only contain information about a specific event such as a natural disaster, but also information about other topics that are not necessarily related to the event. Generally relevant social media posts are detected through algorithms that filter posts according to predefined keywords. However, this approach is limited to the identification of posts that contain the predefined keywords, which may often change due to the dynamic nature of social media networks (e.g. ad-hoc hashtags), and it requires time for preparation and maintenance of keyword sets for multiple languages.
E2mC develops a workflow based on a topic modelling algorithm to detect latent semantic topics in the social media corpus that needs little to no a-priori information about the underlying structure of the corpus. The results of multiple use cases, including the Napa Valley Earthquake 2014 or Hurricane Harvey 2017, show that distinct disaster-relevant topics can be identified. These topics are then used for semantically classifying the social media posts. The accuracy of the topic model is assessed using pre-labelled social media posts and is used to determine the parameter values for optimum performance of the algorithm. In a next step, the disaster-relevant tweets are clustered in a spatial hot spot analysis to investigate if a local neighbourhood is affected by the disaster with a defined confidence interval. This allows authorities to identify areas that have been affected by a disaster in a short period of time after the natural disaster occurred. Apart from this output, the results can be used for validating/complementing results of other data sources like remote sensing-based information products.

Architecture.
The general architecture of E2mC comprises components for data acquisition, storage, management and analysis and graphical user interfaces for the EMS operators and the crowd. The workflow is triggered either by the Early Warning System (EWS) component in case of a flood or manually by an EMS operator. Various parameters have to be selected for the different analysis components and the crawlers before start. In case of a flood the EWS component provides information including probability, timing, location and magnitude of an event. EMS operator chooses parameters for Witness workflow and can use the parameters obtained by EWS. When the EMS operator starts the Witness workflow, crawlers begin to collect social media posts. The collected data are stored via a centric RESTful API with six resources e.g., user, event, keyword, post, image and tag in a relational database model.

Improving Rapid Mapping with Social Network Information.
Associating a location to images and video can bring significant benefits to rapid mapping but this association is often challenging as most social media items are not geotagged and the location associated to a content does not necessarily coincide with the location of the post. This means that a possible solution will be to infer the location of an image based on the location associated with the related post, as originally geotagged content is too limited for rapid mapping purposes and many messages in social media contain implicit references to names of places.
Using the August 2016 earthquake in Italy as a case study, less than 4% of tweets had locations that could be disambiguated by building a local context – the algorithm therefore needs to be extended to connect tweets based on implicit interactions among messages. For example, tweets that share a similar content or belong to the same conversation should be analyzed to improve geolocation.
A global context for each post is built based on its “neighbours” and their respective locations, which allows to overcome the limits of the local context and to improve the disambiguation of the location – in the case study of the 2016 Italy earthquake, this was observed in more than 20% of the tweets.
Crowdsourcing can also be used to validate social information gathered from automatic analysis and eliminate false positives. This can be done through volunteers who answer whether a specific information is relevant or not. Volunteers can also contribute to reduce uncertainty regarding the localization of an image and find a more accurate location.

RESULTS AND OUTLOOK.

So far, the E2mC project has conducted experimentations based on two cold cases (Floods in UK in 2014 and Hurricane Harvey in the USA in 2017) adn an increasing number of warm cases.
Results have shown the high potential of contribution of social media and crowdsourcing in the improvement of satellite based results (in particular, related to flood events) mainly related to the following areas:
- Smart extraction of relevant contents from the social media data stream to populate the Social Crisis Map, as a “living” online map product (e.g. layered on the Copernicus EMS Reference Map) available starting from the first hours after the event, in order to fill the information shortage before the acquisition of the post event satellite image (e.g. define a new EMS Rapid Mapping product: the Social Crisis Map);
- use of social media data stream as a support to improve satellite image interpretation and product validation.
The following aspects have been identified as main points for development during the remaining duration of the project:
- improvement of the geolocation accuracy (using available public datasets as a reference),
- inclusion of additional social media sources.

