BEGIN:VCALENDAR VERSION:2.0 PRODID:-//pretalx//pretalx.earthmonitor.org// BEGIN:VTIMEZONE TZID:Europe/Amsterdam BEGIN:STANDARD DTSTART:20001029T030000 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10 TZNAME:CET TZOFFSETFROM:+0200 TZOFFSETTO:+0100 END:STANDARD BEGIN:DAYLIGHT DTSTART:20000326T020000 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3 TZNAME:CEST TZOFFSETFROM:+0100 TZOFFSETTO:+0200 END:DAYLIGHT END:VTIMEZONE BEGIN:VEVENT UID:pretalx-global-workshop-2026-HWVSXP@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T100000 DTEND;TZID=Europe/Amsterdam:20261007T103000 DESCRIPTION:Keynote Lecture by Matteo Mattiuzzi. Abstract and title to be p rovided. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Keynote - Matteo Mattiuzzi URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/HWVSXP/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-USFXYN@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T103000 DTEND;TZID=Europe/Amsterdam:20261007T110000 DESCRIPTION:The creatures of our world depend on forests - living structure s that provide habitat\, food\, and water. We draw meaning from trees that shape our perspectives about nature. Now\, with access to unprecedented t echnology\, we can understand more complex aspects of tree structure - the architectural form - that were previously hidden. Accurately quantifying forest structure is crucial for assessing climate change mitigating strate gies and for guiding conservation efforts. \n\nHere\, we explore how we u se cutting-edge 3D laser mapping from below and above the canopy to unders tand trees and forest structure around the world. As part of the Global Te rrestrial Laser Scanning (GTLS\, global-tls.net) Database initiative\, we are collecting ultra-high resolution 3D structural data in forests in unpr ecedented detail - leveraging this rich dataset for updated tree-level sca ling\, architecture\, and biomass. To complement this work\, we are lookin g at the forest canopy from above at a global scale with the NASA / UMD Gl obal Ecosystem Dynamics Investigation (GEDI) to capture and investigate ve rtical structural signatures of different forests across the planet. \n\n We are now beginning to understand the dimensions of how a more comprehens ive understanding of tree and forest architecture has direct implications for accurate carbon accounting\, habitat mapping\, and biodiversity conser vation. Moving forward we will apply our newly developed 3D tree traits to inform structural characterizations of forests with GEDI\, while continui ng to fill data gaps by collecting ground-based laser scanning data at new sites around the world. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Understanding the 3D Signatures of Forests Across the Planet with O pen EO - Atticus Stovall URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/USFXYN/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-GVGR7V@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T110000 DTEND;TZID=Europe/Amsterdam:20261007T113000 DESCRIPTION:Healthy soils are an indispensable natural resource providing m any ecosystem services\, such as producing biomass to secure our food supp ly\, storing large amounts of carbon – a higher amount with respect to f orests worldwide – storing and purifying our drinking water\, and provid ing a habitat for a variety of organisms. At the same time\, soils are a l imited\, non-renewable and irreplaceable natural resource.\n\nIn this cont ext\, the EU Soil Strategy for 2030 outlines the path to healthy soils in Europe by 2050 through\nvoluntary and legislative measures by Member State s\, leading to the approval of the Soil Monitoring Law\nin 2025. Member St ates are carrying out a variety of activities to activate their existing e xpertise in soil\nmonitoring and promote technologies such as optical and SAR remote sensing\, to be used in this frame.\n\nCurrently\, there are se veral initiatives and projects that explore the potential of Earth Observa tion for soil\nmapping and monitoring for large areas. This will open the path to establish an operational service for soil information that could b e available for the public\, such as the ones provided in the frame of Cop ernicus Land Monitoring Service. In this talk\, we address the opportuniti es and challenges of mapping soils with\noptical Earth Observation includi ng spaceborne imaging spectroscopy and try to answer the following\nquesti ons: Which technological developments have been achieved in the last decad es? What steps are\nnecessary to establish a robust Earth observation-base d monitoring system for soils? And especially\, how can the imaging spectr oscopy community contribute to this process?\n\nThe talk presents the dema nd for soil-related information\, which\, depending on the application\, m ust fulfil various spatial and temporal requirements\, as well as a specif ic level of detail. One of the major\nchallenges is developing methods and techniques that can handle the heterogeneous regional\ncharacteristics of the landscape. We present examples of regionalized models for temporal ba re soil\ncompositing in Europe to be used as an important input data set ( Karlshoefer et al.\, 2025) and local\nensemble models for soil organic car bon estimation in Germany (Broeg et al.\, 2024). Another challenge\ninvolv es the coverage and repetition of imaging spectroscopy data\, as monitorin g soil erosion requires\nfrequent updates on vegetation coverage. In such cases\, using multispectral and hyperspectral data in\ncombination with de ep learning algorithms to obtain sub-pixel information about vegetation co ver (i.e.\,\nfractional vegetation cover) is promising (Schwind et al.\, 2 024). Finally\, estimating the accuracy and\nuncertainty of information pr oducts\, especially those covering large areas remains challenging. Often\ ,\nvalidation data is scarce and unsuitable for the accuracy assessments o f large areas. We discuss various\nstrategies of assessing accuracy\, such as producing pixel-wise uncertainty maps (Ochoa et al.\, 2025)\,\nevaluat ing the mapping methods itself (Karlshoefer et al.\, 2025) and developing UAV-based strategies\nwith high transferability potential.\n\nThe describe d strategies address the typical challenges of processing large areas\, su ch as countries or\ncontinents\, which include regional differences and da ta scarcity. It is crucial to expand the scientific scope in order to over come these challenges and provide frequent\, accurate and reliable soil da ta for\nextensive regions. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Opportunities and challenges of mapping soils with EO for large are as - Dr. Uta Heiden URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/GVGR7V/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-CNMX9F@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T113000 DTEND;TZID=Europe/Amsterdam:20261007T120000 DESCRIPTION:Keynote talk by Markus Reichstein. Abstract and title to be pro vided. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Keynote - Markus Reichstein URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/CNMX9F/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-GCCBPS@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T123000 DTEND;TZID=Europe/Amsterdam:20261007T130000 DESCRIPTION:Over the past five years\, the Government of Catalonia has made a firm and sustained commitment to developing a competitive and innovatio n‑driven space sector. This effort has combined Catalonia’s strong cap abilities in digital technologies\, scientific research and industrial eng ineering with coordinated collaboration across public institutions\, unive rsities\, research centres and private companies.\nAs part of this initiat ive\, several small satellite missions for communications and Earth observ ation\, such as the GENIOT and GENEO series\, have been deployed with supp orting ground infrastructure and data platforms. Initially conceived as te chnology demonstrators\, these missions have enabled the validation of ope rational capabilities and the development of innovative services in areas such as disaster risk management\, environmental monitoring and territoria l management among others.\nThis talk presents the role of small satellite s in enabling innovative use cases highlighting the supporting infrastruct ure\, service adoption and future developments under the Catalonia Space 2 030 Strategy\, approved in November 2025\, which foresees the deployment o f eight satellite missions by 2030 focused on Earth observation and advanc ed communications and resilience‑oriented applications. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:The role of small satellites in strengthening innovative use caes w ithin the framework of the Catalonia Space Strategy - Estefania Blanch URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/GCCBPS/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-HZRYT7@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T130000 DTEND;TZID=Europe/Amsterdam:20261007T133000 DESCRIPTION:Keynote talk by Johan van den Hoogen. Abstract and title to be provided DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Keynote - Johan van den Hoogen URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/HZRYT7/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-E9SDB8@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T133000 DTEND;TZID=Europe/Amsterdam:20261007T134500 DESCRIPTION:Statistical modeling and uncertainty analysis plays a critical role in evaluating climate and environmental data. Concepts such as standa rd error of the mean and design-based estimation seem to be increasingly u sed to manipulate prediction errors and tradable changes. Advanced trend e stimation and change-point models are essential for accurately identifying long-term shifts in essential climatic variables such as soil organic car bon and above ground biomass. Subtracting two above-ground biomass (AGB) m aps can create false data because map uncertainties propagate into the dif ference\, compounding the errors from both individual maps and inflating a pparent change signals. Rather than revealing true environmental dynamics\ , naive subtraction often produces an apparent "change" that is actually j ust statistical noise. Quantile Regression Random Forests (QRRF) offer a p owerful\, non-parametric approach to estimating the true distribution of e rrors by retaining all observations within the terminal leaf nodes of the forest\, rather than just calculating the conditional mean. This allows th e model to estimate the full conditional cumulative distribution function and extract specific percentiles to form prediction intervals. We demonstr ate how this method can be used to determine tradable carbon sequestration without taking additional risks. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Quantification of temporal changes in Earth-Observation-based estim ates: examples with soil carbon & above ground biomass - Tom Hengl (OpenGe oHub) URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/E9SDB8/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-TW9RF7@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T150000 DTEND;TZID=Europe/Amsterdam:20261007T151500 DESCRIPTION:Foundation models for Earth systems have advanced rapidly for w eather and climate prediction\, but remain largely confined to physical va riables\, omitting the human systems that drive emissions\, shape land use \, build infrastructure\, and mediate vulnerability. We argue that this ga p is fundamentally a data problem: the information exists but is fragmente d across incompatible grids\, projections\, temporal frequencies\, and for mats. We present two complementary contributions that address this challen ge.\nFirst\, WorldTensor is a harmonised global dataset that aligns over 7 50 environmental and socioeconomic variable families onto a common 0.25° latitude–longitude grid and annual temporal framework. It integrates cli mate\, emissions\, land use\, satellite vegetation indices\, gridded popu lation and GDP products\, power plant registries\, and natural hazard and conflict catalogues into a single ML-ready NetCDF corpus. Constructing Wor ldTensor required solving nontrivial harmonisation problems including regr idding across heterogeneous native resolutions\, rasterising point and vec tor datasets into spatially meaningful fields\, and reconciling temporal c overages spanning daily observations to sparse multiyear socioeconomic sna pshots. The dataset and processing code will be released under open licens es.\nSecond\, TerraNova is a foundation model designed to learn from World Tensor's multimodal structure. It combines coordinate-based spatial encodi ng\, learned country-level embeddings\, Fourier temporal encoding\, and a hypernetwork decoder to jointly predict climate\, land surface\, socioecon omic\, and infrastructure variables in a unified multi-task framework. Ear ly results demonstrate successful learning across multiple heterogeneous E arth system tasks simultaneously\, validating that foundation models can l earn shared representations across the coupled human–Earth system.\nToge ther\, WorldTensor and TerraNova provide an open\, end-to-end pipeline fro m harmonised planetary data to multimodal foundation model training\, supp orting applications in climate impact assessment\, cross-domain pattern di scovery\, and evidence-based environmental policy. DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:WorldTensor and TerraNova: Open Data and Foundation Models for the Coupled Human–Earth System - Carlos Rodriguez-Pardo URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/TW9RF7/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-K77QLY@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T150000 DTEND;TZID=Europe/Amsterdam:20261007T151500 DESCRIPTION:The growing availability of global aboveground biomass (AGB) ma ps from Earth Observation (EO) is changing how carbon stocks can be quanti fied\, monitored and reported. Rapid advances in EO\, cloud computing and GeoAI have expanded the range of available products\, from coarse-resoluti on long time series to emerging global maps at up to 10 m resolution. At t he same time\, inconsistencies in spatial support\, temporal coverage\, mo deling approach and uncertainty structure continue to limit comparability and reduce confidence in their use for carbon accounting\, climate reporti ng\, REDD+ and other policy-facing applications. What is increasingly need ed is not simply more biomass maps\, but a framework that can validate the m consistently\, explain where they differ and clarify what those differen ces mean for actual use.\nThis contribution presents an integrated framewo rk developed within the Open-Earth-Monitor ecosystem with four connected c omponents: (1) a harmonized global biomass reference dataset\, AGBref\; (2 ) a validation and estimation framework that explicitly accounts for spati al uncertainty and representativeness\; (3) a systematic inter-comparison of global AGB maps across methods\, resolutions and epochs\; and (4) demon strations of how product differences affect downstream uptake. A key novel ty is the use of AGBref across all components. AGBref combines National Fo rest Inventories\, permanent plots and airborne LiDAR-derived biomass maps in a multi-epoch\, multi-resolution reference system with uncertainty inf ormation\, providing a common backbone for independent validation and more transparent interpretation of biomass products. The framework moves beyon d validation based only on global summary statistics. In addition to agree ment with reference data\, it examines how biomass products represent spat ial heterogeneity and landscape structure. This is particularly important with the emergence of very high-resolution biomass maps\, which may show s imilar overall accuracy but still differ substantially in the spatial patt erns they reproduce. \nUse cases are central to the framework. It responds to a clear demand for biomass information that is not only more accurate\ , but also more comparable\, operational and easier to integrate into exis ting analytical systems. For example\, WRI identifies the best available b iomass dataset that can strengthen forest carbon stock and emissions asses sment\, support biomass change analysis\, and remain compatible with Globa l Forest Watch workflows and baseline forest change products. For OECD\, t he need is similar but framed through environmental indicators\, LULUCF-re lated analysis\, and SEEA-based accounting\, where one consistent and tran sparent dataset is preferred over multiple competing products. In both cas es\, independent validation\, comparability with national data\, open acce ss\, interoperability and regular updates are core conditions for uptake.