# TerraFM: A Scalable Foundation Model for Unified Multisensor Earth Observation

arXiv: https://arxiv.org/abs/2506.06281 · HTML: https://mlsanity.com/paper/2506.06281/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 36 references that are in the map.
- First posted on arXiv 6 Jun 2025.

## Builds on (most cited first, 24 of 24)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Unified Perceptual Parsing for Scene Understanding](https://mlsanity.com/paper/1807.10221/) (2018)
- [GeoChat: Grounded Large Vision-Language Model for Remote Sensing](https://mlsanity.com/paper/2311.15826/) (2023)
- [Scale-MAE: A Scale-Aware Masked Autoencoder for Multiscale Geospatial Representation Learning](https://mlsanity.com/paper/2212.14532/) (2022)
- [SkySense: A Multi-Modal Remote Sensing Foundation Model Towards Universal Interpretation for Earth Observation Imagery](https://mlsanity.com/paper/2312.10115/) (2023)
- [FloodNet: A High Resolution Aerial Imagery Dataset for Post Flood Scene Understanding](https://mlsanity.com/paper/2012.02951/) (2020)
- [CROMA: Remote Sensing Representations with Contrastive Radar-Optical Masked Autoencoders](https://mlsanity.com/paper/2311.00566/) (2023)
- [SatlasPretrain: A Large-Scale Dataset for Remote Sensing Image Understanding](https://mlsanity.com/paper/2211.15660/) (2022)
- [Prithvi-EO-2.0: A Versatile Multi-Temporal Foundation Model for Earth Observation Applications](https://mlsanity.com/paper/2412.02732/) (2024)
- [SkyScript: A Large and Semantically Diverse Vision-Language Dataset for Remote Sensing](https://mlsanity.com/paper/2312.12856/) (2023)
- [AnySat: One Earth Observation Model for Many Resolutions, Scales, and Modalities](https://mlsanity.com/paper/2412.14123/) (2024)
- [Towards a Unified Copernicus Foundation Model for Earth Vision](https://mlsanity.com/paper/2503.11849/) (2025)
- [Cross-Scale MAE: A Tale of Multi-Scale Exploitation in Remote Sensing](https://mlsanity.com/paper/2401.15855/) (2024)
- [On Data Scaling in Masked Image Modeling](https://mlsanity.com/paper/2206.04664/) (2022)
- [Bridging Remote Sensors with Multisensor Geospatial Foundation Models](https://mlsanity.com/paper/2404.01260/) (2024)
- [Next Day Wildfire Spread: A Machine Learning Data Set to Predict Wildfire Spreading from Remote-Sensing Data](https://mlsanity.com/paper/2112.02447/) (2021)
- [Decoupling Common and Unique Representations for Multimodal Self-supervised Learning](https://mlsanity.com/paper/2309.05300/) (2023)
- [Panopticon: Advancing Any-Sensor Foundation Models for Earth Observation](https://mlsanity.com/paper/2503.10845/) (2025)
- [Major TOM: Expandable Datasets for Earth Observation](https://mlsanity.com/paper/2402.12095/) (2024)
- [Better, Not Just More: Data-Centric Machine Learning for Earth Observation](https://mlsanity.com/paper/2312.05327/) (2023)
- [Fus-MAE: A cross-attention-based data fusion approach for Masked Autoencoders in remote sensing](https://mlsanity.com/paper/2401.02764/) (2024)