Area D - Board 384

Extensive pasturelands in sub-Saharan African drylands are tightly coupled socio-ecological systems, where semi-nomadic and nomadic pastoralists’ land management and herd mobility practices reflect the strong spatial and temporal heterogeneity of pasture resource availability, accessibility (e.g. land tenure, cultural barriers), and usability (e.g. water availability, forage palatability, topography). This poses a challenge for the characterization of these systems using Earth Observation data, as land cover and land use patterns are fundamentally decoupled.
In this study we investigate the possibility of combining mobile-crowdsourced observations on rangeland resources and grazing practices with satellite imagery to integrate local ecological knowledge into land cover mapping approaches and characterize the land use in semi-arid extensive pastoral systems of Kenya.
We used a dedicated task-based mobile application and task management system to incentivize 113 pastoralists from different local communities to provide information on rangeland composition, water distance and livestock, including pictures and GPS location, during a 6 months period (March - August 2015) covering the long-rains and long-dry season. More than 100,000 observations were submitted on a geographic area of approximately 30000 km² in Laikipia and Samburu counties. We then used different quality-assessments methods to select a subset of high-quality observations to be integrated in the subsequent analyses.
The land cover was mapped using time series of Normalized Difference Vegetation Index (NDVI) imagery acquired from TERRA-MODIS between 2001 and 2016. Unsupervised classification was iteratively performed to separate classes with similar vegetation growth trajectories and to generate a land cover map reflecting spatial variations in rangeland type and production dynamics. Descriptive per-class statistics of crowdsourced indicators (e.g. density of observations during the rain and dry seasons, tree/shrub/grass composition, stocking rate, distance from water) were then calculated and used to define a classification legend combining the different information sources.
The obtained results show that the integration of the local ecological knowledge from pastoralists into the land cover classification scheme is of critical value to understand rangeland resource use spatial patterns. Areas characterized by more productive rangelands are often poorly accessible and/or usable, particularly during the dry season, because of the lack of water in close range, complex topography, or land tenure constraints (e.g. conservancies). As a result, grazing is mostly concentrated on more accessible, but less productive, areas in the lowlands, suggesting not-optimal resource use and potential land degradation issues.
Thus, the final land use map combining mobile-crowdsourced with satellite remote sensing observations could be an effective support tool for informed decision-making on land use planning and pasture resource management.

Area D - Board 385

Salinization is a common problem for agriculture in arid and semi-arid regions and has string impact on crop yields and agriculture. Soil salinity has significantly affected the agricultural lands productivity in the west side of Urmia lake. Mapping soil salinity is difficult due to the soils large spatial and temporal variability. According to this an efficient method for detecting soil salinity is remote sensing science, which is widely used to detect the soil salinity and its progress. Considering these cases, this study focuses on the increase of saline lands and soil salinity in the west side of Urmia lake, using the Landsat and Sentinel 2-A satellite data in 2000, 2005, 2010 and 2017 years. In order to study the progress of soil salinity, west side basins of Urmia lake are Nazluchay, Shahrchay, Barandouzchay and rouzechay. Urmia lake is chosen because of serious salt aggregation during of lake drought in the past decade. The transformations applied to the images in order to detect the salinity and its improvement and to enhance the correctness and resolution of maps are Principal Component Analysis (PCA), ND visualization, Tasseled Cap, salinity indices such as SI, SI1, NDSI (Normalized Difference Salinity Index) and NDVI (normalized Difference Vegetation Index) index. In this study the band combination used to give the best match and best visualization in order to detect the salinity is SI, Red and NIR. To obtain the final salinity map Maximum-likelihood classification method is used which can show the best results in detecting the salinity and extracting the land use and soil salinity map. Kappa coefficient and overall accuracy is derived to indicate the accuracy of classification method and the accuracy of salinity map. The results showed that the images can highly detect and separate the salt areas from crop lands and water. The salinity indices show high accuracy to detect the saline lands in the satellite images and the images used has high ability to detect the salinity and saline lands. In this study kappa coefficient rate for the mentioned band combination is 0.99 and accuracy is 98 percent which confirms that chosen classification method is best classifier in order to detect saline lands with high accuracy, also shows that extracted maps are highly accurate and it confirms that the used salinity index can detect saline soils with high accuracy.