\ nThe framework is therefore designed not only to compare maps\, but also t o test their implications for reporting and accounting contexts. One prior ity application is carbon accounting across overlapping but distinct frame works such as UNFCCC reporting and SEEA-based environmental accounting. Th ese frameworks share a need for spatially explicit\, transparent and compa rable biomass information\, yet differ in accounting logic\, reporting pur pose\, and treatment of stocks and changes. The framework creates a basis for examining how the same EO-based biomass product performs across these contexts\, where comparability holds\, and where important differences eme rge. This is especially relevant for countries with limited or infrequent National Forest Inventory data\, and for recurrent accounting processes th at require methods and datasets that can be updated regularly and consiste ntly through time.\nTo support uptake\, the framework is implemented throu gh open\, cloud-based tools such as Plot2Map within ESA-MAAP\, enabling re producible integration of plot-level reference data with large-scale EO pr oducts for validation\, visualization and comparison. The platform also se rves as a demonstration space for testing how biomass products can support institutional needs in global forest assessment\, environmental indicator s\, policy analysis\, SEEA and LULUCF applications\, and country-facing mo nitoring workflows. This is particularly relevant for users such as WRI\, OECD and national agencies that require transparent\, scalable\, open and regularly updateable biomass information. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Bridging Data\, Methods and User-Uptake in Global Biomass Mapping: An Open Framework for Validation\, Estimation and Inter-Comparison - Arnan Araza\, Martin Herold URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/K77QLY/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-TURRVV@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T151500 DTEND;TZID=Europe/Amsterdam:20261007T153000 DESCRIPTION:Accurate characterization of tropical forest vertical structure is critical for carbon accounting and ecosystem monitoring\, yet most mac hine-learning pipelines reduce GEDI's rich waveform information to a singl e scalar\, typically canopy height or a high relative-height percentile. T his simplification discards the ordered height distribution that GEDI enco des across its full relative height (RH) profile\, and that its own biomas s algorithms depend on. We introduce Biomazon\, an open\, ML-ready multimo dal benchmark dataset at 20 m resolution over the Amazon Basin\, designed to support joint prediction of the full GEDI RH profile (RH0 to RH100) tog ether with above-ground biomass density (AGBD). The dataset pairs GEDI-der ived targets with multi-sensor predictors including Sentinel-1\, Sentinel- 2\, ALOS-2 PALSAR-2\, Copernicus DEM\, Dynamic World land cover\, and geos patial foundation model embeddings\, all co-registered on a common grid wi th standardized spatial splits and evaluation protocols to enable reproduc ible comparison of methods. We formulate RH prediction as structured outpu t learning with a monotonicity constraint that enforces physical consisten cy across percentiles\, and we provide baseline results from systematic ab lations over model scale\, sensor contributions\, and the role of AlphaEar th embeddings\, both as standalone predictors and in fusion with raw modal ities. Results are contextualized against existing gridded products to ass ess practical relevance. Biomazon addresses a gap in current benchmarking by shifting the task formulation from scalar regression toward structure-a ware modeling\, and by providing the community with an open\, multi-sensor dataset and protocol for investigating when and how different data source s\, including learned representations\, contribute to forest structure and biomass retrieval in tropical forests. DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:Biomazon: A Multimodal Benchmark for Full Vertical Structure and Bi omass Modeling in the Amazon Basin - Sayan Mandal\, Rocco Sedona URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/TURRVV/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-7DU3RA@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T151500 DTEND;TZID=Europe/Amsterdam:20261007T153000 DESCRIPTION:Semi-natural grasslands are critical ecosystems that provide a range of essential services\, with their role as habitats for diverse spec ies being among the most significant. However\, over the past century\, se mi-natural grasslands that once covered vast areas across Europe have larg ely been transformed into intensively managed agricultural lands\, abandon ed\, or converted into forests. These large-scale land-use changes have le d to considerable biodiversity loss\, making the conservation and restorat ion of semi-natural grasslands an important component of sustainable lands cape management. \n\nWe utilized 97 in-situ herbaceous biomass samples col lected during the summer of 2019 from alvar grasslands in Western Estonia\ , all restored between 2015 and 2019. Samples were collected from 20 × 20 cm plots nested within 2 × 2 m botanical plots. Sentinel-1 and Sentinel- 2 imagery from the same period was used\, with median band values and deri ved indices (e.g.\, NDVI\, BSI\, SAVI\, VH/VV) included as predictors. \n\ nRandom Forest models were developed using Sentinel-1 and Sentinel-2 spect ral bands and derived indices as predictors. Model robustness was evaluate d using 5-fold cross-validation. Two approaches for linking field and sate llite data were tested: point sampling and a 3 × 3 kernel mean\, with poi nt sampling performing slightly better. \n\nThe model achieved an RMSE of 98 ± 54 g/m²\, an MAE of 71 ± 30 g/m²\, and an R² of 0.32 ± 0.08\, r eflecting the high spatial variability of semi-natural grasslands. SHAP an alysis identified SAVI\, NDVI\, and the vegetation red edge band B8A as th e most important predictors\, while Sentinel-1 variables contributed less to model performance. \n\nThese results highlight the dominant role of opt ical data in herbaceous biomass estimation and demonstrate that simple poi nt-based sampling can outperform spatial averaging approaches. The propose d methodology provides a practical and scalable solution for monitoring gr assland restoration. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Monitoring herbaceous biomass and restoration of semi-natural grass lands using machine learning on Sentinel-1 and Sentinel-2 imagery - Iris L uik URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/7DU3RA/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-9WB3A7@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T153000 DTEND;TZID=Europe/Amsterdam:20261007T154500 DESCRIPTION:Vegetation in terrestrial ecosystems plays a key role in the ca rbon cycle\, and understanding its spatiotemporal patterns and associated drivers is crucial for ecological research. This study explores the relati ons between remote sensing vegetation Gross Primary Production (GPP) and c limate explanatory variables such as the Standardized Precipitation Evapot ranspiration Index (SPEI) and soil moisture anomalies (SMA). \n\nThe study focused on the climatically diverse Ebro River basin (85\,600 km²)\, Spa in's river largest catchment\, using monthly data from 2016 to 2024. The a rea is bounded between the three meteorological domains of this region of SW Europe: Atlantic\, European continental and Mediterranean. \n\nDuring t he processing phase\, harmonized monthly products at 1 km spatial resoluti on were generated from multiple satellite and in-situ sources. GPP was agg regated from the MOD17A2HGF product\, SPEI was derived in-situ meteorologi cal data (Trypidaki et al 2024) by AEMET\, and monthly SMA were computed f rom Sentinel-1 synthetic aperture radar (SAR) data using a dual-polarizati on algorithm (DPA) (Fan et al. 2025).\n\nWe explore vegetation–climate r elationships using correlation and GeoAI ML approaches\, including Random Forest (caret R package) and Accumulated Local Effects (ALEPlots R package ) between GPP and climate variables. Model stability and variable importan ce were evaluated using multiple metrics.\n\nOur findings highlight the po tential\, requirements and limitations of GeoAI tools compared to classica l statistical methods\, in handling nonlinear relationships and multicolli nearity.\n\nReferences: \n\nFan\, D.\, Zhao\, T.\, Jiang\, X.\, García-Ga rcía\, A.\, Schmidt\, T.\, Samaniego\, L.\, Attinger\, S.\, Wu\, H.\, Jia ng\, Y.\, Shi\, J.\, Fan\, L.\, Tang\, B.-H.\, Wagner\, W.\, Dorigo\, W.\, Gruber\, A.\, Mattia\, F.\, Balenzano\, A.\, Brocca\, L.\, Jagdhuber\, T. \, … Peng\, J. (2025). A Sentinel-1 SAR-based global 1-km resolution soi l moisture data product: Algorithm and preliminary assessment. Remote Sens ing of Environment\, 318\, 114579. https://doi.org/10.1016/j.rse.2024.1145 79\n\nTrypidaki E.\, Pesquer L.\, Domingo-Marimon C\, "Spatiotemporal Anal ysis for Enhanced Drought Monitoring and Agricultural Applications in the Ebro Basin\, Spain\," 2024 IEEE International Workshop on Metrology for Ag riculture and Forestry (MetroAgriFor)\, Padua\, Italy\, 2024\, pp. 603-608 \, https://doi.org/10.1109/MetroAgriFor63043.2024.10948835 DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Understanding Vegetation–Climate Relationships Using GeoAI: A Spa tiotemporal Analysis in the Ebro River Basin - Eirini Trypidaki URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/9WB3A7/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-MZEPKT@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T153500 DTEND;TZID=Europe/Amsterdam:20261007T155000 DESCRIPTION:Discrete Global Grid Systems (DGGS) tessellate the earth’s su rface into zones of equal area and very similar shape\, minimizing spatial distortions in geospatial data processing. Here we present DGGS.jl\, a to ol to create and visualize data cubes with DGGS coordinates. We applied it to create such a data cube of Europe from quarterly cloudless Sentinel-2 mosaics of the Copernicus Data Space Ecosystem. DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:From Sentinel-2 STAC to DGGS Native Data Cubes with DGGS.jl - Danie l Loos URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/MZEPKT/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-UMRHWT@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T154500 DTEND;TZID=Europe/Amsterdam:20261007T160000 DESCRIPTION:In alignment with the European Green Deal’s strategies\, fore st monitoring is crucial to fill the existing information gaps and create a comprehensive forest knowledge base. Based on the 4th Spanish National Forest Inventory collections updated in 2017\, we assessed remote sensing data derived from various sources as complementary information to support forest ecosystem monitoring in Catalonia over the last decade. Our study areas focused on regional natural parks designated by Natura 2000 where de ciduous tree species (Fagus sylvatica\, Castanea sativa) are well represen ted to analyse key biophysical variables known as Essential Biodiversity V ariables (EBVs) such as LAI and FAPAR. The data products compared in this study include high-resolution vegetation maps with new algorithms provide d through cloud-based platforms such as Copernicus Data Space Ecosystem an d Google Earth Engine. Specifically\, we referenced Sentinel-2 based EBV s from BioPAR by VITO and World Reforestation Monitor by ETH Zurich. In di scussion our challenges and opportunities associated with data interoperab ility and quality are addressed. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Key biophysical variables for forest monitoring in Catalonia - Kaor i Otsu\, Imma Serra URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/UMRHWT/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-CDLLLB@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T160000 DTEND;TZID=Europe/Amsterdam:20261007T161500 DESCRIPTION:This study explores the application of high-dimensional embeddi ngs derived from Sentinel-2 imagery for automated anomaly detection in env ironmental monitoring. By utilizing the SSL4EO self-supervised learning fr amework\, we transform raw satellite data into compact\, informative repre sentations that capture essential spatial and temporal features. The entir e workflow is integrated within the Copernicus Data Space Ecosystem (CDSE) \, ensuring efficient access to large-scale\, analysis-ready archives and enabling rapid processing of planetary-scale datasets. \n\nThe core of our approach lies in leveraging SSL4EO to bypass the need for massive labeled datasets\, creating a standardized intermediate format that bridges raw i magery with advanced AI tasks. These embeddings serve as the foundation fo r an anomaly detection pipeline designed to pinpoint deviations from expec ted seasonal or spatial trends\, such as flooding\, wildfires\, or shifts in vegetation health. To ensure interoperability and ease of use in geospa tial analytics\, results are stored in the GeoParquet format\, which suppo rts both reproducibility and high-performance data handling. \n\nTo confir m the framework's robustness\, we conducted extensive validation across di verse geographical regions and seasonal cycles\, including challenging win ter conditions with snow cover and low solar illumination. The pipeline de monstrated high resilience\, producing consistent embeddings even in the p resence of partial cloud cover. Furthermore\, we evaluated the system’s portability across heterogeneous computing environments. Testing on the CR EODIAS cloud platform (using both CPU and GPU nodes) alongside high-perfor mance computing (HPC) infrastructures like EOHPC and SpaceHPC proved that the solution scales effectively and maintains functional integrity across different hardware architectures. \n\nThe results indicate that combining self-supervised embeddings with anomaly detection creates a powerful tool for environmental intelligence. The proposed framework is suitable for a w ide range of operational applications\, from tracking urban growth to moni toring climate-induced environmental changes. By providing a portable and scalable bridge between raw Copernicus data and actionable insights\, this study highlights the transformative potential of GeoAI in supporting glob al and regional decision-making processes. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Satellite-Based Anomaly Detection using GeoAI Embeddings: A Scalabl e Workflow - Marcin Kluczek URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/CDLLLB/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-QF38UE@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T161500 DTEND;TZID=Europe/Amsterdam:20261007T163000 DESCRIPTION:Abstract\nFlash droughts are increasingly recognized as a disti nct class of hydroclimatic extremes marked by rapid soil-moisture depletio n over only a few weeks. Because of their abrupt onset\, these events can severely affect agricultural production\, terrestrial ecosystems\, and reg ional water resources before conventional drought indicators fully capture their development. Recent drought episodes across Europe have reinforced the need to better understand how flash drought behaviour may evolve under continued climate warming. Yet substantial uncertainty remains\, partly b ecause many previous assessments rely on meteorological indicators or sing le datasets that do not adequately represent the subsurface soil-moisture processes governing rapid drought emergence.\nThis study proposes to inves tigate the future evolution of flash drought characteristics across Europe using root-zone soil moisture simulations from a multi-model ensemble of global hydrological models\, including WaterGAP\, H08\, and CWatM\, forced by bias-adjusted CMIP6 climate projections. Daily soil-moisture outputs w ill be aggregated to pentad scale (5-day averages) to reduce high-volatili ty while preserving the rapid depletion signals associated with flash drou ght onset. Root-zone soil moisture will then be transformed into grid-spec ific climatological percentiles\, enabling a temporally and spatially cons istent identification of flash drought events across models and time perio ds. A key methodological contribution of this study is the integration of pentad-scale root-zone soil moisture percentiles with a multi-model hydrol ogical ensemble\, enabling a consistent and process-relevant assessment of future flash drought dynamics across Europe.\nUsing this framework\, the study will assess potential changes in flash drought frequency\, duration\ , severity\, and onset speed under historical conditions and future projec tions for SSP1-2.6\, SSP2-4.5\, and SSP5-8.5. The analysis is intended to advance understanding of future rapid drought development and to support d rought risk assessment\, climate adaptation planning\, and early-warning s trategies across Europe.