## Built on by (most cited first, 25 of 25)

- [OpenEarthAgent: A Unified Framework for Tool-Augmented Geospatial Agents](https://mlsanity.com/paper/2602.17665/) (2026)
- [No One Knows the State of the Art in Geospatial Foundation Models](https://mlsanity.com/paper/2605.12678/) (2026)
- [EarthShift: a benchmark for measuring robustness to real-world distribution shifts in Earth observation](https://mlsanity.com/paper/2605.29330/) (2026)
- [PRUE: A Practical Recipe for Field Boundary Segmentation at Scale](https://mlsanity.com/paper/2603.27101/) (2026)
- [UniverSat: Resolution- and Modality-Agnostic Transformers for Earth Observation](https://mlsanity.com/paper/2606.23503/) (2026)
- [GeoCrossBench: Cross-Band Generalization for Remote Sensing](https://mlsanity.com/paper/2511.02831/) (2025)
- [Any Model, Any Place, Any Time: Get Remote Sensing Foundation Model Embeddings On Demand](https://mlsanity.com/paper/2602.23678/) (2026)
- [A Unified Foundation Model for All-in-One Multi-Modal Remote Sensing Image Restoration and Fusion with Language Prompting](https://mlsanity.com/paper/2604.05629/) (2026)
- [MAESTRO: Masked AutoEncoders for Multimodal, Multitemporal, and Multispectral Earth Observation Data](https://mlsanity.com/paper/2508.10894/) (2025)
- [Brewing Stronger Features: Dual-Teacher Distillation for Multispectral Earth Observation](https://mlsanity.com/paper/2602.19863/) (2026)
- [GeoSANE: Learning Geospatial Representations from Models, Not Data](https://mlsanity.com/paper/2603.23408/) (2026)
- [GeoMeld: Toward Semantically Grounded Foundation Models for Remote Sensing](https://mlsanity.com/paper/2604.10591/) (2026)
- [SpectralEarth-FM: Bringing Hyperspectral Imagery into Multimodal Earth Observation Pretraining](https://mlsanity.com/paper/2605.21075/) (2026)
- [SLED: Scalable Location Encoding via Distillation](https://mlsanity.com/paper/2608.06612/) (2026)
- [Assessing the value of Geo-Foundational Models for Flood Inundation Mapping: Benchmarking models for Sentinel-1, Sentinel-2, and Planetscope for end-users](https://mlsanity.com/paper/2511.01990/) (2025)
- [COP-GEN: Latent Diffusion Transformer for Copernicus Earth Observation Data](https://mlsanity.com/paper/2603.03239/) (2026)
- [Spatial Representation Learning Beyond Pixels: Unifying Raster Data and Vector Semantics for Human-Centric Geospatial Foundation Models](https://mlsanity.com/paper/2606.02374/) (2026)
- [Beyond Accuracy: Assessing Calibration of Geospatial Foundation Models and Their Sensitivity to Distribution Shifts](https://mlsanity.com/paper/2608.16614/) (2026)
- [Neural Processes Maintain Calibrated Biomass Estimates Across Spatiotemporal Gaps and Disturbance](https://mlsanity.com/paper/2604.03874/) (2026)
- [Multimodal Large Language Models for Remote Sensing Image Understanding: Domain-Specific or General-Purpose?](https://mlsanity.com/paper/2607.20284/) (2026)
- [Multi-Year Geospatial Reasoning using Interannually-Consistent Historical Predictions as a Free Input Modality](https://mlsanity.com/paper/2608.05979/) (2026)
- [Earth Observation Foundation Models for Terrestrial Ecohydrology: From Representation Learning to Process Inference](https://mlsanity.com/paper/2608.15282/) (2026)
- [AlignJEPA: Predictive Vision-Language Alignment for Remote Sensing Foundation Models](https://mlsanity.com/paper/2608.15456/) (2026)
- [Lightweight Adaptation of General-Purpose VLMs for Multispectral and SAR Image Understanding](https://mlsanity.com/paper/2609.02187/) (2026)
- [Autoregressive Drillhole Modelling Under Distribution Shift](https://mlsanity.com/paper/2610.01204/) (2026)

## Abstract (arXiv)

Modern Earth observation (EO) increasingly leverages deep learning to harness the scale and diversity of satellite imagery across sensors and regions. While recent foundation models have demonstrated promising generalization across EO tasks, many remain limited by the scale, geographical coverage, and spectral diversity of their training data, factors critical for learning globally transferable representations. In this work, we introduce TerraFM, a scalable self-supervised learning model that leverages globally distributed Sentinel-1 and Sentinel-2 imagery, combined with large spatial tiles and land-cover aware sampling to enrich spatial and semantic coverage. By treating sensing modalities as natural augmentations in our self-supervised approach, we unify radar and optical inputs via modality-specific patch embeddings and adaptive cross-attention fusion. Our training strategy integrates local-global contrastive learning and introduces a dual-centering mechanism that incorporates class-frequency-aware regularization to address long-tailed distributions in land cover.TerraFM achieves strong generalization on both classification and segmentation tasks, outperforming prior models on GEO-Bench and Copernicus-Bench. Our code and pretrained models are publicly available at: https://github.com/mbzuai-oryx/TerraFM .