Area D - Board 386

The accuracy and precision of satellite retrievals of sea surface temperature (SST) in nearshore coastal waters are not well known, owing to a lack of in situ data available for evaluation. Over the past five years we have been testing tools for measuring SST in nearshore waters when surfing. Here we demonstrate that one tool, the Smartfin (a surfboard fin that measures water temperature among other variables, see https://smartfin.org), is well suited for evaluating, and consequently improving, satellite retrievals of SST in the nearshore. We first report on the accuracy and performance of the Smartfin in both the field and laboratory. Field comparisons were conducted at beaches and during an Atlantic Meridional Transect (AMT28) field campaign spanning a wide range of environments (with SST varying over 20 degrees Celsius). On AMT28, the Smartfin was found to compare well with various instruments measuring surface water temperature, including skin measurements from the Infrared Sea surface temperature Autonomous Radiometer (ISAR). Initial comparisons with satellite data in the nearshore have revealed large uncertainties in the satellite data, over twice that observed from other published validations using datasets more than a few kilometres offshore from the coastline. These initial comparisons highlight issues with land contamination in the nearshore satellite data, caused by either land directly within the pixel, potentially impacted by errors in geo-location, or possibly through thermal adjacency effects. We show there is a huge potential to expand in situ observations of nearshore SST through tapping into recreational watersports like surfing. Considering the importance of water temperature in controlling the physical, biological and chemical marine environment, and considering the value of the nearshore, using these measurements synergistically with satellite thermal datasets has the potential to improve environmental monitoring. Engaging citizens, such as surfers, in this process will also help improve ocean literacy and marine stewardship.

Area D - Board 387

Two of the oldest instruments in optical oceanography are the Secchi disk and Forel Ule (FU) colour scale. The Secchi disk is a white disk that is lowered into the water and the depth at which it disappears and reappears from sight is logged. This depth is proportional to the clarity or transparency of the water. The FU colour scale is used to classify the colour of natural waters. It consists of 21 colours ranging from blue to green to brown and is used alongside the Secchi disk with the observer typically recording the colour from a submersed Secchi disk (at roughly half the Secchi-depth). Despite the establishment of precision optical instruments, designed to measure the colour and transparency of water electronically, the Secchi disk and FU colour scale are frequently used in modern times, owing partly to their simplicity and low cost. Unlike precision electronic instruments, the observer is implicitly connected with a Secchi and FU measurement, considering both these measurements are dependent on the psychophysiology of the human eye-brain system and the individual's perception of colour and contrast. The Secchi depth and FU colour scale are thus useful tools for teaching the basic concepts of optical oceanography, teaching the measurement process and engaging participants.

Three-dimensional printing is a process where successive layers of material are laid down onto a surface to form virtually any shape from a digital model. Basic desktop 3D-printers can be purchased at low cost, making desktop computer aided manufacture possible in any home. 3D printing has the potential to create cheap, bespoke scientific instrumentation that have formerly required dedicated facilities to manufacture. It can put instrument manufacture into the hands of any citizen who has access to a 3D printer.

We have developed a simple hand-held device designed to measure the Secchi depth and water colour (FU) of lake, estuary and near-shore environments. The device is manufactured with marine resistant materials using predominately a 3D printer. It is cheap to manufacture, lightweight and easy to use. We have produced 100 of these devices for the REVIVAL project (Rehabilitation of Vibrio-infested waters of Lake Vembanad) funded under the India-UK Water Quality Initiative of the Natural Environment Research Council (NERC). These devices were printed at a high school with help from students (aged 11-16) and have been distributed to citizens of Lake Vembanad. Data collected will be integrated with other data streams, including satellite observations, to monitor the health of Lake Vembanad.