\n\n\n\nKeywords: Flash drought\; Root-zone soil m oisture\; Global hydrological models\; CMIP6\; Europe. DTSTAMP:20260625T073855Z LOCATION:Room 18 SUMMARY:Emerging Flash Drought Risk across Europe: Insights from Multi-Mode l Root-Zone Soil Moisture Projections - VAIBHAV KUMAR URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/QF38UE/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-7GBGPT@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T163000 DTEND;TZID=Europe/Amsterdam:20261007T164500 DESCRIPTION:Extreme hydrometeorological extremes are one of the main focuse s of operational early warning systems for natural hazards. The ongoing in tegration of remote sensing datasets into the monitoring pipelines is aime d at contributing to the refinement of the forecasts and the accurate iden tification of the risks. However\, very few studies have specifically addr essed the inherent uncertainties of the remote sensing datasets in the ran ge of extreme events. Multiple factors in the processing of these datasets can impact the capabilities of each type of data to effectively detect po tentially hazardous events due to unrealistic recognition of the tails of the distribution of events. \n \nThis study is devoted to the intercompari son of remote sensing\, model-based and reanalysis products of key variabl es of the water cycle (rain\, soil moisture\, flow) to evaluate the consis tency of common current operational products for the portrayal of extreme events. The procedure comprises specific extreme value analysis of the dis tributions of the datasets with special attention to the characterisation of the magnitude and temporal dimensions of the events. In this way\, metr ics of frequency\, duration and intensity are applied to assess the suitab ility of each product for proper extremes identification against ancillary data of multiple events of well-known impact. \n\nThe results indicate re levant differences among products well before the range of true extreme ev ents\, which partly explains the struggle of current operational monitorin g systems to accurately characterise impactful events. Discussion on the f actors influencing such notable differences in the products apprise of mul tiple aspects of datasets generation and handling that led to distorted ca pabilities in the tail range of the distributions that need review and coo rdination between the actors in charge of the generation and application o f datasets. \n\nThe study encourages further attention to the evaluation o f data in the range of their most relevant application\, risk assessment\, in order to avoid undesired inherited constraints to their application\, jeopardising the confidence in early warning systems or the remotely–sen sed data. DTSTAMP:20260625T073855Z LOCATION:Room 18 SUMMARY:Limitations of current operational systems based on remote sensing and models for the characterization fo extreme hydrometeorological events - Jaime Gaona URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/7GBGPT/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-WHG977@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T164500 DTEND;TZID=Europe/Amsterdam:20261007T170000 DESCRIPTION:Robert N Masolele1\, Katja Berger2\, Zoltan Szantoi3\, Camilo Z amora2\, Johannes Reiche1\n\n1 Wageningen University\, Wageningen\, The Ne therlands\; robert.masolele@wur.nl\n2 GFZ\, German GeoResearch Center Pots dam\, Germany\n3 Directorate of Earth Observation Programmes\, European Sp ace\nAgency (ESA)\, Frascati\, RM\, Italy\n\nCoffee cultivation underpins agricultural economies worldwide\, supporting millions of livelihoods and contributing significantly to global production [1]. At the same time\, co ffee is among the leading commodities associated with global deforestation risks linked to European Union (EU) consumption. However\, accurately map ping coffee farm locations remains challenging due to the heterogeneous la ndscapes in which coffee is grown\, including dense vegetation\, diverse l and cover types\, varying management practices\, and phenological stages [ 2]\, [3]\, [4]. Existing mapping efforts are largely limited to major prod ucers such as Brazil\, Vietnam\, Ethiopia\, and Colombia\, leaving substan tial gaps across other coffee-growing regions [5].\nTo address this\, we f irst present a global benchmarking framework for commodity crop mapping. W e evaluate a combination of Sentinel-1 and Sentinel-2 data\, alongside loc ational variables. Using a comprehensive reference dataset spanning >40 co ffee-producing countries\, we show that models integrating Sentinel-1 and Sentinel-2 data with location encoding achieve the highest performance (F1 -score: 89%)\, outperforming models without contextual information [4].\nB uilding on this\, we apply the best-performing deep learning framework to generate the first high-resolution global map of coffee farm extent\, achi eving an F1-score of 86%. The integration of Sentinel-1 (radar) and Sentin el-2 (optical) data enables robust feature extraction across diverse condi tions\, while location encodings enhance geographic contextualization of c offee systems.\nThis work delivers a consistent\, high-resolution global c offee map\, supporting sustainable land management\, supply chain transpar ency\, and conservation in tropical regions. It directly aligns with the E U Deforestation Regulation (EUDR\, Regulation (EU) 2023/1115)\, which requ ires monitoring the deforestation footprint of seven key commodities\, inc luding coffee relative to the December 31\, 2020 cut-off date. The approac h is being operationalized within cloud-based platforms (e.g.\, Copernicus Data Space Ecosystem)\, facilitating access for policymakers\, certificat ion bodies\, and stakeholders.\n\n[1] R. Grüter\, T. Trachsel\, P. Laube\ , and I. Jaisli\, ‘Expected global suitability of coffee\, cashew and av ocado due to climate change’\, PLoS One\, vol. 17\, no. 1\, p. e0261976\ , Jan. 2022\, doi: 10.1371/JOURNAL.PONE.0261976.\n[2] D. A. Hunt et al.\, ‘Review of Remote Sensing Methods to Map Coffee Production Systems’\, Remote Sensing 2020\, Vol. 12\, Page 2041\, vol. 12\, no. 12\, p. 2041\, J un. 2020\, doi: 10.3390/RS12122041.\n[3] G. Maskell\, A. Chemura\, H. Nguy en\, C. Gornott\, and P. Mondal\, ‘Integration of Sentinel optical and r adar data for mapping smallholder coffee production systems in Vietnam’\ , Remote Sens. Environ.\, vol. 266\, Dec. 2021\, doi: 10.1016/j.rse.2021.1 12709.\n[4] R. N. Masolele et al.\, ‘Mapping the diversity of land uses following deforestation across Africa’\, Sci. Rep.\, vol. 14\, p. 1681\, 2024\, doi: 10.1038/s41598-024-52138-9.\n[5] A. Escobar-López\, M. Á. C astillo-Santiago\, J. F. Mas\, J. L. Hernández-Stefanoni\, and J. O. Lóp ez-Martínez\, ‘Identification of coffee agroforestry systems using remo te sensing data: a review of methods and sensor data’\, Geocarto Int.\, vol. 39\, no. 1\, p. 2297555\, 2024\, doi: 10.1080/10106049.2023.2297555\; WGROUP:STRING:PUBLICATION. DTSTAMP:20260625T073855Z LOCATION:Room 18 SUMMARY:High-Resolution Global Maps of Coffee Farms Extent - Robert Masolel e URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/WHG977/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-PEKSXV@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T164500 DTEND;TZID=Europe/Amsterdam:20261007T170000 DESCRIPTION:The organic carbon flux entering the pedosphere through forest litterfall is a critical indicator of forest ecosystem functioning and a p rimary driver of soil respiration (RS). However\, accurately quantifying l itterfall spatiotemporal dynamics at the global scale remains a major chal lenge due to the scarcity of high-resolution Earth Observation (EO) framew orks coupled with extensive ground observations. \nHere\, we present a nov el GeoAI-driven approach that synthesizes 14\,912 in-situ observations acr oss 843 sites globally with multi-source remote sensing data. By leveragin g machine learning algorithms\, we decoupled complex biogeochemical mechan isms and generated a 500-m spatial resolution global forest litterfall pro duct. Furthermore\, we integrated these high-resolution EO derivatives int o an Olson legacy model to quantify the impact of litterfall on RS across different forest biomes. \nOur results reveal significant spatial heteroge neity in biogeochemical coupling\, highlighting asymmetric microbial respo nses between tropical forests (characterized by high turnover rates) and t emperate/boreal forests (exhibiting biogeochemical inertia). This study de monstrates the profound potential of integrating open Earth Observation da ta and machine learning to monitor global forest dynamics. Our 500-m globa l product provides a vital\, scalable data infrastructure for next-generat ion Earth system models\, biodiversity conservation\, and forest carbon ma nagement. \n(Note: This research has been recently published in Remote Sen sing of Environment\, 2026\, https://doi.org/10.1016/j.rse.2026.115373) DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:Mapping Global Forest Litterfall Dynamics at 500-m Resolution via G eoAI: Implications for Forest Ecosystem Functioning and Soil Respiration - Chunsheng Wang URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/PEKSXV/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-NZJHKC@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T170000 DTEND;TZID=Europe/Amsterdam:20261007T171500 DESCRIPTION:The way we monitor soils\, water resources\, and agricultural l andscapes has been transformed by Earth Observation and environmental mode lling over the last few years. Nevertheless\, real decision-making remains a major challenge\, as turning complex datasets into tools is not yet tri vial for human purposes. \n\nIn this talk\, the audience will learn about lessons learned from designing platforms such as Soils Revealed\, Aqueduct \, Foodscapes\, and Landgriffon. These tools combine satellite indicators\ , global environmental datasets\, and workflows to help users better under stand soil degradation\, water risks\, and food system dynamics\, and to t ranslate environmental data into platforms that support decision-making ac ross scales\, from regional planning to global analysis. \n\nWe will also reflect on the challenges of moving from data to decision tools\, combinin g scientific outputs with user needs\, based on human-centered design to s upport exploration and comparison\, and ensuring that complex environmenta l information is both scientifically rigorous and accessible. By combining Earth observation\, modelling and interface design\, these platforms demo nstrate that environmental data is trustful acctionable knowledge to help inform land and water management. DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:From earth observation to farm decisions: Designing platforms for d ecision making - Sergio Estella URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/NZJHKC/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-HGTPJN@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T170000 DTEND;TZID=Europe/Amsterdam:20261007T171500 DESCRIPTION:While forest monitoring has reached high levels of maturity\, g rassland ecosystems remain a critical "blind spot" in global conservation. To address this\, Global Pasture Watch (GPW) has established a comprehens ive baseline using 30m multi-decadal datasets (2000–2022) covering grass land extent\, vegetation height\, and livestock density. However\, the inh erent heterogeneity and rapid seasonality of these landscapes present sign ificant current challenges for traditional pixel-based classification. To overcome these barriers\, our next steps involve transitioning to next-gen eration machine learning models that utilize Sentinel-2 spatial-temporal e mbeddings. By moving beyond simple spectral signatures to rich\, high-dime nsional latent representations\, we can better capture the nuances of mana ged vs. natural grasslands and monitor Gross Primary Productivity (GPP) wi th unprecedented precision. This evolution in our workflow aims to deliver near-real-time\, actionable insights\, transforming how we track land-use conversion and guide sustainable restoration across the world’s most vu lnerable non-forest biomes. DTSTAMP:20260625T073855Z LOCATION:Room 18 SUMMARY:Global monitoring of grassland and livestock: Current status\, chal lenges and next steps - Leandro Parente URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/HGTPJN/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-JRPSQY@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T170000 DTEND;TZID=Europe/Amsterdam:20261007T170500 DESCRIPTION:Mediterranean pine reforestations are increasing its vulnerabil ity to climate change\, particularly to recurrent drought events -as raisi ng tree decline and mortality rates have been identified-. In Spain\, deca des of limited forest management have contributed to mature dense stands c haracterized by high water stress and reduced ecosystem resilience. In thi s context\, adaptive silviculture emerges as a key strategy to enhance for est stability under changing environmental conditions.\nThis study evaluat es the effects of thinning treatments on the eco-resilience of mature pine plantations across Spain by integrating field data and multi-source remot e sensing observations. First\, spectral information derived from vegetati on indices (e.g.\, NDVI\, EVI or NBR) and shortwave infrared bands (SWIR1) are analyzed for detecting forest cover reduction promoted by thinning. S econd\, these observations and the Standardized Precipitation Evapotranspi ration Index (SPEI) are used for assessing how adaptive silviculture modul ates forest responses to drought events. Third\, dendrochronological data are incorporated to compare radial growth patterns with spectral dynamics\ , enabling cross-validation between traditional growth-based approaches an d remote sensing indicators to evaluate forest resilience. Results show th at forest management induces measurable changes in the spectral behavior\, and that treated stands exhibit faster recovery dynamics following extrem e droughts compared to unmanaged stands.\nBeyond these findings\, the curr ent study aims to frame ongoing research towards a more integrative assess ment of forest resilience\, combining spectral indicators with structural and ecohydrological perspectives. This approach seeks to advance the devel opment of scalable indicators of eco-resilience\, supporting forest manage ment strategies under future climate scenarios. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:The impact of adaptive silviculture on the spectral response and dr ought resilience of Mediterranean pine forests in Spain - Marina Muñiz Ma rtínez URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/JRPSQY/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-GDJSPB@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T171000 DTEND;TZID=Europe/Amsterdam:20261007T171500 DESCRIPTION:We present S2BIOPHYS\, the first global dataset of annual veget ation biophysical properties (LAIe\, FAPAR\, FCOVER) at 20 m resolution fr om Sentinel-2 (2019–2025). The product combines radiative transfer model inversion with iterative hyperparameter optimization using in-situ calibr ation and validation data. It provides per-pixel estimates with uncertaint y and observation counts\, validated against over 11\,000 ground measureme nts. S2BIOPHYS enables scalable monitoring of ecosystem condition\, restor ation\, and biodiversity. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:S2BIOPHYS: A Global Annual 20 m Dataset of Vegetation Biophysical P roperties from Sentinel-2 - Felix Specker URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/GDJSPB/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-E7NZH7@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T171500 DTEND;TZID=Europe/Amsterdam:20261007T173000 DESCRIPTION:Peatlands and other organic soils occupy a small fraction of th e land surface but store a large share of the global soil carbon. Drainage and fire in these systems are major sources of greenhouse gas (GHG) emiss ions\, yet remain poorly mapped. Remote sensing enables global monitoring of proxies for peatland disturbance\, but no monitoring system currently l inks the extent of organic soils\, disturbance\, and emissions at high spa tial resolution. Here we develop a 0.00025° (approximately 30 m) global g eospatial framework that overlays organic soils extent with multi-temporal land cover and land use data\, drainage infrastructure\, plantations\, pe at extraction areas\, coastal wetlands\, and burned area to delineate dist urbed organic soils. Using IPCC Wetlands Supplement default (Tier 1) metho ds\, we estimate CO2\, CH4\, N2O\, and CO emissions for disturbed organic soils over 2001–2024. Baseline results indicate that disturbed organic s oils emitted about 4.9 Gt CO2e yr-1 (4.5–5.1 Gt CO2e yr-1 across five in ventory periods)\, with roughly three quarters from drainage and one quart er from fires. Emissions and disturbed area are heavily concentrated in a small group of countries and land uses\, dominated by Russia and Indonesia \, particularly cropland and settlements in boreal and temperate zones and plantations on organic soils in the tropics. Sensitivity experiments that vary the extent of organic soils\, the drainage radius around infrastruct ure\, and IPCC default (Tier 1) emission factors yield a plausible range o f approximately 3–7 Gt CO2e yr-1. These estimates should not be interpre ted as a correction to existing peatland-specific emission estimates\, but as complementary\, more comprehensive monitoring of disturbed organic soi l systems under a harmonized\, globally consistent framework. The resultin g 30 m global maps of organic soil state\, disturbance\, and emissions dem onstrate how multi-temporal Earth observation can be combined with GHG inv entory methods to monitor peatland disturbance drivers\, identify high-emi tting hotspots\, and provide an updatable resource for inventories\, natio nally determined contributions\, and peatland conservation and restoration planning. DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:Global organic soil disturbance and emissions: leveraging Earth obs ervation–based geospatial data within an IPCC framework - Erin Glen URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/E7NZH7/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-E3KZMH@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T171500 DTEND;TZID=Europe/Amsterdam:20261007T173000 DESCRIPTION:We present an updated generation of the Global Pasture Watch (G PW) gross primary productivity (GPP) products for grassland ecosystems at 30 m spatial resolution. This new release builds on the original Landsat-b ased light use efficiency framework by integrating improved MODIS land sur face temperature (MOD21) and updated photosynthetically active radiation i nputs\, while also introducing an experimental Sentinel-2-based GPP produc t. The combined use of Landsat and Sentinel-2 observations strengthens the capacity to monitor grassland productivity dynamics with improved spatial detail and temporal consistency across local to global scales. Beyond lon g-term ecosystem assessment and grassland monitoring\, the updated GPW GPP products are also intended to support near-real-time (NRT) grassland biom ass estimation within the Time2Graze project. DTSTAMP:20260625T073855Z LOCATION:Room 18 SUMMARY:High-Resolution Grassland GPP Estimation with Landsat and Sentinel- 2 - Mustafa Serkan Isik URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/E3KZMH/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-RDHLEE@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T171500 DTEND;TZID=Europe/Amsterdam:20261007T172000 DESCRIPTION:Understanding where and why forest restoration succeeds remains a key challenge for global monitoring and policy. This project investigat es how satellite-based indicators of vegetation structure and function can capture restoration outcomes across spatial scales. We combine global rem ote sensing data with contextual information on climate\, landscape config uration\, and human pressure to identify drivers of restoration success an d compare intervention strategies. The results aim to inform scalable\, op erational approaches for monitoring forest recovery and supporting evidenc e-based restoration efforts worldwide. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Unraveling patterns and drivers of global forest restoration succes s using remote sensing - Felix Specker URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/RDHLEE/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-JRQQU8@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T171500 DTEND;TZID=Europe/Amsterdam:20261007T172000 DESCRIPTION:We present recent developments in the xcube ecosystem motivated by its involvement in the Open Earth Monitor project\, focusing on xcube server processing facilities and integration into scalable cloud workflows . We describe xcube server's established and newly-implemented API impleme ntations and processing facilities\, with particular attention to OGC API – Processes\, openEO\, the relationship to the eozilla processing framew ork\, and applicability to Open Earth Monitor science use cases. These cap abilities are complemented by further developments in the associated xceng ine package\, which converts Jupyter notebooks into EO application package s. With these developments\, xcube\, eozilla\, and xcengine provide a powe rful and versatile toolkit for developing and deploying EO workflows\, bot h locally and in the cloud. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:EO processes and workflows with xcube - Pontus Lurcock URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/JRQQU8/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-KZU8DF@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T172000 DTEND;TZID=Europe/Amsterdam:20261007T172500 DESCRIPTION:Savanna ecosystems cover approximately one fifth of Earth's lan d surface and provide critical ecosystem services to over one billion peop le\, yet their dynamic vegetation layer remains difficult to monitor consi stently at scale. Spaceborne lidar from the Global Ecosystem Dynamics Inve stigation (GEDI) mission provides vegetation structure measurements\, such as canopy height and cover\, but its spatially sparse sampling necessitat es extrapolation using satellite remote sensing data. Temporal consistency of these wall-to-wall mapping products remains a key challenge\, particul arly in heterogeneous savanna systems characterized by pronounced seasonal ity and complex disturbance dynamics.\nThis study compares two approaches for mapping GEDI-derived canopy height and cover across Kruger National Pa rk\, South Africa. The first uses hand-crafted Sentinel-1/2 features deriv ed from phenology-informed time series analysis. The second uses TESSERA f oundation model embeddings (pixel-wise representations of annual Sentinel- 1/2 time series) as open\, analysis-ready features with lightweight regres sion heads. Both approaches use phenology-aligned GEDI samples anchored to leaf-on conditions as training data\, and are evaluated using temporal cr oss-validation and independent airborne lidar data acquired across multipl e sites in the study area\, with particular focus on temporal transferabil ity and label efficiency.\nThe comparison addresses a question of growing practical relevance: does explicit phenological knowledge embedded in task -specific feature engineering outperform the implicit temporal representat ions learned by large-scale foundation models\, and under what conditions? The results will inform scalable\, open\, and reproducible approaches to vegetation structure monitoring in African savannas\, with direct relevanc e for biodiversity conservation and carbon stock assessment. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Comparing Foundation Model Embeddings and Phenology-Informed Featur e Engineering for Temporally Consistent Mapping of Savanna Vegetation Stru cture - Marco Wolsza URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/KZU8DF/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-YE9XVA@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T173000 DTEND;TZID=Europe/Amsterdam:20261007T173500 DESCRIPTION:Authors: Linda Luck (GFZ)\, Ben Brede (GFZ)\, Johannes Wilk (GF Z)\, Arnan Araza (WUR)\, Geike De Sloover (UGent)\, Bert Gielen (Universit y of Antwerp)\, Martin Herold (GFZ)\n\nTerrestrial laser scanning (TLS) an d Uncrewed Aerial Vehicle laser scanning (ULS) provide highly detailed and reliable measurements of vegetation structure and have become an importan t tool for forest and ecosystem research. Despite its high value\, openly available TLS/ULS datasets remain scarce. Previous initiatives to establis h centralised data collection have faced challenges in gaining sufficient participation and maintaining long-term feasibility\, highlighting the nee d for alternative approaches to improving accessibility and usability of T LS/ULS data.\n\nIn addition to making some of our data publicly available\ , we are developing a reproducible processing workflow that can be applied across sites and shared among partners. This workflow is currently being implemented for several contrasting ecosystems\, including the deciduous m ixed forest of Hohes Holz (Germany)\, the savanna ecosystem at Las Majadas (Spain)\, and additional sites currently under preparation. By standardis ing processing and derived outputs\, the approach enables consistent gener ation of structural metrics that can be shared and integrated across proje cts.\n\nAs a first application example\, we present the ESA Forest DTC pro ject\, where tree-level structural metrics derived from high-resolution TL S scans with manually corrected segmentation are combined with large-scale automated segmentation and feature extraction from UAV-based lidar to sup port ecosystem modelling.\n\nLooking ahead\, cloud-based and online proces sing services - such as those provided by ForesSens and currently being de veloped within the 3D-Trees project - may represent a promising pathway fo r enabling broader access to TLS processing and derived products without r equiring specialised local computing infrastructure. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Standardising Terrestrial and ULS Laser Scanning Processing for Cro ss-Site Data Sharing and Applications - Linda Luck URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/YE9XVA/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-ZWFDRC@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T173500 DTEND;TZID=Europe/Amsterdam:20261007T174000 DESCRIPTION:Large-scale reforestation is central to climate mitigation and ecosystem restoration\, yet monitoring when and where restoration activiti es occur remains a major challenge. Existing satellite based approaches ty pically detect forest recovery only after canopy development\, limiting th eir usefulness for timely monitoring\, reporting\, and verification (MRV) of restoration efforts.\nWe present a scalable framework for detecting ref orestation interventions within one year of planting using multi-sensor sa tellite time series. The approach integrates Sentinel-1 radar and Sentinel -2 optical data to learn a characteristic temporal signature of restoratio n\, capturing transitions from stable pre-intervention conditions to distu rbance and early vegetation recovery. Training leverages a global dataset of verified reforestation sites combined with a synthetic control strategy to generate spectrally matched non restoration samples.\nOur results show that early stage reforestation can be identified at the pixel level acros s diverse climate zones\, substantially improving the temporal resolution of forest monitoring. The poster will present global scale examples\, mode l outputs\, and temporal signatures illustrating how restoration signals e merge prior to canopy closure.\nThis work supports more timely and transpa rent monitoring of reforestation efforts and has direct relevance for carb on accounting\, climate finance\, and large-scale restoration initiatives. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Early Detection of Reforestation Interventions Using Multi-Sensor S atellite Time Series - Angela John URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/ZWFDRC/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-8YKTEM@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T174000 DTEND;TZID=Europe/Amsterdam:20261007T174500 DESCRIPTION:This study presents the development and multi-regional applicat ion of the Normalized Radar Burn Ratio (NRBR)\, a novel Synthetic Aperture Radar (SAR)-based index designed to improve burned area detection under c hallenging observational conditions. Unlike traditional optical indices su ch as the differenced Normalized Burn Ratio (dNBR)\, NRBR exploits the com plementary behavior of Sentinel-1 C-band co-polarized (VV: vertical transm it–vertical receive) and cross-polarized (VH: vertical transmit–horizo ntal receive) backscatter signals\, enhancing the contrast between burned and unburned surfaces by capturing fire-induced structural changes in vege tation.\nThe NRBR formulation is based on the normalized difference betwee n polarization-specific Radar Burn Ratios\, effectively integrating post- to pre-fire backscatter dynamics while reducing speckle noise and topograp hic effects. Initial validation in Mediterranean ecosystems demonstrated t hat NRBR improves burned area delineation compared to conventional radar i ndices\, achieving strong agreement with optical-based metrics and competi tive segmentation performance when implemented within a U-Net deep learnin g framework.\nBuilding on these results\, the index was further evaluated across diverse fire-prone regions including Portugal\, Spain\, California\ , and Canada\, encompassing Mediterranean\, chaparral\, and boreal ecosyst ems. The results indicate that NRBR achieves performance comparable to\, a nd in some cases exceeding\, optical approaches\, particularly in cloud-pr one or smoke-affected conditions where optical data are limited. Additiona lly\, a SAR–optical fusion strategy combining NRBR and dNBR further impr oves mapping accuracy and spatial consistency at large scales.\nOverall\, NRBR demonstrates strong potential as a robust and scalable alternative fo r burned area mapping\, providing consistent performance across different land cover types and environmental conditions. Its cloud independence and sensitivity to vegetation structural changes position it as a valuable too l for operational wildfire monitoring and next-generation multi-sensor map ping frameworks. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Multi-Sensor Fusion for Large-Scale Burned Area Mapping: The role o f NRBR - Yonatan Tarazona Coronel URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/8YKTEM/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-VMEVPF@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T175500 DTEND;TZID=Europe/Amsterdam:20261007T180000 DESCRIPTION:Forest structural diversity - the spatial heterogeneity of cano py architecture across vertical and horizontal dimensions - is a fundament al component of ecosystem functioning. Yet its continuous global mapping r emains constrained by the sparse orbital sampling of spaceborne LiDAR miss ions such as the Global Ecosystem Dynamics Investigation (GEDI). \nHere\, we integrated globally distributed GEDI-derived structural diversity metri cs with dense-vector representations from a geospatial vision foundation model pretrained on multi-source satellite imagery. Specifically\, we used the Google Satellite Embedding dataset\, derived from the AlphaEarth Foun dations model\, which provides globally consistent 64-dimensional embeddin gs at 10 m resolution from multi-source satellite imagery. Our analysis sp ans GEDI's full tropical-to-temperate sampling domain (52°N–52°S)\, en compassing 14 major biomes from temperate conifer to tropical moist broadl eaf forests.\nRandom forest regression models were fitted within a spatial ly blocked cross-validation framework stratified by biogeographic region. Cross-validated R² was consistently high across structural diversity dime nsions\, with low inter-fold variance indicating robust transferability ac ross held-out biogeographic regions. Predicted structural diversity reveal ed strong but metric-dependent spatial gradients\, reflecting the distinct axes of canopy architecture — from height and complexity to vertical pr ofile shape - captured across the global sampling domain.\nOur results dem onstrate that geospatial foundation-model embeddings capture information a cross both vertical and horizontal dimensions of forest canopy architectur e\, thus providing a scalable pathway for wall-to-wall inference of forest structural diversity from existing spaceborne observations DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Foundation-model embeddings predict global variation in forest stru ctural diversity - Marco Girardello URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/VMEVPF/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-KL7URJ@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T180000 DTEND;TZID=Europe/Amsterdam:20261007T180500 DESCRIPTION:Forest and grassland restoration constitutes a central objectiv e of global initiatives\, including the United Nations Decade on Ecosystem Restoration. Nevertheless\, the synergistic mechanisms and quantitative l inkages between enhanced restoration potential and improved ecosystem serv ices (ESs) remain insufficiently understood. In this study\, we developed the Forest and Grassland Restoration Potential Achievement Efficiency (FGR PAE). By integrating remote sensing data\, ecosystem service assessments\, and nonlinear modeling\, we constructed a comprehensive framework to eval uate restoration benefits in the Yellow River Basin (YRB)\, a representati ve region of large-scale ecological restoration. This framework systematic ally investigates the long-term spatiotemporal dynamics of FGRPAE\, as wel l as its interactive patterns with ecosystem services and underlying nonli near response mechanisms. The results show that FGRPAE increased at an ave rage annual rate of 0.0083\, corresponding to a cumulative growth of 53.73 %. During 2000-2010\, FGRPAE increased at an annual rate of 0.0096\, yield ing a cumulative increase of 32.45%. In contrast\, from 2010 to 2020\, the growth rate decelerated to 0.0063 per year\, with a total increase of 16. 