Geospatial Representation Learning

Follow this topic Ask about this topic

Geospatial representation learning studies how to encode locations and geographic data so that useful spatial structure can be reused across tasks such as image geolocation, land-cover mapping, urban prediction, and mobility modeling. Common approaches include contrastive and masked self-supervision, coordinate and graph encoders, and models that align satellite imagery with maps, text, sensor data, or human activity. These representations can reduce the need for task-specific labels, but their usefulness depends on the data modality, spatial scale, and downstream task. A central challenge is reliable transfer: performance can change substantially across regions and under distribution shifts, so spatially blocked evaluation, calibration, and robustness testing matter alongside accuracy.

Latest papers

133 papers19 papers in the last four weeks, up 12% on the four weeks before. 0.1% of all new papers.

LoCA: Enabling Multi-Modality in Geospatial Location Encoders
Oct 23, 2026Panagiotis Koromilas, Eleanna Vali, Diogenis Antonopoulos +7Geospatial Representation Learning Multimodal Contrastive Learning
University of Athens · The Cyprus Institute · Archimedes, Athena Research Center +7
Earth embeddings are decomposable records of the land surface
Oct 23, 2026Robin Young, Srinivasan KeshavGeospatial Representation Learning
Department of Computer Science and Technology, University of Cambridge, William Gates Building, Cambridge CB3 0FD, UK
GeoLEON-CLIP: Graph-Anchored Multimodal Contrastive Learning for Geographic Region Representation
Oct 23, 2026Radoslaw Malek, Szymon Soltysiak, Aleksandra Kawala-Sterniuk +1Geospatial Representation Learning Multimodal Contrastive Learning
Wrocław University of Science and Technology, Department of Artificial Intelligence, Wrocław, Poland
OlmoEarth v1.3: Embeddings
Oct 23, 2026Gabriel Tseng, Favyen Bastani, Patrick Alan Johnson +13Geospatial Representation Learning Multimodal Embedding
Allen Institute for AI
Towards Consistent Geospatial Representations: Aligning One Encoder to Many Foundation Models
Oct 20, 2026Arthur Ouaknine, David RolnickGeospatial Foundation Models Geospatial Representation Learning
McGill University · Mila – Quebec AI Institute
Emergent Sparsity: How Many Effective Variables Survive Coarse-Graining in AlphaEarth Embeddings?
Oct 18, 2026Johannes HirnDimensionality Reduction Intrinsic Dimensionality
Image Processing Laboratory (IPL), Universitat de València, 46980 Paterna, Valencia, Spain · Universidad Cardenal Herrera-CEU, 46115 Alfara del Patriarca, Valencia, Spain
Learning Administrative Boundary-Aware Location Encoders via Reverse Geocoding
Oct 16, 2026Lakshay Sharma, Alex MarinGeospatial Representation Learning Boundary Representation Modeling
Instacart New York, NY, USA · Thomson Reuters Labs Seattle, WA, USA
CartoVec: General-Purpose Location Embeddings from Open Map Data
Oct 12, 2026Furkan GursoyGeospatial Representation Learning Spatial Context Modeling
Geospatial Vector Geometry as a Native Spatial Modality for Vision-Language Models
Oct 12, 2026Huang Fei, Boyi Chen, Hao Xue +3Vision-Language Models Geospatial Reasoning
The Hong Kong University of Science and Technology (Guangzhou), China · The Hong Kong University of Science and Technology, Hong Kong SAR, China
Learning Holistic Scene Layout against View Shifts in Cross-View Geo-Localization
Oct 9, 2026Chao Yang, Songlian Li, Zhigang TuSpatial Context Modeling Cross-View Geo-Localization
the State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, China
PlacePulse: Learning Transferable Place Representations from Structural Geographic Data
Oct 9, 2026Lucía García-Duarte, Miguel Álvarez, Diego Vicente +2Geospatial Reasoning Benchmarks Geospatial Representation Learning
(CARTO)
Automated Detection and Encoding of Cartographic Signs in Historical Maps
Oct 9, 2026Remi PetitpierreHistorical Document Processing Geospatial Representation Learning
EPFL, Swiss Federal Institute of Technology in Lausanne. Digital Humanities Laboratory (DHLAB), School of Computer and Communication Sciences (IC). CM 1 468, Station 10, 1015 Lausanne, Switzerland. · EPFL. Laboratory of Urban Sociology (LASUR), School of Architecture, Civil and Environmental Engineering (ENAC).
S $^3$ Geo: Structure-Semantic Synergistic Learning for Cross-View Geo-Localization
Oct 8, 2026Ziqian Mo, Hill Zhang, Haosheng Tan +2Cross-View Geo-Localization Geospatial Representation Learning
The Hong Kong University of Science and Technology (Guangzhou) Guangzhou, China · Claremont McKenna College Claremont, California, USA
CETUS: How Far Do Representations Trained on Earth Transfer to Cassini SAR of Titan?
Oct 6, 2026Kevin LeeRemote Sensing Image Understanding Geospatial Representation Learning
University of California at Los Angeles, Los Angeles, CA · NASA Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA
T-JEPA: A Temporal Joint-Embedding Predictive Architecture for Learning Better Remote Sensing Representations
Oct 5, 2026Bowen Peng, Li Liu, Yongxiang Liu +3Geospatial Representation Learning Self-Supervised Time Series Representation Learning
National University of Defense Technology, Changsha, China
LEON: Location Embeddings from OSM Neighborhoods via Hexagonal Graph Masked Autoencoders