Area D - Board 388

Remote sensing is a key tool for improving the current understanding and monitoring of inland water quality. Wider application of this approach is contingent on in situ observations of the optical properties of water bodies for the parametrization of the satellite derived data. In many parts of the world, logistic and other constraints make it impossible to gather the required information at an adequate intensity. We describe a technology-based tool which may be widely used by citizen scientists in collaboration with researchers to remedy this severe data gap.
As water colour is a very informative indicator of the ecological state of marine and fresh-waters, a smart phone application was developed and has been used for many years as part of the fully operational EU Citclops project (www.eyeonwater.org). One of the most obvious characteristics of water to the human eye is water colour, which indicates the water column constituent’s in terms of phytoplankton, suspended sediments and dissolved organic matter. In the smart phone application, the photograph of the water surface is compared with a Forel-Ule colour chart by the citizen scientist and other metadata is recorded such as date, time, location, device type etc. The Forel-Ule colour chart has 21 colours ranging from oceanic blue through coastal/lake green waters to turbid brown inland waters. It is therefore important that we attempt to find the linkage between the water colour (and Forel-Ule number) with the optical characteristics of the water body.
In July 2017 we started a program to adapt Eye on Water for Australian conditions. The ongoing support and development of visualization and improvements of the Eye on Water application have been managed by Maris BV in the Netherlands. As project partners, Maris have developed an Australian version with several new features and the ability to transfer the quality assured data to an Australia server for analysis. The aim of the analysis includes comparing the spatio-temporal characteristics of the citizen science data with satellite derived Forel-Ule data.
This project plans to build on work (Novoa et al, 2015, van der Woerd & Wernand, 2018 and Wang et al, 2018) where the satellite remote sensing reflectance is converted into a Hue angle (αw), a measure colour on the chromaticity diagram and then to a Forel-Ule value. Monitoring changes in the water colour can be linked to temporal or process related changes (van der Woerd & Wernand, 2018) or linked to a trophic state (Wang et al, 2018). Importantly the data acquired by citizens will potentially be high value, as they will often be in areas that are impacted by seasons or short duration events like floods and cyclones.
This project aims to harness the enthusiasm of citizen scientists interested in the water quality of their local region. The engaged organisations and communities have been trained on use of the phone application, deployment of a Secchi disk, and basic water quality testing. The consolidated data is viewable on a website and various visualisation tools facilitate analysis by the citizen scientists. The program is growing rapidly and this presentation will outline the middle stage of the project’s implementation, and outline the plan of integrating spatially and temporally these non-traditional data into satellite water quality monitoring.

Area D - Board 389

On 29th of August 2017 the German Aerospace Center (DLR) Institute for Data Science was established in the city of Jena, Germany. This institute will focus on data related research such as big data analysis, data mining, software and IT security, or artificial intelligence. One of the six workgroups of this institute is dedicated to citizen science related scientific questions. The workgroup will benefit from the close cooperation with local and regional research institutes such as the Friedrich-Schiller-University Jena (Institutes for Informatics and Geography) and the Bauhaus University in Weimar (Departments for Media Informatics, Big Data Analytics/Web Technology and Information Systems).
Although the tremendous potential of citizen science has been recognized and partly used within various thematic fields, some important issues lack thorough research. This includes i) sociologic and psychologic aspects with regard to the engagement and motivation of people, ii) the processing of citizen science data, e.g. semantic attribution, knowledge discovery, data quality & reliability assessment, and iii) strategies to optimize the exploitation of the data, e.g. automation, visualization, or handling of heterogeneous data. The citizen science workgroup will undertake dedicated research on the mentioned topics. Thematically, the research will have close links to crisis management using social media data and volunteered geographic information as well as to earth observation (remote sensing), e.g. local to global reference data generation for model training and product validation. The activities of the workgroup will benefit from a permanent citizen science lab acting as contact point for interested citizens, providing tools and infrastructure to support citizen projects, and hosting capacity building and education activities.

Area D - Board 390

The potential of using Citizen Science (CS) approaches and new digital technologies in a suitable setting for experimenting and generating new EO products and services is faced by social, scientific, technical and economic challenges. This requires a dynamic approach in which the facilitation of interactions amongst actors, incl. citizens, the wider public, NGOs and non-Space sector actors, takes centre stage in accelerating Earth Observation (EO) data-related innovations and their use.

The approach of the Citizen Science Earth Observation Lab (CSEOL) is firmly based in the open innovation paradigm (i.e. the co-creation of innovation beyond organisational boundaries) and on best practices for creating multi-stakeholder involvement in the innovation process. Specifically, LivingLabs provide an enabling environment to support the innovation process for all involved actors and stakeholders and allows to take into account their motivations.

The principle idea behind the CSEOL methodology is to create an innovation channel which drives the process of generating many ideas on how to explore the potential of CS to exploit EO data towards concrete, implementable projects. The CSEOLab will serve as a central place for idea generation and the co-design of CS EO projects.

Following a competitive Call for Ideas in 2019 to the CS and EO community, CSEOL will generate and fund four pilot projects with the purpose of demonstrating the benefits and value of CS for the exploitation of EO.

The thematic focus of the pilots is proposed to be based on existing ESA programs (e.g. CCI, EO4SD, Copernicus services, Sentinel missions) and usual EO data nomenclature (Data levels), relating to the use and exploitation of EO for a) ESA missions, b) Climate change & sustainable development, and c) Downstream services. This is complemented by topics on how CS can be used for EO exploitation in terms of links with science, society, business and policy/decision making.

