The German sustainability strategy uses three indicators to measure the loss of land resources due to urbanization: The daily rate of urban expansion, the loss of open space, and the density of settlements. The quantitative objective is to reduce land take from a rate of 52 hectares per day in 2019 to less than 30 hectares per day by 2030, and further to zero by 2050. The data source used to measure the indicators are land use statistics aggregated from the cadastral survey. However, changes in the nomenclature of this dataset have negatively affected the reliability of the time series since the introduction of the ALKIS cadastral system in Germany from 2015 onwards. Alternatives on a European level such as Urban Atlas, Imperviousness and CLC land cover maps are frequently used for monitoring urban land changes instead. But these datasets also come with limitations including a low update frequency (every 3 or 6 years), coarse mapping units, and the inability to capture land use characteristics that match the intentions of policy objectives for urban land use change. They are therefore not fully suitable to monitor the annual land take in line with sustainable land use objectives.

On this background, the incora project [in German: Inwertsetzung von Copernicus Daten für die Raumbeobachtung / Adding value to Copernicus data for spatial monitoring] conducted by ILS - Research Institute for Regional and Urban Development gGmbH in cooperation with the BBSR - Federal Institute for Research on Building, Urban Affairs and Spatial Development and mundialis GmbH explores the potential of Earth observation techniques to support land use monitoring progress. Based on Sentinel-2 satellite images, our project applied machine learning, automatic training data generation and sub-pixel analysis techniques to produce annual Germany-wide land cover and imperviousness maps. A dedicated companion poster covers in detail the Sentinel-2 data processing steps while this contribution presents the overall project and its results.

The mapping products and the following change detection results were used to quantitatively measure urban land, open space, as well as their changes. By integration of auxiliary data, our project further calculated the annual urban land take, annual loss of open spaces, annual change of settlement density, structural characteristics as well as other indicators that are of interest to stakeholders. The project results demonstrate the added value of Copernicus data for support sustainable land use, including firstly, the derived products are spatially continuous and highly consistent in time series, which provides a solid base to study urban land dynamics at the national, regional and municipality levels. Secondly, it provides not only quantitative measurement but also the spatial distribution of urban land changes, which can be used for the analysis of urban sprawl and land fragmentation.

Expert feedback emphasizes that the monitoring of sustainable land use targets requires high validity of mapping results. We thus highlight our imperviousness mapping, an innovative subpixel analysis method, which improves the measurement accuracy in annual land take (see attached figure). The key advantages of this method are that firstly, it deals with the mixed pixel problem which often causes misclassification in the urban area. Secondly, the change layer contains less noise caused by the random changes in atmospheric conditions, sun angle, soil moisture, vegetation phenology. One disadvantage is that imperviousness is restricted to one land cover only. This was exactly compensated by the land cover classification map. Image classification provides specific land use classes and is, therefore, suitable to detect and characterize land use flows from previously open spaces to urban land in more detail. The conference presentation will show how satellite image analysis helps to validate and improve the information sourced from land use statistics, and how the incora project results contribute to more up-to-date and consistent information on land take for the monitoring of sustainable development targets.

The mitigation of human micro-nutrient deficiency (MND) is a major aim of United Nations Sustainable Development Goal 2 to “End hunger, achieve food security and improved nutrition and promote sustainable agriculture” by 2030. MND has been coined the “hidden hunger” because food supply may be sufficient but lack quality in terms of vitamins and minerals. Hidden hunger can lead to serious health problems including impaired physical and mental human development, susceptibility to various diseases, and reduced learning capacity. The prevalence of MND is assessed at the national level with Food Consumption Tables (FCTs) which report the nutrient content of grain or other crop yield. FCTs are seldom updated and many countries who lack resources employ FCTs from other countries. In doing so, it is assumed that the nutrient content is relatively stable over space and time. This assumption is a major source of uncertainty in nutrition analysis, because in reality nutrient content varies according to various crop growth conditions.

Satellite image data may be able to improve estimates of nutrient content, because they are able to monitor crop growth and development frequently over large areas. The bulk of data however consist of a few broad ( > 10nm) spectral bands. These bands are too coarse to distinguish many of the biophysical properties and biochemical processes related to nutrient content. Hyperspectral image data consist of several narrowbands that are able to make this distinction. Until recently, these data were collected mainly with spectroradiometers in the laboratory, field or mounted to drones and aircraft. The PRecursore IperSpettrale della Missione Applicativa (PRISMA) platform was launched on March 22, 2019. It is the first spaceborne hyperspectral mission in almost 20 years and ushers in a new era in such missions.