Oct 4, 2026Szymon Soltysiak, Radoslaw Malek, Jedrzej Kusnierz +3Masked Autoencoders Geospatial Representation Learning
Department of Artificial Intelligence, Wroclaw University of Science and Technology, Wroclaw, Poland
How to Reduce Localization Ambiguity? Geometry-Semantic Constrained BEV Representation Learning for Satellite-Ground Localization
Sep 30, 2026Junming Feng, Panwang Xia, Qiong Wu +8Representation Learning Cross-View Geo-Localization
The Hong Kong Polytechnic University, Hong Kong · Southern University of Science and Technology · Wuhan University, Wuhan, China +2
Planetary Feature Fields are Scalable Earth Representations
Sep 29, 2026Arjun Rao, Sebastian Loeschcke, Anthony Fuller +3Implicit Neural Representations Geospatial Representation Learning
University of British Columbia · University of Copenhagen · Carleton University +2
When local gains fail to transfer: Frozen Earth-observation embeddings across wildfires
Sep 28, 2026Philipp Stark, Alexandros Sopasakis, Ola HallGeospatial Foundation Models Geospatial Representation Learning
Department of Human Geography Lund University Sölvegatan 10, SE-223 62 Lund, Sweden · Centre for Mathematical Sciences Lund University Sölvegatan 18, SE-221 00 Lund, Sweden
AlphaEarth distinguishes cities but compresses urban variation
Sep 24, 2026Andrew RenningerGeospatial Foundation Models Geospatial Representation Learning
School of Geographical & Earth Sciences, University of Glasgow
Recoverable Geographic Location Information in Earth-Observation Embeddings
Sep 24, 2026Peiwen Zhang, Kristie Hu, Jovana Knezevic +2Geospatial Foundation Models Geospatial Representation Learning
University of Waterloo · University of Cambridge · Aalto University
Annual Earth-observation embeddings encode wildfire disturbance and support simplified burned area mapping
Sep 22, 2026Jovana Knezevic, Clement Atzberger, Zhengpeng Feng +3Remote Sensing Image Segmentation Spatial Generalization
Conservation Research Institute, University of Cambridge, Cambridge, United Kingdom. · Department of Plant Sciences, University of Cambridge, Cambridge, United Kingdom. · Department of Computer Science and Technology, University of Cambridge, Cambridge, United Kingdom. +2
MIND the Gap: A Geographic Implicit Neural Representation with Adjustable Spatial Scale
Sep 21, 2026Isaac Corley, Arjun Rao, Esther Rolf +7Geospatial Reasoning Benchmarks Geospatial Representation Learning
Taylor Geospatial · University of British Columbia · Vector Institute +7
Chronosphere: Space-Time Tessellation of Local Climate Experts
Sep 18, 2026Daniel Cher, Eric Xing, Kexing Li +3Geospatial Representation Learning Representation Learning
Washington University in St. Louis · Taylor Geospatial Institute
MarsRecon: Self-Supervised and Multimodal Surface Representations for Mars
Sep 17, 2026Akshay Naik, Marius F. R. Juston, Jay MahajanGeospatial Representation Learning Self-Supervised Visual Representation Learning
Department of Electrical and Computer Engineering The Grainger College of Engineering University of Illinois Urbana-Champaign Urbana, IL 61801-3080, USA · Department of Computer Science University of Illinois Urbana-Champaign Urbana, IL 61801-3080, USA
From Foundation Embeddings to Cropland Maps: Label Efficiency, Temporal Transferability and Independent Human Validation
Sep 15, 2026Mohammad Ammar Mughees, Giovanni Montefoschi, Zhongxin Chen +1Agricultural Remote Sensing Geospatial Foundation Models
Department of Civil and Environmental Engineering, Politecnico di Milano, Milan, Italy · Food and Agriculture Organization of the United Nations, Rome, Italy
When Correlations Mislead: Confounder-Aware Multi-View Urban Region Representation Learning
Sep 14, 2026Sean Bin Yang, Ying Sun, Zongyi Xu +6Causal Representation Learning Graph Representation Learning
Aalborg University, Aalborg, Denmark · Chongqing University of Posts and Telecommunications, Chongqing, China · East China Normal University, Shanghai, China
Physically Typed and Geometry-Aware Representations for Earth Foundation Models
Sep 12, 2026Rajiv RanjanGeospatial Foundation Models Geospatial Representation Learning
Plaksha University
Geospatial Foundation Models Capture Health-Relevant Dimensions of Place Beyond Conventional Social Risk Indices
Sep 10, 2026Nathaniel Hendrix, Carl Y. Zhang, Chris Heitzig +2Geospatial Foundation Models Geospatial Representation Learning
Center for Professionalism and Value in Health Care, American Board of Family Medicine 1016 16th St NW Ste 800 Washington, DC 20036 United States of America · Stanford Center on Longevity, Stanford University Palo Alto, CA USA · Center for Population Health Sciences, School of Medicine, Stanford University Palo Alto, California, United States of America +1
Applying foundation model embeddings towards urban livability evaluation
Sep 8, 2026Ayush Khot, Wen Zhou, Shaowen WangFeature Attribution Geospatial Foundation Models
Siebel School of Computing and Data Science, University of Illinois at Urbana-Champaign, Urbana, IL, USA · Department of Geography and Geographic Information Science, University of Illinois at Urbana-Champaign, Urbana, IL, USA · Department of Geography and Geographic Information Science & Siebel School of Computing and Data Science, University of Illinois at Urbana-Champaign, Urbana, IL, USA

