Digital Earth: Big Earth Data Concepts 2024/2025 PDF

Digital Earth Big Earth Data Concepts Winter 2024/2025 | Martin Sudmanns & Dirk Tiede Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Organisation and structure “This lecture focuses on a recent trend in Earth o...

Digital Earth Big Earth Data Concepts Winter 2024/2025 | Martin Sudmanns & Dirk Tiede Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Organisation and structure “This lecture focuses on a recent trend in Earth observation (EO) called “big Earth data” and covers or touches topics such as accessing & processing of massive amount of EO data, data cubes, artificial intelligence (knowledge-based systems & machine learning), and possible application areas for continental- or global-scale remote sensing image processing. In a mixture of theoretical and hands-on sessions students will learn to understand current trends and their backgrounds as well as applying new concepts and approaches in remote sensing.” − 9 sessions, 3 hours per session − 3 ECTS − Mixture of hands-on and theoretical parts − Blackboard support (materials and discussions) − Final exam and graded assignments − De-register after 2 session if you don‘t plan to complete the course − No recording of the sessions − Remote sensing background required, but can be aquired in the beginning of the semester Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 2 Grading criteria Attendance: 70 % of the classes Active participation Submit and pass all exercises / assignments Pass the final exam Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 3 Generative AI usage policy What could possibly go wrong … Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 4 Sessions overview Date Topic Comments 11.10.2024 Introduction to big Earth data / Digital Earth 18.10.2024 Data selection and Analysis-Ready Data in a big Assignment 1 Earth data environment 25.10.2024 Copernicus: The European Union's Earth Assignment 2 Observation Programme 08.11.2024 Google Earth Engine 15.11.2024 Google Earth Engine Assignment 3 06.12.2024 EO Data Cubes 20.12.2024 Sentinel-2 Semantic EO Data Cube Austria Assignment 4 10.01.2025 Time vs space and spatial concepts 17.01.2025 Deep Learning 31.01.2025 Exam Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 5 Who we are Dr. Martin Sudmanns PostDoc at research group EO Analytics, Department of Geoinformatics - Z_GIS Research group co-head EO Analytics, Department of Geoinformatics - Z_GIS Research topics: Big Earth data, Data Management, (time series) EO data analysis Main research / contribution: Sentinel-2 semantic data cube for Austria Dr. Dirk Tiede Associate Professor remote sensing and GIS, Uni Salzburg Deputy head of the Department of Geoinformatics – Z_GIS Research group co-head EO Analytics, Department of Geoinformatics - Z_GIS Involved in a range of research projects, mainly related to EO data & analysis. Research fields include EO based environmental monitoring and support of humanitarian relief operations https://eo-analytics.zgis.at Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 6 About Big Earth Observation Data at Z_GIS Main people involved: Dr. Dirk Tiede, Dr. Martin Sudmanns, Hannah Augustin, Thomas Strasser, Dr. Andrea Baraldi, Luke McQuade, Felix Kröber, Antonio Pandza, Vanessa Streifeneder … Part of the EO Analytics group: https://eo- analytics.zgis.at Several research projects since 2015: AutoSentinel 2/3, SemEO, Sen2Cube.at, SemantiX, SIMS, INTERFACE, ESG-Pro, Space4AD, LEONSEGS, Prometheus International network Copernicus (EU), CSIRO (Australia), Uni Geneva, Uni Bern (Switzerland), ERATOSTHENES Centre of Excellence (Cyprus), German Aerospace Center, Digital Earth Africa … Lecture during winter term Big Earth Data: Digital Earth Concepts Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 7 References and literature − Sudmanns, M., Augustin, H., Killough, B., Giuliani, G., Tiede, D., Leith, A.,... & Lewis, A. 2022. Think global, cube local: an Earth Observation Data Cube’s contribution to the Digital Earth vision. Big Earth Data, 1-29. − Sudmanns, M., Tiede, D., Lang, S., Bergstedt, H., Trost, G., Augustin, H., Baraldi, A. and Blaschke, T., 2020. Big Earth data: disruptive changes in Earth observation data management and analysis?. International Journal of Digital Earth, pp.1-19. − Guo, H., 2017. Big Earth data: A new frontier in Earth and information sciences. Big Earth Data, 1(1-2), pp.4-20. − Baumann, P., Mazzetti, P., Ungar, J., Barbera, R., Barboni, D., Beccati, A., Bigagli, L., Boldrini, E., Bruno, R., Calanducci, A. and Campalani, P., 2016. Big data analytics for earth sciences: the EarthServer approach. International Journal of Digital Earth, 9(1), pp.3-29. − Sudmanns, M., Lang, S. and Tiede, D., Big Earth Data: From Data to Information. GI_Forum 2018, 1, pp.184-193. − Guo, H., Goodchild, M.F. and Annoni, A., Manual of Digital Earth. Springer Nature 2020. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 8 References and literature − Manual of Digital Earth: − Open access book: https://link.springer.com/book/10.1007/978- 981-32-9915-3 − “This open access book offers a summary of the development of Digital Earth over the past twenty years. By reviewing the initial vision of Digital Earth, the evolution of that vision, the relevant key technologies, and the role of Digital Earth in helping people respond to global challenges, this publication reveals how and why Digital Earth is becoming vital for acquiring, processing, analysing and mining the rapidly growing volume of global data sets about the Earth.” Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 9  Introduction Big Earth Data Concepts Winter 2024/2025 | Martin Sudmanns & Dirk Tiede Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at CORONA Information The CORONA program was a US-American strategic military program (reconnaissance satellite) where film capsules were parachuted back to Earth and caught in-flight by the U.S. Air Force. It was the predecessor of today’s Keyhole program. More info in Dawson 2018. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 11 Digital Earth Digital Earth is the name given to a concept by former US vice president Al Gore in 1998, describing a virtual, digital representation of the Earth that is georeferenced and connected to the world’s digital knowledge archives. − More than a virtual globe − A digital Earth exist in parallel to the physical Earth with translation layers between them (sensors  visualization) − Problems can be described, solved and transferred in the digital Earth − Earth observation “part”: Big Earth Data Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 12 Digital Earth: Scientific Journals https://www.tandfonline.com/toc/tjde20/current https://www.tandfonline.com/toc/tbed20/current Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 13  https://earthengine.google.com/timelapse/ Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 14 What is big Earth data? Unprecedented volume, variety and velocity of Bring users to the data, not EO data the data to the users New user types and applications Increasing use of EO data, documented by number of registered users at the ESA portals Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 15 Data Deluge: Sentinel-2 example >80 mil. Sentinel-2 Scenes in the archive 3 TB new data volume every day 70 % of the Earth is cloud covered Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 16 Characteristics of Big Earth Data Data are Open Access Data not for a single purpose, but constant stream Volume, velocity, variety Increased acquisition frequency Decreased time between acquisition and data provision Image archive date back some decades (e.g., MODIS, Landsat) Availability of computing resources Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 17 Characteristics of Big Earth Data Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 18  Scanning the Earth‘s surface: Optical vision Satellite / Platform Swath / Sensor. Multiple sensors Field of view per satellite possible Acquisition during day only Different wavelengths (bands) Atmospheric / weather influeces © ESA, artist‘s view Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 19 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 20  Scanning the Earth‘s surface: Radar vision Satellite / Platform Day & night acquisition (asc. / desc. orbit) Sensor & Antenna Swath. Different Modi possible Less atmospheric influeces © ESA, artist‘s view Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 21 Scanning the Earth‘s surface https://sentinel.esa.int/web/sentinel/user-guides/sentinel-2-msi/revisit-coverage Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 22 Scanning the Earth‘s surface Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 23 From data streams to image files Space segment: Different topic... Ground segment (1): Lots of technical decisions (input & output, reprocessing) Raw data stream Defining processing baseline Ground segment (2) / data provider: Image file Providing data, e.g., in download portals In Future: using exploitation platforms and online processing End user: Creating value from data Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 24 Spatial vs temporal resolution Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 25  Effects of seasonality on optical acquisitions https://scihub.copernicus.eu/twiki/pub/SciHubWebPortal/AnnualReport2017/COPE-SERCO-RP-17-0186_-_Sentinel_Data_Access_Annual_Report_2017-Final_v1.4.1.pdf Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 26 Requirements of application domains Kennedy, R.E., A., S., Cohen, W.B., Gomez, C., Griffiths, P., Hais, M., Healey, S.P., Helmer, E.H., Hostert, P., Lyons, M.B., Meigs, G.W., Pflugmacher, D., Phinn, S.R., Powell, S.L., Scarth, P., Sen, S., Schroeder, T.A., Schneider, A., Sonnenschein, R., Vogelmann, J.E., Wulder, M.A., Zhu, Z., 2014. Bringing an ecological view of change to landsat-based remote sensing. Front. Ecol. Environ. 12, 339– 346. doi:10.1890/130066 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 27 Temporal scale of EO missions Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 28 Temporal scale of EO missions Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 29 Spatial coverage: Sentinel-2 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 30 Spatial coverage: Sentinel-2 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 31  Data access / Ground segment (Copernicus) Copernicus data access: Different user types, access options, … Generic, worldwide portals: Private, Private-Public- Partnerships Specific, (geographically) local implementations Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 32 Sensors (V)HR Thermal Hyperspectral HR Optical Lidar HR Radar VHR Optical Glimmer / Night Light In-Situ Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 33 Which perspective do you want? Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 34  Technology Overview Big Earth Data Concepts Winter 2024/2025 | Martin Sudmanns & Dirk Tiede Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Changes in workflows Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 36 Changes in workflows https://www.youtube.com/ watch?v=AYB5Sw80Vfs Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 37 Technologies: Data Cubes Yellow: Operational Red: Planned or under construction https://medium.com/opendatacube/what-is-open-data-cube-805af60820d7 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 38  55 27 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 39 Technologies: Google Earth Engine Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 40 Technologies: STAC https://data.inpe.br/bdc/web/en/stac-spatiotemporal-asset-catalog-2/ https://knowledge.dea.ga.gov.au/guides/setup/gis/stac/ Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 41 Technologies: Convolutional Neural Network − Machine-learning approach − Mapping of an input image to a target classes using neural networks − Topology-preserving − Requires usually lots of training data Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 42 Technologies: Beyond this class − Much more technology exist − Docker / Kubernetes − OGC APIs − ML/DL Libraries in R, Python, Julia, … − Swarm/Dask on HPC − Data Lakes / Digital Twins − … Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 43  Applications Big Earth Data Concepts Winter 2024/2025 | Martin Sudmanns & Dirk Tiede Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Applications End users don‘t care about (big Earth) data, they want reliable information Space 4.0: Make MONEY Answers Big Earth data is supposed to be a business Expectations „New kids on the block“: Questions Data science, machine learning, … …? Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 45 Sustainable Development Goals Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 46  No images any more: Treating pixels as observation in space/time Executing the algorithm within a cloud over the Internet No snapshot: “This is a nice map, but how has it changed? ”Mike Wulder: https://www.youtube.com/watch?v=GzIOzTGB-Vs http://www.ga.gov.au/scientific-topics/hazards/flood/wofs/about-wofs Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 47  No study areas any more: Processing the entire Earth „Zero download model“: Data -> Analysis -> Map Information layers in the context of global initiatives Data is open access & free (for end-users) https://lcluc.umd.edu/content/global-forest-change Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 48  Information in real-time General-purpose collection of data Multi-sensor approach: Virtual constellations Continous stream of data and update of information https://emergency.copernicus.eu/ mapping/system/files/components/ EMSR352_AOI08_GRA_PRODU CT_r1_RTP04_v1.jpg Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 49 Example: blackshark.ai https://blackshark.ai Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 50 Example: Missing roads detection for Open Street Map Bing Maps is releasing mined roads around the world. We have detected 48.9M km of all roads and 1165K km of roads missing from OSM. Mining is Not all performed with Bing countries are Maps imagery between covered 2020 and 2022 including Maxar and Airbus. The data is freely available for download and use under the Open Data Commons Open Database License https://github.com/microsoft/RoadDetections/ (ODbL). Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 51 Example: JRC Global Surface Water Explorer https://global-surface-water.appspot.com/ Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 52 Example: ESRI World Imagery Wayback Machine https://livingatlas.arcgis.com/wayback/ Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 53 Space 4.0 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 54 Microsoft Flight Simulator: Aerial / Satellite images and real-time weather https://www.youtube.com/watch?v=ReDDgFfWlS4 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 55  Data selection in a big Earth data environment* Big Earth Data Concepts Winter 2024/2025 | Martin Sudmanns & Dirk Tiede *Within this course we focus on optical image data, mainly Sentinel-2 and Landsat. Radar imagery will not be covered Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 56 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 57 Challenges Useful information produced from a set of images: depending on the task Automated quality check and feasibility evaluation is difficult Transfer data from storage to processing engine Kempeneers, P. and Soille, P., 2017. Flexibility to switch between sensors Optimizing Sentinel-2 image selection in a Big Data context. Big Earth Data, 1(1-2), pp.145- 158. if time is critical Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 58 Copernicus Data Space Ecosystem (CDSE) https://dataspace.copernicus.eu Access to almost all Sentinel data Graphical User Interface (GUI) and Application Programming Interface (API) Open and free (download limited to two products/images per account at the same time) Newest images are online for immediate download, older images are indicated as offline, but will be restored on demand Cloud-based processing options for users available Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 59 EO-Browser EO-Browser (company: Sinergise / Planet) Included (partly) in the Copernicus Data Space Visual data selection through time Manipulation of data visualization (band combination, indices etc.) Time-lapse animations Uses AWS for fast access to Sentinel-2 data (reprocessed to a specific data format, freely accessible via the Amazon open data policy) Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 60  ESA Sentinel-2 processing chain: Level -1 C and Level-2 A data Big Earth Data Concepts Winter 2024/2025 | Martin Sudmanns & Dirk Tiede Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 64 Levels in the processing chain of EO data Processing levels of satellite data are not standardized and depend on the organization (e.g. ESA, NASA other satellite providers) and the different satellites Definition by ESA and CNES, compliant with the norms of the Committee on Earth Observation Satellites (CEOS): Level 1C is a monodate ortho-rectified image expressed in TOA reflectance Level 2A is a monodate ortho-rectified image expressed in surface reflectance, provided with a cloud/cloud shadow/snow/water mask Level 3A is a monthly composite of Level2A Cloud/Cloud shadows free pixels Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 65 Levels in the processing chain of EO data ESA Sentinel-2 processing chain: L0 to L1B not released to users L1C TOARF All data acquired by the MSI instrument are systematically processed to Level-1C by the Payload Data Ground Segment (PDGS). Only the Level-1C and Level-2A products are released to Users. Level-2A products are generated either by the PDGS using the Sen2Cor processor, or on the User side through the Sen2Cor Toolbox (standalone or within SNAP). Since version 2.8 ESA states, that the Sen2Cor processor produces the exactly same results as processing on the user side using the Toolbox (same parametrisation) Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 66 Levels in the processing chain of EO data ESA Sentinel-2 processing chain: L0 to L1B not released to users L1C TOARF Level-1 processing uses consolidated Level-0 data as input. Level-1A processing is focused on decompressing relevant mission source packets. The Level-1A product is not available to users. Level-1B processing uses the Level-1A product and applies the required radiometric corrections. The Level-1B product is not made available. Level-1C processing uses the Level-1B product and applies radiometric and geometric corrections (including orthorectification and spatial registration, calibrated into Top of Atmosphere reflectance (TOARF) values). A rough cloud cover estimation is calculated (metadata which can be used in the data selection step in the different portals), the granules, also called tiles, are 100 x 100 km² ortho-images in UTM/WGS84 projection. https://sentinel.esa.int/web/sentinel/user-guides/sentinel-2-msi/processing-levels/level-1 