In our study, we used PRISMA hyperspectral narrowbands to predict the concentration of eight important nutrients (Ca, N, S, P, K, Fe, Mg, Zn) in four important global staple grains (corn, rice, soybean, wheat). We compared performance of PRISMA to Sentinel-2. Sentinel-2 is a multispectral broadband platform, but contains five experimental red-edge and near infrared narrowbands. The images were collected at the three main stages of crop development (vegetative, reproductive, maturity) in the 2020 growing season. A field campaign was conducted concurrently to sample grains over 60×60m2 survey frames. The nutrient content of the grains was determined in a laboratory using inductively coupled plasma - optical emission spectroscopy and a carbon, hydrogen, and nitrogen analyzer. The nutrients were predicted using Random Forest.

The accuracy of the PRISMA-based predictions (mg kg-1) of the nutrient composition of wheat ranged between a minimum of R2 = 0.49 (RMSE = 0.25) for N and a maximum of R2 = 0.74 (RMSE = 9.20) for Zn. The prediction accuracies for rice ranged from R2 = 0.54 (RMSE= 448.01) for P and to R2 = 0.73 (RMSE = 80.33) for S, for corn from R2 = 0.51 (RMSE = 279.17) for P and R2 = 0.73 (RMSE=99.92) for S. For soybean, the highest prediction accuracy was obtained for Ca (R2 = 0.88, RMSE=241.99). Nutrient composition predictions for wheat using Sentinel-2 ranged from R2 = 0.4 (RMSE = 0.29) for N to a maximum of R2 = 0.76 (RMSE = 373.5) for K, from R2 = 0.39 (RMSE= 12.17) for Fe. Good prediction results were obtained for rice, especially for Ca (R2 = 0.55, RMSE = 90.38), P (R2 = 0.51, RMSE = 235.4) and K (R2 = 0.50, RMSE = 288.7). Nutrient composition of soybean showed better prediction accuracy than the other crops with R2 >0.78 for all target nutrients. The sensitive spectra for various nutrients varied across the four investigated crops but they were mainly concentrated in the visible and Short-wave infrared (SWIR) regions for RPISMA data and, NIR, red-edge and SWIR regions for Sentinel-2.

Our study represents a first important step towards using remote sensing imagery for predicting the nutrient concentrations of crops. Future work will focus on testing the robustness of our predictions across larger geographic areas and several growing seasons.

In the framework of the United Nations (UN) 2030 Agenda for Sustainable Development and the New Urban Agenda (Habitat III), local and regional authorities require indicators at the intra-urban scale to design adequate policies in support of the Sustainable Development Goal (SDG) 11: Make cities and human settlements inclusive, safe, resilient and sustainable. Nevertheless, the current literature provides mainly national, regional and city scale indicators. Earth Observations (EO) data have been recently recognized as an essential source of information to achieve the SDG 11 targets and progress measurements with respect the SDG 11 indicators. However, the complexity of EO data handling and processing in SDGs monitoring and reporting mechanisms makes difficult a direct integration in evidence-based decision-making process.
In order to fill such gaps, this work presents the development and implementation of a set of workflows aimed at the automatic computation of some SDGs 11 indicators at intra-urban scale. A workflow is a process for generating knowledge from observation/simulation data and scientific models. The Virtual Earth Laboratory (VLab) framework (Santoro et al., 2020) was used as a cloud-based platform for sharing and facilitating the invocation of such scientific workflows from urban planners or technical employers of public administration without an extensive expertise on the EO domain. VLab implements all required orchestration functionalities to automate the technical tasks required to execute a model on different computing infrastructures, minimizing the possible interoperability requirements for both model developers and users.
A first workflow has been designed to extract essential variables devoted to the study of urban ecosystems such as the settlement map (built-up) and the population density map. The former can be obtained from a semi-automatic Sentinel-2 data classification procedure or directly by downloading the available European Settlement Map from the Copernicus Land Monitoring Service. Both maps are available at 10 m spatial resolution. Concerning the second variable, a specific workflow was developed for generating a population density map at a fine grid size (100 m X 100 m) from the ancillary population data per census area (Aquilino et al., 2020).
Additional workflows have been implemented for the computation of SDG 11.2.1 “Proportion of population that has convenient access to public transport” and SDG 11.3.1 “Ratio of land consumption rate to population growth rate.” The output maps are generated at regular grid of 100 m spatial resolution size.
Ancillary input, such as the population census data, as well as the local public transport map and additional auxiliary data need to be obtained from local authority providers. The diffusion of open data web portals results promising for the acquisition of such data for many cities.
The height of the buildings information (e.g., LiDAR data) is an optional input that, if available, allows to generate a population density map by applying the improved approach suggested by (Aquilino et al., 2021).
The workflows are validated considering the three Italian towns of Bari, Bologna and Reggio Calabria.
For reproducibility in other cities, the workflows are flexible to a wide variety of formats and geographical reference systems as input. GDAL/OGR standard formats are accepted. An advanced version of workflows is available for those expert users who intend to customize configuration parameters of the models. Besides, VLab frameworks make available a set of Web APIs designed to enable application developers to create dedicated Web applications based on models already available in VLab. By exploiting these latter functionalities, a dedicated web application will be developed as a tool for urban planner and policy-making to make EO data integration in SDGs measuring and monitoring operational for countries.