Oct 23, 2026J/K move · Enter open · S save

LoCA: Enabling Multi-Modality in Geospatial Location Encoders

Panagiotis Koromilas, Eleanna Vali, Diogenis Antonopoulos, Eleni Saka+6

University of Athens · The Cyprus Institute · Archimedes, Athena Research Center+7

Geospatial Representation Learning Multimodal Contrastive Learning Remote Sensing

Open paper AskPDF arXiv

Location encoders map geographic coordinates to high-dimensional descriptors through contrastive alignment with paired Earth Observation imagery. We prove that this construction is structurally limited: any Lipschitz location encoder produces an output of intrinsic dimension at most 2, regardless of architecture or bandwidth, which forces the contrastive loss to strictly exceed its asymptotic optimum. Empirically, joint training also collapses the encoder's embedding covariance to low rank. Although dimension-limited, current location encoders sidestep this spectral collapse by training against a frozen image encoder, which we reinterpret as a covariance regularizer that supplies a high-rank target for the location encoder to fold onto. However, extending this approach to multimodal inputs is challenging, as location encoders should simultaneously map a single location to multiple, potentially incompatible, descriptors. We identify two requirements a regularizer must satisfy for location encoding in multimodal regimes: co-alignment across modalities, since adapters alone cannot align independently pretrained encoders; and high effective rank, since the location encoder inherits its target's spectrum. We propose LoCA (Location Contrastive Alignment), a two-stage pipeline that satisfies both. Stage 1 jointly trains Sentinel-1 and Sentinel-2 encoders with contrastive learning, producing a co-aligned, high-rank multimodal target. Stage 2 freezes the image encoders and trains a location and a tabular metadata encoder as projection adapters with multi-view contrastive learning. Against five state-of-the-art encoders on 17 tasks ranging from land cover and wildfire forecasting to geophysical, socio-economic, biodiversity and climate applications, LoCA achieves the most task wins, is the only model to outperform every other on a majority of tasks, and wins 68.8% of all per-task comparisons.