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 67 ESA Level-2A data Top of Bottom of Classification mask Level 2A for optical EO data according ESA Atmosphere Atmosphere and quality definition1 consists of three components: (TOA) Level 1C (BOA) Level 2A indicators 1. Image corrected for atmospheric, adjacency and topographic effects 2. Classification mask (pre-requisite for atmospheric correction) 3. Quality indicators (e.g. also as masks) 1Level 2A Products Algorithm Theoretical Basis Document (https://earth.esa.int/c/document_library/get_file?folderId=349490&name=DLFE-4518.pdf) From: https://sentinels.copernicus.eu/web/sentinel/user-guides/sentinel-2-msi/processing-levels/level-2 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 68 ESA Level-2A data Look-Up-Tables for extended atmospheric conditions (not for high latitudes / polar regions) ➔ Ground Image Processing Parameters (GIPP) https://www.libradtran.org/doku.php See also: https://sentiwiki.copernicus.eu/__attachments /1692737/S2-PDGS-MPC-ATBD-L2A%20- %20Level%202A%20Algorithm%20Theoretic al%20Basis%20Document%202021%20- %202.10.pdf?inst-v=7e368646-a179-477f- af62-26dcc645dd8a Sorurce: http://step.esa.int/thirdparties/sen2cor/2.5.5/docs/S2-PDGS-MPC-L2A-SUM-V2.5.5_V2.pdf Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 69 ESA Level-2A data Sen2Cor Toolbox for Sentinel-2A processing Basis of the ESA processing chain, but also freely available (stand-alone or within SNAP) The finality of the algorithm (parameter selection, quality of the SCM etc.) and its transferability is still https://www.youtube.com/watch?v=ryGROtiHPYI an issue Download: http://step.esa.int/main/third-party- plugins-2/sen2cor/ Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 70 71 TOARF calibration: S-2 Level 1C, no atmospheric & topographic correction Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Atmospheric & topographic correction : S-2 Level 2A (Sen2Cor approach – ESA) Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at 73 Atmospheric & topographic correction : S-2 Level 2A (Sen2Cor scene classification map – 20m resolution) Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Sentinel-2 Collection 1 Sentinel-2 Collection 1 The continuous updates of the processing baseline creates inconsistent time series. In the collection 1, all images are re- processed with a consistent time series All images are Level-2A (BOA) and compliant with CEOS ARD All images will have a DOI for unique referencing Not yet completely available, currently being produced: https://sentinels.copernicus.eu/web/sentinel/sentinel-data- access/sentinel-products/sentinel-2-data-products/collection-1-level- 2a Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 74  Analysis-ready data (ARD) Big Earth Data Concepts Winter 2024/2025 | Martin Sudmanns & Dirk Tiede Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 75 Analysis-ready data (ARD) Analysts spend too much time in data preparation The community seeks a solution in providing data, which can be direct input to analysis Obstacles can be Data discovery & data access File format transformations Geometric transformation Incompatible processing levels … Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 76 Analysis-ready data (ARD) Existing definitions (selection) USGS “U.S. Landsat Analysis Ready Data (ARD) are consistently processed to the highest scientific standards and level of processing required for direct use in monitoring and assessing landscape change.” CEOS “CEOS Analysis Ready Data for Land (CARD4L) are satellite data that have been processed to a minimum set of requirements and organized into a form that allows immediate analysis with a minimum of additional user effort and interoperability both through time and with other datasets.” Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 77 CEOS Analysis-ready data for land (CARD4L) http://ceos.org/ard/#slide3 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 78 CEOS ARD overview Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 79 CEOS ARD CARD4L Specification Agreement in February 2019 Goals (selection): Maximise impact and potential of EO data Cost-effective and practical distribution of EO data Targeting a wide range of users Interoperable with other ARD (PFS) Procedure (selection): Peer-review of specific products (CARD4L Product Alignment) Promotion of CARD4L compliant data Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 80 CEOS ARD Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 81 CEOS ARD Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 82 Example: Digital Earth Africa Sentinel-1 datasets for entire Africa CEOS ARD compliant Surface reflectance data for Sentinel-2 and Landsat 5,7,8,9 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 83 Example: Digital Earth Africa Right: Landsat 9, Left Sentinel 2, aquisitions at the same day Sentinel-2 and Landsat 9 one-day coverage Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 84 Example: Digital Earth Africa Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 85 The FAIR principle The FAIR Guiding Principles for scientific data management and stewardship Findable: Rich metadata, persistent identifier, indexed metadata. Accessible: Standardised, open, free protocol for metadata access. Interoperable: Dictionary, references to other metadata. Wilkinson, M.D., Dumontier, M., Aalbersberg, I.J., Appleton, G., Axton, M., Baak, A., Blomberg, N., Boiten, J.W., da Silva Santos, L.B., Bourne, P.E. and Bouwman, J., 2016. The FAIR Guiding Principles for Reusable: scientific data management and stewardship. Scientific data, 3. Evans, Ben; Druken, Kelsey; Wang, Jingbo; Yang, Rui; Richards, Clare; Wyborn, Lesley (2017): A Data License, provenance, metadata quality. Quality Strategy to Enable FAIR, Programmatic Access across Large, Diverse Data Collections for High Performance Data Analysis. In Informatics 4 (4), p. 45. DOI: 10.3390/informatics4040045. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 86 Limitations – Example global cloud cover ARD data should reduce pre-processing effort for analysts, so it should be reliable/trustable. Is this always the case? Have a look at the global cloud cover estimations on the following slides: First you see the visualization of Sentinel-2 cloud covers estimation from the EO- compass (the values reflect the Level 1C estimations provided from ESA) Second image shows the same analysis based on MODIS data (less resolution but this doesn't matter on a global scale) What are the differences? Any ideas why? Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 87 Copernicus Scientific Hub – Example global cloud cover Sentinel-2 cloud estimation Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 88  MODIS cloud estimation Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 89 Example: Cloud cover overestimation Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 91  Example: Cloud cover overestimation ▪ Query for all Sentinel-2 Level 1C images available (query conducted on July 10, 2019) for granule 19JDJ and relative orbit R96 (Andes, border region between Chile and Argentina). ▪ Selection criteria: cloud cover estimate of >= 80 % in their metadata. ▪ The result shows the thumbnails of 90 images found in the database; most of them have no or much less than 80% cloud cover, images are returned from all seasons and show minimal variance over the year. ▪ A query for images with 90% estimated cloud cover in 2019. High reflectance in the blue band usually indicates cloud cover (here in blue) while low reflectance values in the blue band are indicating cloud free data (cirrus clouds are not specifically considered in this analysis). Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 93 Example: Cloud cover overestimation - conclusions This specific example should raise awareness for big EO data analyses, where problems in ARD processing (here: cloud cover overestimates) can significantly influence data selection and analysis workflows for specific regions Geographic characteristics and biases of data still need to be considered and checked prior to analyses when processing big Earth observation data – even in "so called" analysis ready data stacks Pre-processing is undergoing constant changes / improvement (see e.g. https://sentiwiki.copernicus.eu/web/s2-processing#S2Processing-L2AProcessingBaselineS2-Processing-L2A-Processing-Baselinetrue ) Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Copernicus: The European Earth Observation Programme Big Earth Data Concepts Winter 2024/2025 | Martin Sudmanns & Dirk Tiede Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Copernicus European system for Earth observation 3rd largest data provider worldwide Online: http://www.Copernicus.eu Founded 1998 by EC and ESA Satellites & in-situ measurement systems Space Component: Sentinel Satellites Earth Observation Satellite Navigation Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 96 Copernicus https://www.copernicus.eu/sites/default/files/2018-10/History_Factsheet_vf.pdf Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 99 The 20 years of Copernicus (1998 – 2018) https://www.youtube.com/ watch?v=ozQtgxsqsUc Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 100 The 25 years of Copernicus (1998 – 2023) https://www.youtube.com/ watch?v=vyoXKUuAf5E Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 101 European Space Agency (ESA) Builds and operates satellites (for the Copernicus Programme and others) Member states or cooperation Sentinel family deliver 16 Tb data per day https://epthinktank.eu/2017/01/31/european-space-policy-historical-perspective-specific- aspects-and-key-challenges/eu-and-esa-member-states/ Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 102 Sentinel satellites http://www.space-airbusds.com/media/image/copernicus-poster-840x297_eng_1_1.jpg Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 105 Sentinel satellites Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 106 Sentinel satellites Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 107 Sentinel-1 Radar satellite C-band synthetic aperture radar (SAR) with 5.405 GHz Near-polar, sun-synchronous orbit Antenna length is 12.3m 4 different acquisition modes Sentinel 1 A in orbit, B not functional, C to be launched in 2025 Acquisition frequency: 12 days (Equator) Main application areas (selection): Sea ice, Oil spill, Ocean current/ wave Sentinel-1 Website Land deformation / earthquake http://www.esa.int/Our_Activities/Observing_the_Earth/Cop ernicus/Sentinel-1/ Emergency mapping/ monitoring (ex. flood) Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 108 Sentinel-1 GRD (Ground Range Detected) product Projected onto ground Intensity only 10 m resolution Flood mapping, deforestation, urban expansion,… applications Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 109 Sentinel-1 SLC (Single Look Complex) product Amplitude Phase Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 110 Sentinel-2 Optical instrument MSI (Multi-Spectral Instrument) 290 km swath 13 spectral channels (443 nm - 2190 nm) Spatial resolution: 10 – 60 m Sentinel 2 A + B + C in orbit Acquisition frequency: 5 days Main application areas (selection): Agriculture, forestry Land-use (change) Sentinel-2 Website Water bodies http://www.esa.int/Our_Activities/Observing_the_Earth/Cop ernicus/Sentinel-2/ Emergency mapping / monitoring Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 111 Sentinel-2 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 112 Sentinel-3 Multiple instruments Ocean and Land Colour Instrument (OLCI) Sea and Land Surface Temp. Radiometer (SLSTR) Synthetic Aperture Radar Altimeter (SRAL) Microwave Radiometer (MWR) Sentinel 3 A + B in orbit Acquisition frequency: 1 day Main application areas (selection): Sea surface temperature / ocean colour Weather forecast Sentinel-3 Website Fire detection http://www.esa.int/Our_Activities/Observing_the_E arth/Copernicus/Sentinel-3/ Vegetation, glacier Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 113 Sentinel-3 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 114 Current Sentinel missions … Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 115 … possible future Sentinel missions Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 116 EO-Browser EO-Browser (company: Sinergise) Visual data selection through time Manipulation of data visualization (band combination, indices etc.) Time-lapse animations Uses AWS for fast access to Sentine-2 data (reprocessed to a specific data format, freely accessible via the Amazon open data policy) Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 117 The Copernicus ecosystem See PDF Copernicus ecosystem Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 