These innovative CS EO projects require a supportive yet flexible set up to allow creative exploration of innovative ideas about the use of EO data (e.g. new products and services, new mission concepts, public education, story telling) with a diverse range of actors (citizens, established CS communities, SMEs and private sector players, etc.) and using diverse and emerging digital technologies.

In the short run, CSEOL will result in greater awareness of citizens and participation of CS and non-space actors in using and exploiting EO data in four pilots, taking advantage of new and emerging digital technologies. In the long run, CSEOL will generate a continuous stream of ideas and new CS projects that support the exploitation and validation of EO and help shape ESA missions and concepts by involving the general public, Citizen Scientists and a wide range of non-space actors.

Area D - Board 391

In a wide-ranging science communication project, data visualisation software and content has been developed for ESA’s Climate Change Initiative (CCI). In the spirit of open science, a novel approach has been applied to put the results of this project directly into the hands of decision makers and the public. The CCI is developing key datasets, based on the best available Earth observation technologies, for use in understanding changes to the Earth’s climate. Twenty-two essential climate variables are being developed, with data stretching back in some cases up to forty years.

The CCI visualisation activity introduces the project and visualises its data products using linear and interactive 3D graphics, on desktop computers, tablet devices and the web. Two target audiences are addressed: a technical audience familiar with Earth observation data, and the general public. The project's science communication objectives have been met by developing two software products, with a common interactive graphics engine, and a set of computer graphic animations showing how different climate variables interact.

The CCI Visualisation Tool was rapidly developed for use by scientists to communicate to experts in different fields, programme managers and decision makers at scientific and intergovernmental meetings. The digital book app, Climate from Space, builds on this work with a more user-friendly interface and rich text and image content on mass-market tablet platforms for the general public. In both cases care is taken to follow best practice for scientific data visualisation and science communication. In the digital book, 16,000 words of engaging text places the work in scientific context and explains how the climate variables relate to the reader's everyday experience. In the animations, computer graphic illustrations of microscopic processes such as carbon absorption in the ocean are combined with visualisations of the CCI's global climate data products. Future work will focus on web implementation of the software and tailoring the content for educational use.

Area D - Board 392

IceKing is an app for sustainable tourism on glaciers that is also used to crowd-source photographs from tourists for glacier research. Our vision is to unite the power of the community and space technology, producing valuable data on resource utilisation and distribution, to protect planet Earth. We will achieve this by creating a virtuous cycle between sustainable tourism on glaciers and scientific research.
Glaciers are among the most reliable indicators of climate change, but they are also an increasingly popular travel destination, with around 10 million visitors taking a glacier trip each year. Sentinel-1 SAR and Sentinel-2 optical data, together with ESA historical datasets, can be used to monitor key parameters that show us how glaciers are evolving and their health. However, scientists also need in-situ images to validate and improve their data models and our solution harnesses the power of crowdsourcing, obtaining photos from tourists that can help to create cutting-edge algorithms and state-of-the-art monitoring capabilities for glaciers worldwide. Our intention is to combine information acquired thanks to GNSS services (e.g. GPS, Galileo) and Copernicus services, by integrating in-situ information collected via crowdsourcing by glacier guides and tourists. In the long term this can also serve the space community by validating scientific models with crowdsourced ground-based data, also supporting the development of new sensors to be integrated either on future satellites, High-Altitude Platform Stations and Remote Piloted Aircraft Systems (e.g. drones).
Our app offers an all-round experience, starting with useful information to organise the trip, from the choice of where to stay to local glacier guides. During the trip, tourists can join a global expedition to study glaciers, by uploading their geo- referenced photos of different glacier spots. After the trip, users will receive additional services such as an online diary thanks to the available GNSS data and the photographs taken on the glacier. Their participation to the glacier campaign will increase their curiosity and engagement towards the research community. Moreover, they will receive easy to read infographics translating the scientific data into meaningful information on the evolution of glaciers in response to climate change. The app is being developed through funding from the ESA Technology Transfer program, in cooperation with the Institute for Interdisciplinary Mountain Studies in Innsbruck and in accordance with the glacier classification system of the World Glacier Monitoring Service.

Area D - Board 394