Earth Observation (EO) Foundation Modelling (FM) holds great promise for simplifying and improving the use of EO data for diverse real-world tasks. However, most existing models require additional adaptation before they can be used and are structured rigidly around particular data sources or training approaches. To address this, we take inspiration from large language models, where diverse tasks, both pre-training and downstream, are implicitly captured through next-token prediction over unified token sequences, leveraging the structure and diversity of the training data. We introduce SITS-DECO (Satellite Image Time Series-DECoder Only), a proof-of-concept generative model that applies this unified-sequence framing to EO data. Using a simple GPT-style decoder-only architecture, and demonstrate its ability to perform useful EO tasks (pixel-wise, multi-temporal, multi-modal crop-type classification) in a purely generative framework. Through symbolic prompting, we show that the model can perform multiple supervised and self-supervised tasks within a single unified architecture, without task- or modality-specific adaptation. Despite its simplicity and lack of spatial context, SITS-DECO outperforms much larger EO foundation models on crop-type classification (PASTIS-R) demonstrating that dense temporal sequence modelling is a critical missing ingredient in the current paradigm. This work exemplifies a data-centric modelling paradigm in which capability arises from the diversity and structure of the training data rather than from architectural complexity. SITS-DECO provides a lightweight, practical route to multi-modal, multi-task EO modelling, and a conceptual bridge toward future generative EO foundation models.

The Mar Menor, Europe's largest coastal lagoon, located in Spain, has undergone severe eutrophication crises. Monitoring chlorophyll-a (Chl-a) is essential to anticipate harmful algal blooms and guide mitigation. Traditional in situ measurements are spatially and temporally limited. Satellite-based approaches provide a more comprehensive view, enabling scalable, long-term, and transferable monitoring. This study aims to overcome limitations of chlorophyll monitoring, often restricted to surface estimates or limited temporal coverage, by developing a reliable methodology to predict and map Chl-a across the water column of the Mar Menor. The work integrates Sentinel 2 imagery with buoy-based ground truth to create models capable of high-resolution, depth-specific monitoring, enhancing early-warning capabilities for eutrophication. Nearly a decade of Sentinel 2 images was atmospherically corrected using C2RCC processors. Buoy data were aggregated by depth (0-1 m, 1-2 m, 2-3 m, 3-4 m). Multiple ML and DL algorithms-including RF, XGBoost, CatBoost, Multilater Perceptron Networks, and ensembles-were trained and validated using cross-validation. Systematic band-combination experiments and spatial aggregation strategies were tested to optimize prediction. Results show depth-dependent performance. At the surface, C2X-Complex with XGBoost and ensemble models achieved R2 = 0.89; at 1-2 m, CatBoost and ensemble models reached R2 = 0.87; at 2-3 m, TOA reflectances with KNN performed best (R2 = 0.81); while at 3-4 m, RF achieved R2 = 0.66. Generated maps successfully reproduced known eutrophication events (e.g., 2016 crisis, 2025 surge), confirming robustness. The study delivers an end-to-end, validated methodology for depth-specific Chl-amapping. Its integration of multispectral band combinations, buoy calibration, and ML/DL modeling offers a transferable framework for other turbid coastal systems.

F oundation models constitute a significant advancement in computer vision: after a single, albeit costly, training phase, they can address a wide array of tasks. In the field of Earth observation, over 75 remote sensing vision foundation models have been developed in the past four years. However, none has consistently outperformed the others across all available downstream tasks. T o facilitate their comparison, we propose a cost-effective method for predicting a model's performance on multiple downstream tasks without the need for fine-tuning on each one. This method is based on what we call "capabilities encoding. " The utility of this novel approach is twofold: we demonstrate its potential to simplify the selection of a foundation model for a given new task, and we employ it to offer a fresh perspective on the existing literature, suggesting avenues for future research.

The Leaf Area Index (LAI) is a critical parameter to understand ecosystem health and vegetation dynamics. In this paper, we propose a novel method for pixel-wise LAI prediction by leveraging the complementary information from Sentinel 1 radar data and Sentinel 2 multi-spectral data at multiple timestamps. Our approach uses a deep neural network based on multiple U-nets tailored specifically to this task. To handle the complexity of the different input modalities, it is comprised of several modules that are pre-trained separately to represent all input data in a common latent space. Then, we fine-tune them end-to-end with a common decoder that also takes into account seasonality, which we find to play an important role. Our method achieved 0.06 RMSE and 0.93 R2 score on publicly available data. We make our contributions available at https://github.com/valentingol/LeafNothingBehind for future works to further improve on our current progress.