118  > Access to Copernicus Data - presentation provided by the Copernicus Support office: https://www.copernicus.eu/en/copernicus-services Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 119 Operational Copernicus applications: Sen2Agri http://www.esa-sen2agri.org/ Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 120 Operational Copernicus applications: Sen2Agri http://www.esa-sen2agri.org/ Cloud-free composite Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 121 Operational Copernicus applications: Sen2Agri http://www.esa-sen2agri.org/ Cloud-free composite Cropland mask Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 122 Operational Copernicus applications: Sen2Agri http://www.esa-sen2agri.org/ Crop type map Cloud-free composite Cropland mask Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 123 Operational Copernicus applications: Sen2Agri http://www.esa-sen2agri.org/ Vegetation status Crop type map Cloud-free composite Cropland mask Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 124 Operational Copernicus applications: Sen4Cap http://esa-sen4cap.org/ Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 125 Operational Copernicus applications: Sen4Cap http://esa-sen4cap.org/ Pilot countries Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 126 Operational Copernicus applications: Sen4Cap http://esa-sen4cap.org/ Pilot countries Output products: Cultivated crop type map Grassland mowing product Vegetation status indicator Agricultural practices monitoring product (http://esa-sen4cap.org/content/eo-products) Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 127 GTIF|Green Transition Information Factory https://gtif.esa.int/explore Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 128  Google Earth Engine Big Earth Data Concepts Winter 2024/2025 | Martin Sudmanns & Dirk Tiede Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 129 What is Google Earth Engine? For a detailed description and how to cite: Gorelick, N., Hancher, M., Dixon, M., Ilyushchenko, S., Thau, D., Moore, R., 2017. Google Earth Engine: Planetary-scale geospatial analysis for everyone. Remote Sens. Environ. 202, 18–27. https://doi.org/10.1016/j.rse.2017.06.031 "Big Data" analysis and visualization platform from Google Inherently parallel system (lot of computational power in the background) Runs mostly on the server ⇒ no need for sophisticated client software, high computational power, high storage capacity nor high internet bandwidth "Designed for scientists, not software engineers" Goals: make it easy, enable non-traditional users Earth Engine is free for research, education, and nonprofit use. For commercial applications, Google offers paid commercial licenses. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 130 What is Google Earth Engine? Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 131 What is Google Earth Engine? Petabyte of data (permanently increasing): Imagery Landsat 4-8 7 bands, 30m MODIS 250m Daily Global Sentinel-2 12 bands, 10m/20m Sentinel-1 10m SAR Geophysical Digital Elevation Land Cover Surface Temperature 29 Weather Forecasts, Climate Models https://developers.google.com/earth-engine/datasets/catalog +300 other datasets Upload your own data (up to 10 GB, Raster and Vector data) Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 132 What is Google Earth Engine? Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 133 How to use? Javascript API Interactive Code Editor (N. Gorelick) Node.js* Python API Python module Web Apps with Appengine Jupyter Notebooks* Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 134 Useful links Starting page (Overview, Timelapse etc.): https://earthengine.google.com/ Standard interface / standard tools (extended if signed up): https://explorer.earthengine.google.com API / JavaScript Code: https://code.earthengine.google.com/ Data Sets: https://developers.google.com/earth-engine/datasets/ Tutorials/Help: ▪ Starting page: https://developers.google.com/earth-engine https://developers.google.com/earth-engine/getstarted https://developers.google.com/earth-engine/help ▪ Code Engine Overview https://developers.google.com/earth-engine/playground ▪ API references / Syntax: https://developers.google.com/earth-engine/api_docs ▪ Tutorials: https://www.eefabook.org https://courses.spatialthoughts.com/end-to-end-gee.html Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 135 GEE Concepts GEE Concepts* *The following slides are based on a presentation by Noel Gorelick (GEE) Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 136  Data Models Feature Line / Point / Polygon List of Properties TNC Ecoregions Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Data Models Feature Image Stack of Georeferenced bands Each band has its own: Mask, Projection, Resolution A list of properties, including: Date, Bounding-box Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Data Models Feature Image Collection Bag of Elements Table of Features Directory of Images Filter, Sort, Join, Map, Reduce Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Map Apply a function to each element of a collection A "For Each" operation Examples Compute area of each feature Cloud cover of each image Mosaic for each month Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Reduce Apply a function to everything in a collection "Aggregation" Examples Summed area over all features Median-pixel composite Train a classifier Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Reducers in Earth Engine 8 ways to reduce 40+ reducers Reducer.allNonZero Reducer.min Image.reduce Reducer.and Reducer.minMax Image.reduceNeighborhood Reducer.anyNonZero Reducer.mode Image.reduceRegion Reducer.or Reducer.count Image.reduceRegions Reducer.countEvery Reducer.percentile Image.reduceToVectors Reducer.histogram Reducer.product ImageCollection.reduce Reducer.intervalMean Reducer.sampleStdDev FeatureCollection.reduceColumns Reducer.linearFit Reducer.sampleVariance Reducer.linearRegression Reducer.stdDev FeatureCollection.ReduceToImage Reducer.max Reducer.sum Reducer.mean Reducer.toCollection Reducer.median Reducer.toList Reducer.variance Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Reduce Bands Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Reduce Image Collection Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Reduce Neighborhood Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Reduce