Aquilino, M.; Adamo, M.; Blonda, P.; Barbanente, A.; Tarantino, C. (2021). Improvement of a Dasymetric Method for Implementing Sustainable Development Goal 11 Indicators at an Intra-Urban Scale, Remote Sensing, Special Issue “Earth Observations for Sustainable Development Goals”, 13, 2835, https://doi.org/10.3390/rs13142835

Aquilino, M.; Tarantino, C.; Adamo, M.; Barbanente, A.; Blonda, P. (2020). Earth Observation for the Implementation of Sustainable Development Goal 11 Indicators at Local Scale: Monitoring of the Migrant Population Distribution, Remote Sensing, Special Issue “EO Solutions to Support Countries Implementing the SDGs”, 12(6), 950, ISSN: 2072-4292, doi:10.3390/rs12060950

Santoro, M., Mazzetti, P., & Nativi, S. (2020). The VLab Framework: An Orchestrator Component to Support Data to Knowledge Transition. Remote Sensing, 12(11), 1795. https://doi.org/10.3390/rs12111795

The SDGs are a universal agenda to address the world’s most pressing societal, environmental, and economic challenges. Robust monitoring mechanisms and timely, accurate and comprehensive data are essential in guiding policies and decisions for successful implementation of the SDGs. Yet official statistics alone cannot provide all of the data needed to populate the SDG indicator framework. Along with EO data, citizen science, briefly defined as public participation in scientific research and knowledge production, offers a new solution and an untapped opportunity to complement traditional sources of data for monitoring progress towards the SDGs. The complementarity of citizen science and EO approaches for SDG monitoring has also been acknowledged in recent literature. For example, Fraisl et al (2020) carried out a systematic review of all SDG indicators and citizen science initiatives, demonstrating that citizen science data are already contributing and could contribute to the monitoring of 33 per cent of the SDG indicators. As part of this review, Fraisl et al. also identified overlap between contributions from citizen science and EO for SDG monitoring based on the mapping exercise undertaken by GEO (2017). The GEO analysis demonstrated that EO data can contribute to the monitoring of 29 SDG indicators, and Fraisl et al showed that citizen science could support 24 out of these 29 SDG indicators, which shows the complementarity between both data sources. One specific citizen science tool integrating citizen science and EO approaches that could complement and enhance official statistics to monitor several SDGs and targets is Picture Pile. Designed to be a generic and flexible tool, Picture Pile is a web-based and mobile application for ingesting imagery from satellites, orthophotos, unmanned aerial vehicles or geotagged photographs that can then be rapidly classified by volunteers. Picture Pile has the potential to contribute to the monitoring of fifteen SDG indicators covering areas, such as deforestation, post disaster damage assessment and identification of slums, among others, which can provide reference data for the training and validation of products derived from remote sensing.

This talk presents the potential offered by Picture Pile and other citizen science tools and initiatives to complement and enhance official statistics to monitor several SDGs and targets. Another example of the use of citizen science for SDG monitoring that will be highlighted in this talk is the Citizen Science for the SDGs project implemented in Ghana for monitoring SDG indicator 14.1.1b on plastic debris density. This project is a partnership between IIASA, the Ghana Statistical Service, the Ghana Environmental Protection Agency, UNEP, Ocean Conservancy, Earth Challenge, and others. The main achievement of the project is that citizen science data for monitoring beach litter have been integrated into the official SDG monitoring and reporting mechanisms of Ghana, which makes Ghana the first country to report on SDG indicator 14.1.1b and the first country using citizen science data for that purpose. Additionally, these data will serve as inputs to Ghana’s Ocean Plan, currently under development, as well as other relevant policies to address the marine litter problem. At the end of the talk, recommendations will be provided for how to enable partnerships and collaborations across data communities and ecosystems in order to bring citizen science and EO data into official statistics for SDG monitoring and reporting.