Analysis of the satellite images by Soar Earth, an Australian startup that aggregates satellite imagery and crowdsourced maps into an online digital atlas, suggests that the workers have already excavated around 26 million cubic meters of earth and rock--78 times the volume of the world's tallest building, the Burj Khalifa. Official drone footage of The Line's construction site, released in October, indeed showed fleets of bulldozers, trucks, and diggers excavating its foundations. Visit The Line's location on Google Maps and Google Earth, however, and you will see little more than bare rock and sand. The strange gap in imagery raises questions about who gets to access high-res satellite technology. And if the largest urban construction site on the planet doesn't appear on Google Maps, what else can't we see? The Line is as controversial as it is futuristic.

Analyzing the planet at scale with satellite imagery and machine learning is a dream that has been constantly hindered by the cost of difficult-to-access highly-representative high-resolution imagery. To remediate this, we introduce here the WorldStrat dataset. The largest and most varied such publicly available dataset, at Airbus SPOT 6/7 satellites' high resolution of up to 1.5 m/pixel, empowered by European Space Agency's Phi-Lab as part of the ESA-funded QueryPlanet project, we curate nearly 10,000 sqkm of unique locations to ensure stratified representation of all types of land-use across the world: from agriculture to ice caps, from forests to multiple urbanization densities. We also enrich those with locations typically under-represented in ML datasets: sites of humanitarian interest, illegal mining sites, and settlements of persons at risk. We temporally-match each high-resolution image with multiple low-resolution images from the freely accessible lower-resolution Sentinel-2 satellites at 10 m/pixel. We accompany this dataset with an open-source Python package to: rebuild or extend the WorldStrat dataset, train and infer baseline algorithms, and learn with abundant tutorials, all compatible with the popular EO-learn toolbox. We hereby hope to foster broad-spectrum applications of ML to satellite imagery, and possibly develop from free public low-resolution Sentinel2 imagery the same power of analysis allowed by costly private high-resolution imagery. We illustrate this specific point by training and releasing several highly compute-efficient baselines on the task of Multi-Frame Super-Resolution. High-resolution Airbus imagery is CC BY-NC, while the labels and Sentinel2 imagery are CC BY, and the source code and pre-trained models under BSD. The dataset is available at https://zenodo.org/record/6810792 and the software package at https://github.com/worldstrat/worldstrat .

Satellite images are snapshots of the Earth surface. We propose to forecast them. We frame Earth surface forecasting as the task of predicting satellite imagery conditioned on future weather. EarthNet2021 is a large dataset suitable for training deep neural networks on the task. It contains Sentinel 2 satellite imagery at 20m resolution, matching topography and mesoscale (1.28km) meteorological variables packaged into 32000 samples. Additionally we frame EarthNet2021 as a challenge allowing for model intercomparison. Resulting forecasts will greatly improve (>x50) over the spatial resolution found in numerical models. This allows localized impacts from extreme weather to be predicted, thus supporting downstream applications such as crop yield prediction, forest health assessments or biodiversity monitoring. Find data, code, and how to participate at www.earthnet.tech

Deep learning methods have received increasing interest by the remote sensing community for multi-temporal land cover classification in recent years. Convolutional Neural networks that elementwise compare a time series with learned kernels, and recurrent neural networks that sequentially process temporal data have dominated the state-of-the-art in the classification of vegetation from satellite time series. Self-attention allows a neural network to selectively extract features from specific times in the input sequence thus suppressing non-classification relevant information. Today, self-attention based neural networks dominate the state-of-the-art in natural language processing but are hardly explored and tested in the remote sensing context. In this work, we embed self-attention in the canon of deep learning mechanisms for satellite time series classification for vegetation modeling and crop type identification. We compare it quantitatively to convolution, and recurrence and test four models that each exclusively relies on one of these mechanisms. The models are trained to identify the type of vegetation on crop parcels using raw and preprocessed Sentinel 2 time series over one entire year. To obtain an objective measure we find the best possible performance for each of the models by a large-scale hyperparameter search with more than 2400 validation runs. Beyond the quantitative comparison, we qualitatively analyze the models by an easy-to-implement, but yet effective feature importance analysis based on gradient back-propagation that exploits the differentiable nature of deep learning models. Finally, we look into the self-attention transformer model and visualize attention scores as bipartite graphs in the context of the input time series and a low-dimensional representation of internal hidden states using t-distributed stochastic neighborhood embedding (t-SNE).