02%\, approximately 49.52% of the increase observed during 2000-2010. Conc urrently\, the proportion of areas exhibiting spatial trade-offs between F GRPAE and comprehensive ecosystem services (CES) rose by 76.12% between 20 10 and 2020. A nonlinear enhancement relationship was identified between F GRPAE and CES. However\, CES gains plateau when FGRPAE exceeds approximate ly 50%. This study shifts the analytical focus from restoration intensity to restoration efficiency\, demonstrating that neglecting spatial trade-of fs between FGRPAE and ESs may compromise the overall effectiveness of ecol ogical restoration. Accordingly\, we propose an optimized spatial configur ation for restoration planning that emphasizes the integrated consideratio n of forest and grassland restoration potential and ecosystem service func tions under resource constraints. The proposed framework supports the tran sition of ecological engineering from “area expansion” to “function enhancement\,” offering actionable policy guidance for optimizing restor ation strategies within ecological carrying capacity limits. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:A Framework to Optimize the Potential Restoration Achievement and E cosystem Services Trade-offs applied in the Yellow River Basin - liuyuan URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/KL7URJ/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-SFZHLV@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T180500 DTEND;TZID=Europe/Amsterdam:20261007T181000 DESCRIPTION:Abstract\nFlash droughts—characterized by abrupt onset and ra pid soil moisture depletion—are emerging as a consequential hydroclimati c extreme across Europe. Their fast evolution\, strong sensitivity to atmo spheric evaporative demand\, and reinforcement through land–atmosphere c oupling challenge traditional drought monitoring approaches that were larg ely developed to track slowly evolving deficits. Despite growing attention \, continental-scale understanding of how flash droughts initiate\, propag ate\, and vary across Europe’s diverse climate regimes remains limited.\ nHerein\, we propose a framework for flash drought detection and character ization using three complementary soil moisture perspectives: ASCAT satell ite observations\, ERA5-Land reanalysis\, and GloFAS hydrological model so il moisture. The analysis covers 2007–2024 at dekadal (10-day) resolutio n. Flash drought onset is diagnosed from rapid short-timescale declines in near-surface soil moisture percentiles derived from ASCAT and ERA5-Land\, while GloFAS is used to assess whether—and where—these surface drying signals propagate into catchment-scale hydrological response. To ensure c omparability across datasets with differing process representations and ef fective soil depths\, all soil moisture variables are expressed in a commo n percentile space\, which isolates anomalous moisture states relative to local climatology.\nUsing this unified framework\, we quantify key flash d rought characteristics—including frequency\, mean duration\, severity\, and mean onset speed—across Europe and examine how these metrics vary ac ross major climate regimes. The findings highlight pronounced regional het erogeneity and systematic cross-system contrasts. ASCAT captures the sharp est and most immediate surface drying signals\, whereas ERA5-Land and GloF AS provide complementary insight into physically consistent drivers and th e potential for downstream hydrological impacts. Overall\, the results emp hasize that flash drought diagnosis benefits from combining observation-in formed onset detection with process-oriented evaluation of drivers and hyd rological propagation. This multi-perspective approach offers a consistent basis for strengthening monitoring and supporting early-warning readiness under Europe’s intensifying hydroclimatic variability.\n\n\nKeywords: F lash drought\; Soil-moisture\; ASCAT\, ERA5-Land\; GloFAS\, Hydrological r esponse\, Land–atmosphere coupling\, Europe. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:From Surface Drying to Hydrological Response: An Integrated Diagnos is of Flash Droughts across Europe - VAIBHAV KUMAR URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/SFZHLV/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-9G3MQF@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T181000 DTEND;TZID=Europe/Amsterdam:20261007T181500 DESCRIPTION:Fungal diseases such as Ascochyta remain a major obstacle to ch ickpea production\, leading to significant yield losses if not detected ea rly. Whilst hyperspectral imaging (HSI)\, combined with machine learning\, has demonstrated strong potential for early and non-destructive detection under controlled laboratory conditions\, its transferability to real-worl d field environments remains a major challenge. This study aims to validat e detection models developed in the laboratory under field conditions usin g multispectral images acquired by a drone. Following promising results ob tained using hyperspectral data (400–1000 nm) and advanced machine learn ing pipelines\, we have extended our approach to drone-based multispectral detection to assess its robustness in real-world agricultural scenarios. Field data were acquired using drone-mounted sensors capturing key spectra l bands relevant to vegetation health and stress detection. A comprehensiv e processing pipeline was implemented\, comprising radiometric correction\ , image pre-processing\, vegetation index calculation\, and feature extrac tion. The previously developed classification framework was adapted and ap plied to multispectral data\, incorporating both spectral and statistical features. The results demonstrate that the proposed approach can be succes sfully applied from the laboratory to field conditions\, achieving high de tection performance with a classification accuracy of over 90% in distingu ishing healthy chickpea plants from infected ones. Furthermore\, the syste m proved capable of detecting signs of infection at an early stage in the canopy\, despite environmental variability such as changes in light intens ity and ground background effects. These results confirm the feasibility o f deploying AI-based disease detection systems using drone-based multispec tral imaging in real agricultural environments. This work represents a sig nificant step towards operational precision agriculture solutions\, enabli ng large-scale monitoring\, early intervention\, reduced chemical inputs\, and improved crop management strategies. Future work will focus on valida tion over multiple seasons\, integration with close-range detection\, and extension to other plant-pathogen systems. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:AI-Driven Early Detection of Chickpea Ascochyta Blight: From Contro lled Hyperspectral Analysis to UAV Multispectral Field Monitoring - Mohame d URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/9G3MQF/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-DJBM9P@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T181500 DTEND;TZID=Europe/Amsterdam:20261007T182000 DESCRIPTION:The coastal border of Semarang and Demak in Central Java\, Indo nesia\, faces unprecedented mangrove deforestation driven by rapid land su bsidence\, sea-level rise\, aquaculture expansion\, and industrialization. Traditional optical remote sensing approaches are severely constrained by persistent cloud cover in this tropical environment\, resulting in detect ion lags of weeks to months that preclude timely intervention. This study presents an iterative Bayesian updating framework for near-real-time mangr ove deforestation monitoring through multi-sensor fusion of Sentinel-1 Syn thetic Aperture Radar (SAR) and optical imagery from Landsat-8/9 and Senti nel-2. We formulate a probabilistic change detection model where posterior deforestation probabilities are sequentially updated with each new satell ite observation\, incorporating VH-polarized backscatter from Sentinel-1 a longside three complementary optical indices: Normalized Difference Vegeta tion Index (NDVI)\, Mangrove Vegetation Index (MVI)\, and Enhanced Mangrov e Index (EMI). Four experimental scenarios were evaluated across the 2018- 2025 period: (1) SAR-Optical Baseline (VH + NDVI)\, (2) Structure-Focused (VH + MVI)\, (3) Moisture/Soil-Focused (VH + EMI)\, and (4) Full Integrate d Suite (VH + EMI + NDVI + MVI). Validation through field surveys\, high-r esolution imagery\, and comparison with existing deforestation maps demons trated that Scenario 4 achieved the highest F1-score (0.89) and lowest det ection lag (8.3 days median)\, reducing false positives from tidal floodin g by 67% compared to single-sensor approaches. The integration of structur al information from SAR and MVI with spectral-moisture signals from EMI an d NDVI enabled robust discrimination between genuine deforestation events and natural tidal dynamics. Mathematical formulations for prior specificat ion\, likelihood functions\, and posterior updating are presented in detai l\, alongside practical implementation considerations for tropical coastal environments. These findings provide actionable guidance for local coasta l management agencies in Semarang-Demak to implement operational near-real -time monitoring systems that can trigger rapid response to illegal loggin g and land conversion. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Iterative Bayesian Updating for Near Real-Time Mangrove Deforestati on Monitoring: A Multi-Sensor Fusion Approach in Semarang-Demak\, Indonesi a - Munawaroh Munawaroh URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/DJBM9P/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-E8HUES@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T182500 DTEND;TZID=Europe/Amsterdam:20261007T183000 DESCRIPTION:Large-scale and highly accurate wheat yield prediction is of gr eat importance for \nensuring food security\, supporting agricultural poli cymaking\, and guiding grain \nallocation. In recent years\, the rapid dev elopment of remote sensing technologies and \ndeep learning algorithms has provided powerful tools for large-scale crop yield \nprediction. However\ , crop yield is jointly influenced by multiple environmental factors\, \ns uch as climate\, soil\, and topography. Existing studies often adopt simpl e feature \nconcatenation or fixed-weight fusion strategies\, lacking adap tive modeling of relative\nmodality importance\, which limits further impr ovement in prediction accuracy. To \naddress this issue\, this study propo ses a Transformer-based multi-modal adaptive Gated \nFusion model (TMMGF). The model employs Transformers to model dynamic time \nseries of remote s ensing spectral data and climate variables\, applies multilayer \nperceptr ons (MLP) to handle static environmental factors including soil and topogr aphy. \nMultiple modalities are then integrated through a gated fusion mec hanism to achieve\nadaptive weighted fusion. This study was conducted acro ss the conterminous United \nStates\, based on county-level winter wheat y ield records from 2008 to 2023. The \nTMMGF was systematically compared wi th an LSTM-based multimodal adaptive \nGated Fusion model (MMGF)\, Transfo rmer single-modal remote sensing model\, \nTransformer single-modal climat e model\, MLP single-modal soil model\, and MLP \nsingle-modal topography model. The results show that TMMGF achieves the best \nperformance\, with an average R² of 0.813\, RMSE of 0.571 t/ha\, and MAPE of 14.49% \nin 10- fold cross-validation\, significantly outperforming the baseline models. I n \nparticular\, compared with the LSTM-based multimodal model MMGF (R² = 0.796\, \nRMSE = 0.598 t/ha\, MAPE = 15.11%)\, TMMGF shows clear advantag es in both \naccuracy and stability. This study demonstrates that a Transf ormer-based adaptive \nmultimodal fusion framework can effectively integra te heterogeneous data sources and \nprovides a promising technical pathway for high-accuracy large-scale wheat yield \nprediction. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Transformer-Based Adaptive Multimodal Fusion Model for Remote Sens ing Large-scale Winter Wheat Yield Prediction - Haoran Meng URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/E8HUES/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-JRVYY9@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T183000 DTEND;TZID=Europe/Amsterdam:20261007T183500 DESCRIPTION:The DynaFun project intends to showcase EFTs as a biodiversity indicator for non-stand shifts in forest ecosystems. EFTs are land delinea tions that result from similar energy and matter processes such as carbon storage or hydrological cycle. As such\, vegetation indices are a well-acc epted proxy indicator for ecosystem functionality. In this study Sentinel- 2 satellite imagery has been used to derive remotely sensed NDVI to produc e an EFT land classification over Catalonia. Plant carbon dynamics are inf erred through the derivation of productivity (NDVImean)\, seasonality (NDV Icovariate) and phenology (NDVIDay of Maximum). Catalonia has a mixed land use land cover (LULC) system that is heavily influenced by its Mediterran ean climate. It causes interannual\, variable environmental changes reflec ted in its vegetation dynamics. Non-stand shifts occur before visible stru ctural changes. This project presents an opportunity to enrich Land use la nd cover and habitat categories to determine their stability\, resilience and drivers of change. This information is crucial for forest management a nd can act as an early warning indicator for biodiversity change for plann ers and policy developers. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Ecosystem Functionality of Catalonian Landscapes: Change Assessment of Ecosystem Functional Types (EFTs) Using Sentinel-2 Derived NDVI - Lynn Fanikiso URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/JRVYY9/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-8YNN39@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261007T183500 DTEND;TZID=Europe/Amsterdam:20261007T184000 DESCRIPTION:Flash flood events are increasing in frequency and intensity in Mediterranean regions\, requiring rapid\, reliable\, and scalable monitor ing approaches to support emergency response and climate adaptation. Earth Observation (EO) offers a powerful means to provide timely spatial intell igence\; however\, single-sensor approaches remain limited by cloud cover\ , revisit frequency\, and data latency. This work presents an automatic\, multi-sensor\, modular\, and open-source flood mapping framework designed to deliver actionable information for emergency responders through near-re al-time flood detection coupled with a first-pass impact assessment.\nThe proposed methodology integrates Synthetic Aperture Radar (Sentinel-1) and multispectral imagery (Sentinel-2 and Landsat 8) with ancillary geospatial datasets within a unified processing pipeline. A change detection approac h is applied to pre- and post-event observations\, followed by automated t hresholding and morphological filtering to generate consistent flood exten t maps. To reduce noise sensitivity\, outputs from multiple sensors are th en fused at the pixel level to generate flood extent\, severity\, and dama ge assessment maps. \nThe framework was validated against ground-truth dat a from the October 2024 flash flood event in Valencia\, with results clear ly demonstrating the value of automated multi-sensor data fusion by increa sing the likelihood of acquiring usable observations by up to ~60%. This m odular architecture is fully reproducible and designed for extensibility\, enabling the integration of additional sensors and seamless deployment wi thin open EO ecosystems and distributed data infrastructures. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Data fusion for flood monitoring - Ana Linares Barrio URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/8YNN39/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-7NRN3X@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T103000 DTEND;TZID=Europe/Amsterdam:20261008T110000 DESCRIPTION:Satellite data have transformed our ability to observe land-use dynamics\, forest change\, and environmental pressures at unprecedented s patial and temporal scales. Yet a critical challenge remains: translating vast volumes of data into actionable information that meaningfully informs policy and leads to real-world implementation. This talk highlights the i mportance of open and transparent data\, with a particular emphasis on the ease of accessing information and answering practical questions using dat a to drive evidence-based decision-making. It underscores the need to inte grate satellite-derived land-use and forest data\, along with information on change and its impacts on climate\, nature\, and people\, into operatio nal platforms and planning processes. By doing so\, governments\, supply c hains\, and local stakeholders can better assess risks\, target interventi ons\, and track outcomes. Transforming data into decision-ready insights i s essential for strengthening land-use governance\, advancing sustainable forest management\, and delivering measurable environmental and social imp acts. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:From Data to information and Policy to Implementation - Fred Stolle URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/7NRN3X/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-AADXE7@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T110000 DTEND;TZID=Europe/Amsterdam:20261008T113000 DESCRIPTION:Keynote talk by Xavier Pons. Abstract and title to be provided. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Keynote - Xavier Pons URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/AADXE7/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-NWX9DD@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T113000 DTEND;TZID=Europe/Amsterdam:20261008T120000 DESCRIPTION:Artificial Intelligence offers powerful tools to understand and respond to the climate crisis\, but there are no shortcuts or silver bull ets. Bigger models or geoengineering schemes alone won't solve our challen ges. What matters is applying AI where it can make a real difference and f or those who need it most: monitoring crops and ensure food security level s\, tracking air quality\, predicting floods and wildfires\, and supportin g vulnerable communities on the frontlines of climate change. In this talk \, I will share how combining satellite data\, local knowledge\, and advan ces in explainable and causal AI can turn information into actionable insi ghts. From anticipating extreme weather to guiding humanitarian responses\ , these applications show how AI can strengthen decision-making and build climate resilience\, helping policymakers act today for a more sustainable tomorrow. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:AI for Climate Resilience: From Data to Decisions that Matter - Gus tau Camps-Valls URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/NWX9DD/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-XM77A8@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T120000 DTEND;TZID=Europe/Amsterdam:20261008T123000 DESCRIPTION:Acknowledging the value of earth observation data and sharing t he frustration in the difficulty of working with it\, just shy of 10 years ago a group of earth observation scientists and practitioners set out to described their vision (https://r-spatial.org/2016/11/29/openeo.html) of h ow they would like to work with this treasure of data without falling into the traps of a vendor lock-in\, custom codes for data access and the need to deal with 15 different data formats. Inspired by the way gdal solved t he latter of these problems for raster data driven GIS systems\, they envi sioned an API agnostic to both the clients programming language and the se rver site implementation of EO processing workflows. I joined this group o f like-minded researchers in the Horizon 2020 project openEO that was fund ed about a year after.\nDuring this keynote I will talk about the importan ce of open-source development and community building in conjunction with t he importance of public funding from the European Commission and the Europ ean Space Agency as the two main supporters of making this idea a reality. OpenEO is based on an approach of co-design and co-development\, involvin g both earth observation data analysts and researchers as well as software developers and industry service providers right from the start. It is now available as part of many operational service offerings\, such as openEO platform\, the Copernicus Data Space Ecosystem and Destination Earth. ope nEO is not only the foundation of individual service offerings\, but even federations of multiple implementations\, sharing data and processing reso urces leveraging efficient interoperability based on cloud native data for mats and harmonized specifications for\, data discovery\, access\, process ing and sharing of results. It is fully embracing the FAIR data principles \, while supporting researchers and data analyst to focus on what the want to do instead of how the implementations needs to be optimized for modern scalable data infrastructures. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:openEO from an idea in a whitepaper to a community standard in the geospatial data processing. - Alexander Jacob URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/XM77A8/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-TLBCMP@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T123000 DTEND;TZID=Europe/Amsterdam:20261008T130000 DESCRIPTION:Greenhouse gas (GHG) emissions from Agriculture\, Forestry and Other Land Uses (AFOLU) account for roughly one-quarter of global net anth ropogenic emissions\, yet consistent monitoring remains challenging. We pr esent a global\, open\, spatially explicit monitoring framework developed by WRI’s Land & Carbon Lab and partners that combines Earth observation data\, geospatial models\, and IPCC inventory methods to map land-related emissions and removals from 2016–2024. Accessible through Global Nature Watch and other platforms\, the system integrates datasets on land cover c hange\, forest dynamics\, fire\, biomass\, soils\, livestock\, and agricul tural management. We highlight applications for governments\, civil societ y\, companies\, researchers\, and policymakers supporting climate action a nd progress tracking. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:From Land Cover Change to Greenhouse Gases: An Open Geospatial Moni toring Framework - Nancy Harris URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/TLBCMP/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-WQBUZF@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T160000 DTEND;TZID=Europe/Amsterdam:20261008T164500 DESCRIPTION:Monitoring forest cover change\, wildfire risk\, and post-fire recovery demands integrating heterogeneous data sources (i.e. satellite im agery\, field observations\, weather feeds\, and alert systems) into a sha red operational picture. Yet existing tools force a choice: either powerfu l but expensive proprietary platforms\, or open-source solutions that requ ire significant server infrastructure and maintenance.\n\nThis workshop in troduces Driades\, an open-source\, self-hosted geospatial tool developed by h4ck1ng.science that runs entirely in the browser with no backend serve r. By leveraging cloud-native formats (Zarr\, GeoParquet\, and PMTiles amo ng others) served directly from S3-compatible object storage\, Driades ena bles users to visualise satellite imagery\, execute spatial SQL queries vi a WASM\, and run basic transformations using WebGPU. Heavy computational t asks (such as machine learning inference for burn scar detection) can be o ffloaded to remote APIs\, keeping the client lightweight while providing a ccess to geospatial foundation models hosted on model registries like Hugg ing Face.\n\nDriades aims to reduce the friction of accessing and sharing geospatial data\, providing non-experts with a simple interface to explore \, annotate\, and distribute interactive results among collaborators witho ut requiring specialised infrastructure. DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:Driades: A Collaborative\, Browser-Based Forest Monitoring Dashboar d Built on Cloud-Native Geospatial Formats - Carlos Vivar Ríos URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/WQBUZF/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-MRHJ3P@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T161500 DTEND;TZID=Europe/Amsterdam:20261008T163000 DESCRIPTION:Hydroclimatic forcing of similar magnitude can produce contrast ing hydrologic responses within the same basin. Here\, we investigate how soil-moisture memory (SMM) regulates the translation of atmospheric anomal ies into basin-scale hydrologic response across the Po River Basin. To add ress this question\, we developed an open and reproducible Earth-observati on workflow based on a multi-source data cube that integrates Sentinel-1 o bservations with hydroclimatic forcing represented by the Precipitation– Evapotranspiration Anomaly Index (PEAI)\, derived from HYPER-P precipitati on and GLEAM evapotranspiration for 2016–2022. This framework enables as sessment of how SMM varies across land-surface types and during major hydr oclimatic transition episodes. \n\nThe analysis reveals marked contrasts a cross the basin. Irrigated agricultural areas exhibit the strongest memory \, with median persistence close to 3 weeks and low instability (~0.19)\, whereas changed surfaces show weaker and more volatile behavior\, with per sistence of about 1.7 weeks and instability approaching 0.24. Non-irrigate d agricultural areas define a distinct intermediate regime\, characterized by lower persistence and higher instability than irrigated areas\, but le ss volatility than changed surfaces. At the basin scale\, major forcing ep isodes affect approximately 80–90% of the basin\, yet response hotspots typically occupy only 20–40%\, indicating that atmospheric anomalies are not expressed uniformly but are selectively filtered by antecedent basin state and land-surface conditions. \n\nEvent-based analysis further shows that the hydrologic expression of forcing reversal depends strongly on ant ecedent SMM conditions. A continuous 28-day drought–flood abrupt alterna tion (DFAA) sequence in May–June 2020\, automatically detected from the 2016–2022 record\, includes a major drought-to-flood transition (DTF) fr om 21 May to 4 June and a major flood-to-drought transition (FTD) from 4 t o 18 June. Although the two phases exhibit near-equivalent PEAI amplitudes \, reversing from -1.195 to 2.176 during the DTF phase (Δ = 3.371) and fr om 2.176 to -1.250 during the subsequent FTD phase (Δ = 3.426)\, the resu lting basin-scale responses are asymmetrical\, indicating that forcing rev ersal of similar magnitude is not translated into equivalent hydrologic ex pression. These results indicate that hydrologic response to forcing rever sal depends more strongly on antecedent soil-moisture memory than on forci ng amplitude alone. \n\nAdditional comparisons among automatically detecte d FTD events with similar forcing trajectories reinforce this interpretati on. Two major transitions\, detected on 5 March 2020 and 13 May 2021\, sho w comparable forcing duration and amplitude but differ substantially in ti ming\, coherence\, and post-transition evolution. These contrasts are cons istent with distinct memory regimes: the 2020 event is associated with low persistence (~0.20) and sustained high instability (~0.75)\, whereas the 2021 event combines very low persistence at transition (~0.13) with rapid recovery toward higher persistence and lower instability thereafter. Overa ll\, the results show that antecedent soil-moisture memory and land-surfac e conditions exert a strong control on how hydroclimatic forcing is transl ated into basin-scale hydrologic response. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Soil-Moisture Memory as a Regulator of Hydrologic Response in the P o River Basin (Italy) - imane serbouti URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/MRHJ3P/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-X97YJS@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T163000 DTEND;TZID=Europe/Amsterdam:20261008T164500 DESCRIPTION:As FAIR principles become increasingly central to open science and research data stewardship\, a persistent gap remains between their bro ad endorsement and their consistent practical validation in real-world rep ositories. This challenge is particularly visible for geospatial datasets\ , where domain-specific requirements such as spatial formats\, metadata ri chness\, and interoperability standards are not always captured by general FAIR assessment approaches. In this work\, a structured FAIR assessment w as applied to datasets produced within the Open-Earth-Monitor Cyberinfrast ructure (OEMC) project. To support the evaluation\, we developed a FAIR as sessment tool tailored to geospatial data entries published on Zenodo\, wh ile designing the underlying framework to remain flexible and transferable to other repository environments. The assessment identifies both strength s and recurring gaps in current data publication practices and provides ac tionable recommendations for improving the long-term usability\, transpare ncy\, and scientific value of geospatial datasets. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:How FAIR Is Geospatial Data? An Assessment of OEMC Datasets - Imma Serra\, Mustafa Serkan Isik URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/X97YJS/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-YKFVG9@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T164500 DTEND;TZID=Europe/Amsterdam:20261008T173000 DESCRIPTION:One of the most significant deliverables of the OEMC project ar e global\, cloud-less Landsat monthly time series from 2000–2025 at 30 m resolution. The Landsat global mosaics (V1) are explained in detail in Co nsoli et al. (2025\; https://peerj.com/articles/18585/). The Landsat V2 is at the order of magnitude more ambitious aiming at monthly products in 16 bit format and will significantly less artifacts. The pipeline uses a fou r-step process for improved quality\, including gap-filling using spatial and temporal neighbours\, data fusion and final gap filling using global m odels. The results of cross-validation show improvements in accuracy in co nsistency. Major project challenges include needing 1PB of storage and sec uring post-2025 commercial services. Landsat V2 can also be used to derive embeddings for 2000-2025. DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:Landsat monthly cloud-free complete consistent mosaics 2000-2025 - Tom Hengl (OpenGeoHub)\, Sajed URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/YKFVG9/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-BPWMLF@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T164500 DTEND;TZID=Europe/Amsterdam:20261008T173000 DESCRIPTION:GeoFoundation embeddings encode huge amounts of Earth Observati on data and by condensing this into a small vector of numbers\, they can m ake many downstream analyses much easier to perform. However\, the embeddi ngs represent a latent state and as such can be abstract to understand. \n This workshop aims to demonstrate how embeddings can be used and explore h ow to visualize them and make them more usable. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Working with and visualizing GeoFoundational AI embeddings - Zhengp eng (Frank) Feng\, Mike Harfoot URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/BPWMLF/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-QGFTAX@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T180000 DTEND;TZID=Europe/Amsterdam:20261008T184500 DESCRIPTION:Sampling is not actual resolution: the 1 km illusion in satelli te hydrology refers to the discrepancy between a dataset's digital samplin g grid and its true physical resolution. The "rush" to create high-resolut ion data has outpaced the ability to validate it on the ground due to miss ing in situ monitoring networks required to independently validate 1 km al gorithms on a global scale. While many modern satellite hydrological produ cts are labeled as "1 km" resolution\, this often reflects how the data is stored on a grid rather than what the sensor actually "sees". These produ cts frequently remain "physically blind" to hyper-local anomalies. The fun damental challenge resides in the intrinsic trade-off between spatial reso lution and temporal frequency.\n\nWith regard to satellite soil moisture p roducts\, active radar sensors (e.g. Sentinel-1) provide true 1 km spatial acuity but suffer from temporal gaps\, while passive radiometers (e.g. SM AP) offer excellent daily tracking but produce oversampled illusions at th e 1 km scale. For practitioners\, the selection of a dataset must be dicta ted by the physical scale of the hydrological event—ranging from farm-sc ale irrigation to continental-scale drought—rather than the digital labe l on the file.\n\nIn the workshop\, a series of real-world stress tests of the "1 km illusion in satellite hydrology" will be presented. These stres s tests will demonstrate the hydrological applications of the 1 km illusio n\, including the mapping of localised summer storms\, the estimation of i rrigation at the field scale\, and the impact of wildfires on water infilt ration. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:The 1-km illusion in remote sensing for hydrology - Luca Brocca URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/QGFTAX/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-NS7Z9N@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T180000 DTEND;TZID=Europe/Amsterdam:20261008T184500 DESCRIPTION:The collaboration among PIs of eddy covariance sites\, Research Infrastructures and Regional Data Hubs gave birth to a new system for sha ring globally distributed\, standardized flux tower datasets: the FLUXNET Data System Initiative. The system is a milestone for the ecosystem flux c ommunity and its stakeholders (e.g. satellite and modeling communities)\, for it allows to shift from static\, periodic releases (FLUXNET2015) to a quasi-real-time approach\, in which datasets become findable as soon as th ey are processed and quality-controlled. The FLUXNET Data System Initiativ e is built upon three pillars: uniform (meta)data formatting\; a unique pr ocessing software (ONEFlux) used by the three Regional Hubs (ICOS\, AmeriF lux and OzFlux/TERN)\; a data access tool based on APIs: the FLUXNET Shutt le\, developed in the context of the OEMC project. Written in Python and a vailable on GitHub\, the Shuttle is a one-step access system that enables users to find and download open-licensed eddy covariance datasets worldwid e with simple queries executed via command line or graphical interfaces. D ifferent search criteria are available to discover the datasets\, no matte rs where they have been collected and by which of the three Regional Hubs they have been processed. The Shuttle enables the definition of new standa rds for flux data interoperability. \n\nParticipants to this workshop will be able to search and download eddy covariance datasets from different si tes. A quick overview of the datasets characteristics (data format\, metad ata available\, variables included) will be provided at the beginning\, an d then attendees will install the Shuttle on their own devices and explore its functionalities. By the completion of two exercises\, participants wi ll become acquainted with potential use cases of global flux tower dataset s\, like comparison of ecosystem responses to stressors across different c limate conditions. DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:The FLUXNET Shuttle: Enabling Access to Globally Distributed Flux T ower Data - Simone Sabbatini URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/NS7Z9N/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-9Q7PZY@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T180000 DTEND;TZID=Europe/Amsterdam:20261008T184500 DESCRIPTION:In this workshop\, the participants will have access to harmoni zed\, analysis-ready\, gap-filled and complete Landsat global mosaics from 1997 onward in cloud-optimized GeoTIFF (COG) format (130 TB of data) in C DSE (https://browser.stac.dataspace.copernicus.eu). Spanning over 25 years and structured in 7 spectral bands (RGB\, NIR\, SWIR-1\, SWIR-2 and therm al)\, this data is instrumental for long-term monitoring applications of l and cover change\, soil proprieties\, vegetation productivity\, land degra dation\, vegetation height and other environmental characteristics. The gl obal mosaics were produced via the Time-Series Iteration-free Reconstructi on (TSIRF) framework over the entire Global Land Analysis and Discovery (G LAD) ARD Landsat archive (https://doi.org/10.7717/peerj.18585). Participan ts will learn about the implemented methodologies and use several python l ibraries (stacstac\, scikit-map) JupyterLab. DTSTAMP:20260625T073855Z LOCATION:Room 18 SUMMARY:Accessing global multi-decade Landsat cloud-free time-series in CDS E - Leandro Parente URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/9Q7PZY/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-TETSF9@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T184500 DTEND;TZID=Europe/Amsterdam:20261008T190000 DESCRIPTION:High-resolution information on woody vegetation structure is in creasingly required for biodiversity monitoring\, urban planning\, and env ironmental assessment\, particularly in heterogeneous urban and peri-urban landscapes where trees and shrubs are poorly represented in conventional land-cover products. While Earth observation (EO) data provide new opportu nities to capture fine-scale vegetation structure\, many workflows remain closely tied to specific data environments and lack transparent\, transfer able implementation.\nThis contribution presents an open and reproducible workflow for mapping trees and shrubs across urban areas in Switzerland at national scale. The approach integrates airborne LiDAR-derived canopy hei ght models with authoritative Swiss reference datasets\, including cadastr al data (AV) and the Topographic Landscape Model (TLM) to extract structur ally distinct woody vegetation elements. Object-based segmentation and rul e-based classification are implemented using LAStools and R\, with explici t processing steps designed for transparency and reproducibility.\nThe wor kflow focuses on the delineation of above-ground woody structures\, distin guishing individual trees and shrub patches based on canopy height\, spati al configuration\, and their relationship to reference datasets such as th e national tree inventory (TLM). While airborne LiDAR provides detailed ve rtical information\, the methodological logic can be adapted to alternativ e height sources such as photogrammetric surface models\, stereo imagery\, or emerging spaceborne products\, where LiDAR is unavailable.\nResults fr om the national case study demonstrate how EO-derived above-ground structu ral information can complement existing cadastral and land-use datasets by providing spatially explicit woody vegetation objects in complex urban la ndscapes. Beyond the Swiss application\, this work discusses key considera tions for developing reproducible EO workflows\, including data dependency management\, scalability\, and transferability. The presented workflow ai ms to support the Open-Earth-Monitor community by providing a transparent and adaptable framework for structural habitat mapping using high-resoluti on EO data. DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:From EO Data to Urban Woody Vegetation Structure: A Reproducible Wo rkflow for National-Scale Tree and Shrub Mapping - Natalia Kolecka URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/TETSF9/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-QEER9E@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T184500 DTEND;TZID=Europe/Amsterdam:20261008T190000 DESCRIPTION:Accurate crop field boundary delineation is foundational for ag ricultural mapping\, yield estimation\, and decision support systems. Yet existing AI models\, trained predominantly on data from the Global North\, perform poorly in underrepresented farming systems such as those in Sub-S aharan Africa (typically under 2 hectares\, irregularly shaped) and South America (characterized by shifting cultivation and complex morphologies). This data gap misleads agricultural statistics\, weak policies\, and ineff icient resource allocation. \n\nWe leveraged AI and open-source remote sen sing data to automatically delineate field boundaries in both regions usin g transfer learning\, adapting pretrained global models to local contexts. In South America\, we annotated over 46\,000 field boundaries for model t raining and generated more than 10 million boundaries continent-wide. In E ast Africa's Great Rift Valley\, we automatically detected over 400\,000 f arms from just 6\,000 samples\, incorporating multi-stakeholder annotation workflows and quality assurance pipelines refined from lessons learned in South America. \n\nOur results show that models trained on limited but hi gh-quality local annotations scale effectively to out-of-sample regions. I n Africa\, delineated fields have enabled field level crop type and yield data collection\, in preparation for field level crop type mapping\, yield estimation and monitoring of agroecological and regenerative agriculture practices. In South America\, they have supported supply chain auditing fo r deforestation-free commitments\, EUDR compliance\, country-level crop fo recasting\, and scope 3 emissions estimation. Across both regions\, the ap proach has strengthened national and subnational agricultural data systems and climate resilience frameworks. \n\nBy demonstrating AI model transfer ability across contrasting geographies\, this work charts a pathway toward open\, inclusive\, and scalable Earth observation systems that close crit ical data gaps in the Global South\, positioning AI as a core enabler of s ustainable agricultural monitoring at national and subnational scales. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Democratizing Field Boundary Delineation in the Global South with A I. - Christine Muthee\, Tristan Grupp URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/QEER9E/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-H7T3W3@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T184500 DTEND;TZID=Europe/Amsterdam:20261008T190000 DESCRIPTION:EO foundational models transform satellite images from a space- time grid of raw values into high-dimensional latent spaces called embeddi ngs. These embeddings encode relationships between pixel values and the co rresponding biophysical characteristics. Seasonal crop phenology (plant li fe cycle events)\, urban patterns\, and forest canopy texture are each rep resented in different combinations of embedding dimensions. Researchers us e these embeddings to train lightweight\, downstream models for specific t asks\, such as LULC (land use and land cover) classification\, biomass est imation\, or deforestation detection. These tasks require only a fraction of the computational power and labelled data.\nThe trend is to build massi ve\, global-scale foundational EO models (such as TESSERA or AlphaEarth). Nevertheless\, there is a strong case for developing dedicated regional fo undational models. Global foundation models inherently seek universal stat istical patterns\, pushing representations toward generalised\, highly sim plified categories. A regional foundational model avoids this homogenizati on by optimising representations for local landscapes. By pre-training a f oundation model on regional Earth observation data cubes\, the latent spac e represents those specific regions. This prevents the model\nfrom importi ng spatial biases learned from entirely different continents\, resulting i n much higher-quality embeddings for local downstream tasks.\nThis present ation will show how to build regional EO foundational models using an easy -to-use API associated with the R/Python package SITS. Users can merge var ious sources\, such as optical\, radar\, topographic\, and climate data. T he resulting EO embeddings will be better suited to regional applications than global products. DTSTAMP:20260625T073855Z LOCATION:Room 18 SUMMARY:Regional Earth Observation Foundational Models: Improving Represen tation of Domain-Specific Patterns - Gilberto Camara\, Felipe Carlos\, Rol f Simões\, Alexandre Assunção\, Felipe Souza URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/H7T3W3/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-A98WPB@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261008T190000 DTEND;TZID=Europe/Amsterdam:20261008T191500 DESCRIPTION:Time series and spatial modeling are commonly used to generate cloud- and gap-free satellite imagery. Most existing approaches reconstruc t the entire dataset using advanced models\, which requires high computati onal resources and time. In this study\, we introduce a new\, computationa lly efficient pipeline to reconstruct monthly Landsat data without gaps or clouds. The pipeline includes four levels of gap filling. In the first st ep\, we apply a clean mask to biweekly Landsat data and create a 7-image w eighted window spanning the current and neighbouring months. For each band and month across the 28-year period\, we generate 25th and 75th percentil e thresholds and calculate a weighted median\, giving 50% weight to the cu rrent month and 25% to neighboring months\, using only values within the 2 5th–75th percentiles. In the second step\, remaining gaps are filled usi ng an annual land cover classification derived from the GLAD dataset and L andsat data from up to ten previous years\, restricted to pixels in the sa me land cover class. The third step fills small gaps of up to 2×2 pixels using a 4×4 averaging kernel. These steps fill approximately 40–60% of land pixels depending on tile location. Finally\, a pretrained temporal mo del is applied to fill the remaining gaps. We tested this pipeline on a CP U server with 96 threads and 1 TB RAM. Each tile can be processed in under 2000 seconds. Parallelization across tiles and bands enables global proce ssing in under six weeks\, significantly reducing the computational time c ompared to full dataset reconstruction\, which would take approximately si x months. The resulting dataset provides clean\, gap- and cloud-free month ly Landsat imagery suitable for a variety of research applications. Limita tions remain\, mostly related to input/output operations\, and future work could apply embedding models to reduce dataset size and produce abstract representations for faster access. DTSTAMP:20260625T073855Z LOCATION:Room 18 SUMMARY:A Multi-Layer Gap-Filling Pipeline for Continuous Monthly Landsat D ata (1997–2025) - Sajed URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/A98WPB/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-BCLNKW@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261009T103000 DTEND;TZID=Europe/Amsterdam:20261009T110000 DESCRIPTION:Keynote talk by Coco Antonissen. Abstract and title to be provi ded. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Keynote - Coco Antonissen URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/BCLNKW/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-GCLJMU@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261009T110000 DTEND;TZID=Europe/Amsterdam:20261009T113000 DESCRIPTION:Earth Observation is entering an era of abundance—global sate llite archives\, growing in-situ networks\, and an expanding open-source e cosystem—yet turning this distributed wealth into decision-ready environ mental information remains difficult because data are heterogeneous\, inco mplete\, and hard to combine across sensors\, resolutions\, and regions. T his keynote outlines Earth Embeddings: compact\, AI-native “mental maps ” that summarize what makes a location unique by learning directly from imagery and context. Using an intuitive “Satellite GeoGuessr” contrast ive training setup\, neural networks learn place-specific visual and conte xtual signatures and distill them into dense vectors that can serve as por table location tokens in downstream models\, enabling reuse across tasks a nd regions. This talk will give an overview over different strategies to g enerate embeddings and outline research gaps and steps forward towards glo bal interoperable FAIR embeddings. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Earth Embeddings: Learning “Mental Maps” for Open\, Interoperab le GeoAI - Marc Rußwurm URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/GCLJMU/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-3MAMRS@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261009T113000 DTEND;TZID=Europe/Amsterdam:20261009T120000 DESCRIPTION:Deep learning represents a powerful tool to interpret Earth obs ervation data at large geographic scales. However\, in cases where abundan t reference data is not available and cannot easily be collected\, new app roaches are needed to benefit from this technology. Several Earth observat ion tasks\, especially in environmental remote sensing\, remain challengin g due to the limited number of samples and the geographic and temporal bia s in the reference data. Furthermore\, mapping biophysical variables from single sensor inputs often leads to high ambiguities. Multimodal models pr etrained in a self-supervised fashion promise to overcome such challenges. \n\nIn this talk\, I will first present our recent research project MMEart h-Bench\, a multimodal benchmark dataset for environmental remote sensing. I will discuss our evaluation of existing pretrained models and present o ur test-time adaptation approach that adapts any model at test time using multimodal data to construct adaptation signals. Lastly\, I will present S uperF\, an approach for multi-image super-resolution. This test-time optim ization approach based on implicit neural representations makes use of rep eated observations with sub-pixel shifts and does not require any high-res olution training data. Under a static scene assumption\, it can be applied to super-resolve e.g. Sentinel-2 time series for any place on Earth.