Coastal eutrophication is a global challenge that can result in harmful algal blooms, hypoxia, fish kills and other negative environmental impacts. To track the status of coastal eutrophication, Sustainable Development Goal 14 (as set by the United Nations General Assembly) includes an indicator, 14.1.1a - Index of Coastal Eutrophication Potential. Indicator 14.1.1a monitors changes in eutrophication directly, by analysing nutrients, or indirectly, by analysing processes that are caused by or are related to nutrient inputs such as algal growth. One of the challenges for tracking global eutrophication is lack of globally available and comparable nitrogen and phosphorus measurements for coastal areas. To address this gap, GEO Blue Planet, Esri and the UN Environment Programme have developed sub-indicators for the methodology to report on SDG 14.1.1a that use global satellite-derived chlorophyll-a products as outlined in United Nation Environment Programme’s Global Manual on Measuring SDG 14.1.1, SDG 14.2.1 and SDG 14.5.1. These global indicators are part of a progressive monitoring approach that seeks to build a foundation of available data that countries can build upon as they develop capacity for reporting on regional and national satellite data and in situ data. In this talk, we will present these global indicators along with efforts to work with member countries to utilize the resulting data for decision making and additional dashboards and visualizations being produced to assist with the identification of potential eutrophication hot spots and inform further analysis. The indicators include an indicator derived from the global ocean, 4km spatial resolution per pixel monthly mean product of the Ocean Colour Climate Change Initiative project’s product for each pixel within a country’s EEZ and an indicator derived from the National Oceanic and Atmospheric Administration’s VIIRS chlorophyll-a ratio anomaly product produced for the globe at 2km spatial resolution. The indicators will be compared to in situ data from member countries. Challenges to data access, product validation and other issues will be addressed.

ADVANCED EARTH OBSERVATION SATELLITE TECHNOLOGY. AN INTEGRATED SYSTEM TO SUPPORT THE ACHIEVEMENT OF SUSTAINABLE DEVELOPMENT GOALS AT THE COLOMBIAN AMAZON
Quiñones, M.J; Vissers, M.; Hoekman, D.; Kooij, B.; Luiken, R.; VanRooij, W & Pratihast, A.K.; Murcia, U.; Gómez, L.A.; Cuchía, A.; Acosta, H.H.; Carvajal, H.E.; Gil, C.; Rojas, A. & Erazo R.
In the frame of the ESA supported “EOSAT 4 Sustainable Amazon” project, advanced earth observation (EO) radar satellite technology was demonstrated to support the achievement of Sustainable Development Goals (SDG) for the Colombian Amazon. In close collaboration with Colombian governmental and non-governmental stakeholders, working for the sustainable development of the Colombian Amazon, EO radar-based products were defined in terms of required thematic information, spatial and temporal resolution.
Standardized Radar pre-processing steps follow automatic routines using in house developed algorithms and other commercial software like Gamma and IDL. Data downloads, interferometric registration, multitemporal filtering and orthorectification, precede thematic processing using in house time series analysis algorithms like SarSentry, SarFlood and baseline mapping algorithms like SarEcomap and SarSoil. Thematic Products include: 1) Deforestation historical change time series 2007-2017, based on ALOS-Palsar data; 2) Deforestation and degradation historical change time series 2017-2020, based on Sentinel-1 time series; 3) Baseline Ecosystem Map 2017 with 40 vegetation structural classes based on the classification of radar and optical data; 4) Flooding dynamics maps and flood frequency map, based on ALOS PALSAR ScanSAR data; 5) Time series 2007-2020 Fire dynamics time series 2017-2020, data from MODIS and VIIRS systems; 6) Soil degradation map-Area of Florencia 2017-2020, based on combined classification of ALOS PALSAR and Sentinel-1; 7) High resolution ecosystem mapping; 8) River border and island dynamics 2017-2020, based on Sentinel-1 time series analysis; 9) Location of Rock formations: Tepuis, based on SRTM analysis and 10) Near real-time monitoring for deforestation and degradation in 2021. Products were presented to the users through an in-house developed web-based GIS platform, created especially for the efficient exploration and integration of the data including some analysis tools.
Product evaluation and validation was done both by producers and users. Validation included the use of georeferenced aerial photographs acquired by the users and the comparison with existing available data sets, like thematic maps and deforestation information form the Biomass, GLAD-2 and RADD systems. In addition, the usefulness of the data was evaluated using a dedicated questionnaire. In general, all products were considered to be of excellent quality in terms of contents, and spatial and temporal resolution to monitor the environmental conditions of the Colombian Amazon. Some unique thematic information like the forest degradation, flood frequency, soil degradation and high-resolution ecosystem mapping exceeded the expectation of the partners, and were considered as unique and necessary to achieve the SDG. In general, the EO products created in the frame of this project can support the monitoring of the implementation of Sustainable Development Goals. Specifically: 1) Creation of new protected areas; 2) Improved management of protected areas; 3) Support Land use plans and Life plans for indigenous communities; 4) Climate change monitoring; 5) Sustainable Forest management; 6) Land and forest restoration projects; 7) Support Productive systems and value chains for green business; and last but not least communication and sensitization on environmental crisis to the Colombian civil society.
An integrated Monitoring system was proposed and evaluated by the users, in order to integrate the developed EO based products into the different activities supporting the achievement of the SDG. The frame consists of three main components: 1) Data production; 2) Data integration and 3) Data communication.