Region Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Reduce Regions Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Band Math + = Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Band Math + = + = Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  A Computation Example Landsat Image Collection Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  A Computation Example Filter by date Landsat Image 2001-01-01 Collection 2003-12-31 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  A Computation Example Surface reflectance Filter by date Landsat Image 2001-01-01 Image Collection 2003-12-31 Metadata Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  A Computation Example Add NDVI band Surface reflectance Filter by date Landsat Image 2001-01-01 Image NIR Band Collection 2003-12-31 Metadata Red Band Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  A Computation Example Composite: max(NDVI) Add NDVI band Surface reflectance Filter by date Landsat Image 2001-01-01 Image NIR Band Index Band: Collection 2003-12-31 Metadata Red Band NDVI Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  A Computation Example RGB Visualization Composite: max(NDVI) Add NDVI band Surface reflectance Filter by date Landsat Image 2001-01-01 Image NIR Band Index Band: Bands: B4, B3, B2 Collection 2003-12-31 Metadata Red Band NDVI min: 0.3 max: 0.8 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  A Computation Example Map: Reproject to Mercator RGB Visualization Composite: max(NDVI) Add NDVI band Surface reflectance Filter by date Landsat Image 2001-01-01 Image NIR Band Index Band: Bands: B4, B3, B2 Collection EPSG:3785 2003-12-31 Metadata Red Band NDVI min: 0.3 Scale: 30m max: 0.8 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Tiling Images are tiled during ingestion Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Tiling Images are tiled during ingestion Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Tiling Images are tiled during ingestion Downsampled by averaging Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Tiling Images are tiled during ingestion Downsampled by averaging Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Tiling Images are tiled during ingestion Downsampled by averaging During computation Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Tiling Images are tiled during ingestion Downsampled by averaging During computation Compute output tiles Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Tiling Images are tiled during ingestion Downsampled by averaging During computation Compute output tiles Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Tiling Images are tiled during ingestion Downsampled by averaging During computation Compute output tiles Find intersecting source tiles Reproject into the output projection Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  Running a Computation Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at  ➔ Exercise – Google Earth Engine I Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 166  ➔ Exercise – Google Earth Engine II Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 167  EO Data Cubes Big Earth Data Concepts Winter 2024/2025 | Martin Sudmanns & Dirk Tiede Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 168 Big EO data context Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 169 Big EO data context Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 170 Big EO data context Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 171 History of data access 1. Tape by mail 2. CD by mail 3. DVD by mail 4. Download over Internet Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 174 History of data access „Zero download model“ 1. Tape by mail 2. CD by mail 3. DVD by mail 4. Download over Internet 5. Data do not move: “Bring user to the data, not data to the user” Sudmanns, M., Tiede, D., Lang, S., Bergstedt, H., Trost, G., Augustin, H., Baraldi, A. and Blaschke, T., 2019. Big Earth data: disruptive changes in Earth observation data management and analysis?. International Journal of Digital Earth, pp.1-19. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 175 History of data access „Zero download model“ 1. Tape by mail 2. CD by mail 3. DVD by mail 4. Download over Internet 5. Data do not move: “Bring user to the data, not data to the user” Sudmanns, M., Tiede, D., Lang, S., Bergstedt, H., Trost, G., Augustin, H., Baraldi, A. and Blaschke, T., 2019. Big Earth data: disruptive changes in Earth observation data management and analysis?. International Journal of Digital Earth, pp.1-19. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 176 History of data access „Zero download model“ 1. Tape by mail 2. CD by mail 3. DVD by mail 4. Download over Internet 5. Data do not move: “Bring user to the data, not data to the user” Implementing this paradigm: Question is how?Sudmanns, M., Tiede, D., Lang, S., Bergstedt, H., Trost, G., Augustin, H., Baraldi, A. and Blaschke, T., 2019. Big Earth data: disruptive changes in Earth observation data management and analysis?. International Journal of Digital Earth, pp.1-19. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 177 Thinking in observations instead of images Which types of data do you see? What do they have in common? Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 178  Our approach Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Earth Observation Data Cubes https://www.lexcube.org/ Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 181 Thinking in observations instead of images Herman Minkovski gave a speech in 1906 Leipzig, Germany about the nature of space and time, which can be freely accessed (in german: https://de.wikisource.org/wiki/Raum_und_Zeit_(Mi nkowski)). He defines a point in a 3D-space and 1D- time as a "world-point" (Weltpunkt) in a reference system "world" (Welt). This serves as an axiom, the smallest observable unit. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 182 Thinking in observations instead of images “[r]eality is […] a three-dimensional space, which we must examine from three different points of view in order to comprehend the whole […] From one point of view we see the relations of similar things, from the second the development in time, from the third the arrangement and division in space” Hettner 1927, translated by Hartshorne 1939 Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 183 Thinking in observations instead of images Time Location fix A, control B, observe C Theme David Sinton. The inherent structure of information as a constraint to analysis: Mapped thematic data as a case study. Harvard papers on geographic information systems, 6:1–17, 1978. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 184 Thinking in observations instead of images Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 185 Thinking in observations instead of images New approach: From scenes to observations in space-time Executing the algorithm on a cloud infrastructure over the Internet No temporal snapshot, it is the result of continuous monitoring Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 186 Earth Observation Data Cubes Based on material by Peter Baumann Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 187 Earth Observation Data Cubes How do you store your pictures from your last vacation? How does Google Earth store the images? Based on material by Peter Baumann Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 188 Earth Observation Data Cubes Temporally stacked EO images, either as view or as physical data structure. Usually coupled with analysis-ready data (ARD). Main goal: Abstracting data storage from users: Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 189 Earth Observation Data Cubes File-path-based access to images Images Data cubes: No Aquisitions, raw data stream images; at least one non-spatial axis Observations in space-time http://desktop.arcgis.com/en/arcmap/10.3/tools/space-time-pattern-mining-toolbox/learnmorecreatecube.htm Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 190 Data Cubes: Querying across dimensions https://www.holistics.io/blog/the-rise-and-fall-of-the-olap-cube/ Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 191 OLAP Data Cubes in a data warehouse environment Stolte, C., Tang, D., & Hanrahan, P. (2003). Multiscale visualization using data cubes. IEEE Transactions on Visualization and Computer Graphics, 9(2), 176-187. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 192 OLAP Data Cubes in a data warehouse environment https://de.wikipedia.org/wiki/Sternschema#/media/Datei:Star_Schema.png https://www.tutorialspoint.com/dwh/dwh_olap.htm Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 193 Earth Observation Data Cubes Store data query-optimised not acquisition-oriented From: http://www.swissdatacube.org/ Different access methods (API, query language) Data cubes as infrastructure Provide a logical view on the data 1. Index external files 2. data as multi-dimensional array Tiede, Dirk; Baraldi, Andrea; Sudmanns, Martin; Belgiu, Mariana; Lang, Stefan (2017): Architecture and prototypical implementation of a semantic querying system for big Earth observation image bases. In European journal of remote sensing 50 (1), pp. 452–463. DOI: 10.1080/22797254.2017.1357432. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 194 Earth Observation Data Cubes Improve viewing data … How do you store your pictures from your last vacation? How does Google Earth store the images? Based on presentation by Peter Baumann Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 195 Earth Observation Data Cubes In OGC terminology: Coverages Example implementation: Rasdaman Based on presentation by Peter Baumann Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 196 Earth Observation Data Cubes Data storage characteristics independent from access methods (clients) Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 197 Earth Observation Data Cubes Example: Data cube of Austria 1 | import datacube 2 | 3 | query = { 4 | 'time': ('2017-01-01', '2018-01-01'), 5 | 'lat': (47.9, 47.6), 6 | 'lon': (12.8, 13.1), 7 | } 8 | 9 | dc = datacube.Datacube() 10| data = dc.load(product='sentinel-2', measurements=['b4', 'b3', 'b2'], **query) The image archive becomes a python object, directly to use Users don’t care about the location of the data and their physical storage Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 198 Data Cube Manifesto A definition of a data cube “A datacube is a massive multi-dimensional array, also called “raster data” or “gridded data”; “massive” entails that we talk about sizes significantly beyond the main memory resources of the server hardware. Data values, all of the same data type, sit at grid points as defined by the d axes of the d dimensional datacube. Coordinates along these axes allow addressing data values unambiguously. A d-dimensional grid is characterized by the fact that each inner grid point has exactly two neighbors along each direction; border grid points have just one. Point clouds, e.g., are not grids.” http://www.earthserver.eu/tech/datacube-manifesto Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 199 Earth Observation Data Cubes Store data query-optimised not acquisition-oriented From: http://www.swissdatacube.org/ Different access methods (API, query language) Data cubes as infrastructure Provide a logical view on the data 1. Index external files 2. data as multi-dimensional array Tiede, Dirk; Baraldi, Andrea; Sudmanns, Martin; Belgiu, Mariana; Lang, Stefan (2017): Architecture and prototypical implementation of a semantic querying system for big Earth observation image bases. In European journal of remote sensing 50 (1), pp. 452–463. DOI: 10.1080/22797254.2017.1357432. Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 200 Earth Observation Data Cubes … and storing data … Salvador Dali (1960): "A Propos of the 'Treatise on Cubic Form' by Juan de Herrera" Operations in a 3D array to trim or slice a data cube Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 201 Earth Observation Data Cubes Space first vs. time first … and storing data … With time series analysis, the reading “direction” is opposite to the storing “direction” with images Reading (n) >> Acquisition (1) Example: 100 images with 120 Mio. pixel / image → reading 1.2 Bil. pixel, even for a 100 pixel time series Digital Earth: Big Earth Data Concepts | Winter 2024/2025 | Martin Sudmanns & Dirk Tiede| {firstname}.{lastname}@plus.ac.at Slide 202 Earth Observation Data Cubes Space first vs. time first … and storing data … There is no optimal solution*, only trade-offs Stacked images: WaterML, Time series, e.g. NASA GeoServer, most WebGIS data rods applications *which serves all applica