\n\n- Personal website: https://langnico.github.io/\n- MMEarth-Bench project: h ttps://mmearth-bench.com/\n- SuperF project: https://sjyhne.github.io/supe rf/ DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Optimizing Representations at Test Time - Nico Lang URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/3MAMRS/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-TJHTLR@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261009T120000 DTEND;TZID=Europe/Amsterdam:20261009T123000 DESCRIPTION:It is estimated that the deep web\, the part of the web that is not indexed through search engines\, accounts for more than 90% of all th e web content. Although in some cases the content may be hidden intentiona lly\, in many others there is just a failure in delivering it to the users \; even after finding the information\, there may still be some challenges in understanding what it covers and accessing it in a suitable format.\n\ nThis talk will elaborate on how geospatial Standards can help us to addre ss these challenges\, and ensure that EO data can live up to the promise o f being used and reused. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:To be FAIR\, we're Open! How open Standards can power Earth Observa tion - Joana Simoes URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/TJHTLR/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-8S3XVA@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261009T123000 DTEND;TZID=Europe/Amsterdam:20261009T130000 DESCRIPTION:Satellite Earth Observation (EO) time series are fundamental to monitoring our planet's changing environment. However\, inconsistent revi sit times and frequent cloud obstruction in optical data (Sentinel-2) ofte n force practitioners to rely on lossy data compositing\, which discards c ritical phenological information.\nIn this keynote\, we introduce TESSERA (Temporal Embeddings of Surface Spectra for Earth Representation and Analy sis)\, a pixel-wise foundation model designed to overcome these challenges . TESSERA leverages multi-modal fusion of Sentinel-1 (radar) and Sentinel- 2 (optical) data\, employing a self-supervised learning framework based on Barlow Twins and random temporal sampling. This approach ensures high rob ustness to irregular sampling and missing data without requiring expensive ground-truth labels.\nA key highlight of TESSERA is its scale and commitm ent to Open Science: trained on a global dataset spanning 2017–2025\, th e model provides high-dimensional temporal embeddings that capture the "sp ectral fingerprint" of the Earth's surface. In alignment with the FAIR pri nciples\, we are committed to making TESSERA an open-access resource for t he community. We will demonstrate how TESSERA achieves state-of-the-art pe rformance in downstream tasks such as crop type mapping and land cover cla ssification with minimal labeled data\, paving the way for the next genera tion of open-source\, distributed GeoAI monitoring systems. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:TESSERA: A Foundation Model for Label-Efficient and Multi-Modal Ear th Observation at Scale - Zhengpeng (Frank) Feng URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/8S3XVA/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-7LHAUP@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261009T133000 DTEND;TZID=Europe/Amsterdam:20261009T140000 DESCRIPTION:Earth observation is generating data at unprecedented scales\, and new analytical methods offer powerful ways to extract insights from it . But realising this potential depends on the infrastructure beneath: how data is stored\, accessed\, and processed—and crucially\, who builds and maintains that infrastructure. Openness is not just a technical preferenc e. It is a practical strategy for sovereignty\, a driver of innovation\, a nd the foundation for communities that can sustain this work over the long term.\nOpenness for sovereignty: In the current political and economic cl imate\, dependence on infrastructure that can become inaccessible\, unaffo rdable\, or restricted is a real risk. Proprietary platforms can change pr icing\, alter terms\, or disappear entirely. The environmental research co mmunity works on challenges spanning decades—climate change\, biodiversi ty loss\, ecosystem degradation. The tools we build today must remain avai lable and adaptable regardless of corporate decisions\, funding changes\, or geopolitical shifts. Open-source software and open standards provide th is guarantee: there is no licence to be revoked\, no single point of failu re\, no dependency on decisions made elsewhere.\nOpenness for innovation: When tools are open\, anyone can extend them. New capabilities emerge beca use the need exists and the ecosystem allows it—no permission required\, no vendor roadmap to wait for. xDGGS\, which brings Discrete Global Grid Systems to the Python ecosystem\, was built by contributors who saw a gap and filled it. Open standards like Zarr mean new tools can interoperate im mediately\, compounding each other's value. This is how innovation actuall y happens: not through proprietary development cycles\, but through commun ities identifying problems and sharing solutions. The pace of improvement accelerates because every contribution benefits everyone.\nOpenness for su stainable communities: Software without a community is software with an ex piration date. Open source survives because people can join\, contribute\, and take ownership. There are no gatekeepers deciding who gets to partici pate. When someone learns from the codebase\, improves it\, and teaches ot hers\, the community grows stronger. Initiatives like the Environmental Da ta Science Book\, the Pangeo community meetings\, and training programmes across Europe are not just about spreading knowledge—they are about ensu ring that the next generation of researchers and developers can maintain a nd extend these tools. Shared ownership means shared responsibility\, and that is what makes infrastructure last.\nThe Pangeo ecosystem: Pangeo embo dies these principles. It provides the toolkit for scalable Earth science —Xarray for labelled arrays\, Dask for distributed computing\, Zarr for cloud-native storage\, xDGGS for grid systems—built by a global communit y and available to all. Pangeo@EOSC\, deployed on European infrastructure through collaboration between EGI and research institutions\, demonstrates that this model works: open tools\, running on open infrastructure\, main tained by a community with shared stakes in its success.\nMaking data anal ysis-ready: Underlying this is the practical challenge of data. Most Earth observation data was not designed for modern workflows—it comes in hete rogeneous formats\, different projections\, inconsistent resolutions. Disc rete Global Grid Systems and cloud-optimised formats like Zarr address thi s by creating common frameworks where data is ready for use the moment it is published. These are open standards that anyone can implement and build upon.\nWhat this talk will cover: Concrete examples of what Pangeo makes possible today\, who is building these tools\, and how openness enables so vereignty\, accelerates innovation\, and grows communities that last. The message is simple: open infrastructure is working\, it is being built by p eople who believe in it\, and there is room for more to join. DTSTAMP:20260625T073855Z LOCATION:Aula Magna SUMMARY:Pangeo: Openness for Sovereignty\, Innovation\, and Sustainable Com munities - Anne Fouilloux URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/7LHAUP/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-JSFE9H@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261009T140000 DTEND;TZID=Europe/Amsterdam:20261009T141500 DESCRIPTION:Elevated forest disturbances and excess tree mortality are incr easingly reported worldwide. Yet existing assessments are either based on patchy terrestrial observations or on large-scale satellite products\, whi ch are limited in resolution to pixel-level\, binary tree loss detection. This leaves a blind spot on fine-scale disturbances where only a few trees are declining in an otherwise intact canopy.\n\nIn this talk\, we give an overview of the deadtrees.earth initiative and how we leveraged crowdsour ced drone data to build globally generalizing models for mapping tree mort ality and disturbances from drones\, airplanes\, and Sentinel-2. This talk will further go into details of our upscaling approach where centimeter-s cale drone data is leveraged to calibrate a model that processes multi-yea r Sentinel-2 time series around the globe. DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:deadtrees.earth - Crowdsourced Drone Data for Global Tree Mortality Maps - Clemens Mosig URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/JSFE9H/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-LEYYXN@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261009T140000 DTEND;TZID=Europe/Amsterdam:20261009T141500 DESCRIPTION:Accurate snow monitoring requires high spatial and temporal res olution to capture rapid processes such as melt and accumulation. However\ , current satellite missions present inherent trade-offs: optical sensors such as Sentinel-2 provide high spatial resolution (tens of meters) but li mited revisit times\, while sensors like MODIS offer daily observations at coarser spatial resolution (∼500 m). In addition\, different sensors re trieve complementary snow properties\, including snow cover extent from op tical data and wet/dry snow conditions from SAR observations. \n\nTo overc ome these limitations\, multi-mission data integration is essential. Furth ermore\, robust estimation of Snow Water Equivalent (SWE) requires the cou pling of remote sensing observations with physically-based or conceptual s now models driven by meteorological forcing. The increasing volume and com plexity of such datasets demand scalable\, cloud-based processing solution s\, particularly for large-scale applications. \n\nIn this contribution\, we present a scalable workflow for large-scale snow water equivalent (SWE) estimation\, aimed at generating daily high-resolution (50 m) SWE data ac ross extensive regions\, such as for example the extratropical Andes withi n the SNOWCOP project and South Tyrol within the Open-Earth-Monitor projec t. The workflow explores alternative cloud-based processing strategies\, i ncluding (i) data access through Copernicus Data Space Ecosystem or other STAC APIs combined with containerized processing environments (Docker)\, e nabling flexible and reproducible workflows without systematic local data download\, and (ii) data-proximate processing using openEO. These compleme ntary approaches allow us to evaluate trade-offs between flexibility\, sca lability\, and computational efficiency for multi-source data fusion and l arge-scale snow monitoring applications. DTSTAMP:20260625T073855Z LOCATION:Room 18 SUMMARY:Large-scale snow monitoring: multi-mission data integration and sca lable processing strategies - Valentina Premier URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/LEYYXN/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-P3DBXT@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261009T141500 DTEND;TZID=Europe/Amsterdam:20261009T143000 DESCRIPTION:Satellite observations from the GRACE mission and its successor GRACE-FO have significantly advanced our ability to monitor terrestrial w ater storage (TWS) at regional to global scales. However\, their limited s patial and temporal resolution hampers the reliable separation of individu al hydrological fluxes\, particularly precipitation. However\, their coars e spatial and temporal resolution makes the individual separation of diffe rent hydrological fluxes from TWS a challenging problem. These limitations in current gravity mission concepts can be addressed by a joint collabora tion between NASA and ESA initiated the Mass-change And Geosciences Intern ational Constellation (MAGIC)\, which can provide enhanced spatio-temporal observations of mass change and therefore enable improved monitoring of h ydrological extremes and dynamics. The primary objective of this work to a ccess how improving the spatial and temporal resolution of future gravity missions impacts precipitation estimation by developing a number of global synthetic experiments. The precipitation data used as forcing of ESM will be compared with the “true” precipitation for testing the reliability of the SM2RAIN approach (Brocca et al.\, 2014) using as input EWH data (i n the past it was implemented by using surface soil moisture data). Simula ted precipitation estimates derived from different gravity mission configu rations (GRACE-C\, NGGM\, and MAGIC) were evaluated against reference prec ipitation to quantify performance improvements. The global correlation ana lysis shows median and mean correlation coefficients of 0.67 and 0.63\, re spectively\, indicating satisfactory performance of the EWH based SM2RAIN framework across most terrestrial regions. Stronger correlations are obser ved over Northern Hemisphere mid-latitudes\, including Europe\, northern A sia\, and North America\, reflecting robust performance in temperate clima tes\, while reduced performance is evident in several tropical regions suc h as central Africa\, parts of the Amazon Basin\, and Southeast Asia. Subs equently\, synthetic experiments were developed using filter and unfiltere d configurations of GRACE-C\, NGGM\, and MAGIC missions. The performance o f NGGM and MAGIC filtered configurations indicates their capability to cap ture precipitation dynamics effectively as compared to unfiltered ones. Th e results of the study clearly highlight the added value of next generati on gravity missions for global hydrological monitoring and develops new sc alable EO based precipitation estimation systems that support emerging ope n and distributed EO infrastructures. The proposed framework enables impro ved assessment of water cycle dynamics as well as enhanced monitoring of h ydrological extremes such as droughts and floods.\n\nReferences\n\nBrocca\ , L.\, Ciabatta\, L.\, Massari\, C.\, Moramarco\, T.\, Hahn\, S.\, Hasenau er\, S.\, Kidd\, R.\, Dorigo\, W.\, Wagner\, W.\, & Levizzani\, V. (2014). Soil as a natural rain gauge: Estimating global rainfall from satellite s oil moisture data. Journal of Geophysical Research: Atmospheres\, 119(9)\, 5128–5141. DTSTAMP:20260625T073855Z LOCATION:Rooms 12+14 SUMMARY:Satellite Gravity Observations for Scalable Global Precipitation Mo nitoring - Muhammad Usman Liaqat URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/P3DBXT/ END:VEVENT BEGIN:VEVENT UID:pretalx-global-workshop-2026-GUPY7L@pretalx.earthmonitor.org DTSTART;TZID=Europe/Amsterdam:20261009T141500 DTEND;TZID=Europe/Amsterdam:20261009T143000 DESCRIPTION:Earth Observation (EO)\, GIS and open geospatial workflows are transforming how biodiversity and ecosystem services can be assessed and a pplied to environmental decision-making. In this talk\, I present an integ rative research framework that combines EO-based mapping\, biodiversity in dicators\, spatial modelling and nature-based solutions to generate ecosys tem service indicators across multiple socioecological contexts. The prese ntation draws on concrete examples from projects in Europe\, Africa and La tin America. These include MaSOT\, which advances the mapping of ecosystem services from Earth Observations\; ASEBIO\, a national-scale assessment o f biodiversity and ecosystem services in Portugal that combines stakeholde r engagement\, scenario analysis and WebGIS tools\; MozambES\, which uses GIS and remote sensing to map mangroves\, assess pressures and support the valuation of mangrove ecosystem services in Mozambique\; and FOREST-LED\, which examines forest loss\, forest expansion and carbon stocks in Spain under global change and natural hazards. Across these projects\, I show ho w taxonomic\, functional and phylogenetic dimensions of biodiversity can b e combined with land-cover dynamics\, ecosystem service indicators and eco nomic valuation to support conservation prioritization and multifunctional landscape management. I also highlight recent studies on biodiversity ind icators of ecosystem services\, ecosystem service change under land-use dy namics\, comparisons between model outputs and stakeholder perceptions\, a nd the integration of eco-environmental factors into landslide susceptibil ity assessment through an eco-DRR perspective. Together\, these examples s how how open and reproducible EO workflows can connect environmental data\ , biodiversity science and applied modelling to produce scalable indicator s for conservation\, risk reduction and sustainability planning. DTSTAMP:20260625T073855Z LOCATION:Room 18 SUMMARY:From Earth Observation to Ecosystem Service Indicators: Integrating Biodiversity\, Spatial Modelling and Nature-Based Solutions Across Scales - Felipe S. Campos URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/GUPY7L/ END:VEVENT END:VCALENDAR