Newsletter N°12 – May 2025

Spectral Image Data Fusion for Multisource Data Augmentation

🧑 Roberta Iuliana Luca | Alexandra Băicoianu | Ioana Cristina Plajer
📅 2025

Multispectral and hyperspectral images are increasingly popular in different research fields, such as remote sensing, astronomical imaging, or precision agriculture. However, the amount of free data available to perform machine learning tasks is relatively small. Moreover, artificial intelligence models developed in the area of spectral imaging require input images with a fixed spectral signature, expecting the data to have the same number of spectral bands or the same spectral resolution. This requirement significantly reduces the number of usable sources that can be used for a given model. The scope of this study is to introduce a methodology for spectral image data aggregation, in order to allow machine learning models to be trained and/or used on data from a larger number of sources, thus providing better generalization. For this purpose, we propose different interpolation techniques, in order to make multisource spectral data compatible with each other. The interpolation outcomes are evaluated through various approaches. This includes direct assessments using surface plots and metrics such as a Custom Mean Squared Error and the Normalized Difference Vegetation Index. Additionally, indirect evaluation is done by estimating their impact on machine learning model training, particularly for semantic segmentation.

Luca, R.I., Baicoianu, A. and Plajer, I.C., 2025. Spectral image data aggregation for multisource data augmentation. European Journal of Remote Sensing, 58(1), p.2492295.

https://www.tandfonline.com/doi/full/10.1080/22797254.2025.2492295

bibtex

TerraMind: Large-Scale Generative Multimodality for Earth Observation

🧑 Johannes Jakubik et al.
📅 2025

“We present TerraMind, the first any-to-any generative, multimodal foundation model for Earth observation (EO). Unlike other multimodal models, TerraMind is pretrained on dual-scale representations combining both token-level and pixel-level data across modalities. On a token level, TerraMind encodes high-level contextual information to learn cross-modal relationships, while on a pixel level, TerraMind leverages fine-grained representations to capture critical spatial nuances. We pretrained TerraMind on nine geospatial modalities of a global, large-scale dataset. In this paper, we demonstrate that (i) TerraMind’s dual-scale early fusion approach unlocks a range of zero-shot and few-shot applications for Earth observation, (ii) TerraMind introduces “Thinking-in-Modalities” (TiM) — the capability of generating additional artificial data during finetuning and inference to improve the model output — and (iii) TerraMind achieves beyond state-of-the-art performance in community-standard benchmarks for EO like PANGAEA. The pretraining dataset, the model weights, and our code is open-sourced under a permissive license.”

https://doi.org/10.48550/arXiv.2504.11171

Call for Projects: Postdoctoral Contracts in Occitanie for O3T Project

📅 Deadline: July 21, 2025

A new call is open to fund postdoctoral research contracts in the Occitanie region, as part of the multidisciplinary project O3T – Earth Observation and Territories in Transition.

The projects must address at least one of the following three research axes:

• Low-cost instrumentation (e.g. field measurements, affordable sensors)
• Data retrieval and exploitation (storage, processing, analysis, visualization, sharing)
• Operationalization, awareness-raising, and co-construction of data and observatories with territories (collaborative frameworks between researchers and local actors)

Targeted environments include urban, rural, coastal, and mountain areas. Projects must respond to concrete operational needs tied to territorial transitions and demonstrate strong practical relevance.

Priority will be given to initiatives that include a 2–6 month immersion in an operational structure (e.g. local authority, NGO, national park, state agency, or social sciences lab active in fieldwork).

More info: https://o3t.univ-toulouse.fr/lancement-dun-nouvel-aap-pour-le-financement-de-contrats-postdoctoraux/

PRIMA 2025 Calls for Proposals – Funding Opportunities Across the Mediterranean sea

📅 Deadlines: 15 July 2025 ( Section 1) | 8 July 2025 (Section 2)

The 2025 calls for proposals under the PRIMA program (Partnership for Research and Innovation in the Mediterranean Area) are now open, supporting research and innovation tackling water management and agri-food systems in the Mediterranean.

Key Themes:

• Water management: Advanced digital tools (AI, IoT) for adaptive management, with a focus on impact assessments for policy development
• Agricultural systems: Precision farming for smallholders, promoting biodiversity, soil health, and inclusive practices
• Agri-food value chains: Encouraging sustainable consumer choices, expanding plant-based protein options, and improving traceability

Funding Structure:

• Section 1 (EU-funded): 3 thematic calls, ~12 projects expected, €2.7M each, deadline: 15 July 2025
• Section 2 (nationally funded): 1 cross-cutting call, ~20 projects expected, €1.5M each, deadline: 8 July 2025

https://prima-med.org/funding-opportunities-prima-2025s-calls-are-now-open/

Reprocessing Sentinel-2 data: A new step towards better quality

As part of the THEIA program, the CDS THEIA team at CNES announces the progressive release of data derived from the reprocessing of the Sentinel-2 level 2A archive on the GEODES platform. This reprocessing covers over nine years of acquisitions, from July 4, 2015, to December 31, 2024. The process starts with the European zone and will extend to other regions, including India, Africa, and the Americas. The goal is to improve data quality by correcting location errors and optimizing multi-temporal overlap.

Access tools are being developed to facilitate downloading the reprocessed data.

https://www.theia-land.fr/blog/2025/03/28/retraitement-en-cours-des-donnees-sentinel-2-une-mise-a-disposition-progressive-sur-geodes/

Sentinel-1 captures ground shift from Myanmar earthquake

“On 28 March 2025, a powerful magnitude 7.7 earthquake struck central Myanmar, sending shockwaves through the region. While the country is still dealing with the devasting aftermath, scientists have used radar images from the Copernicus Sentinel-1 satellites to reveal a detailed picture of how the ground shifted as a result of the quake – offering new insights into the mechanics of the tectonic Sagaing Fault and the scale of the seismic rupture.“

https://www.esa.int/Applications/Observing_the_Earth/Copernicus/Sentinel-1/Sentinel-1_captures_ground_shift_from_Myanmar_earthquake

Biomass satellite launched aboard a Vega-C rocket from Kourou, French Guiana

“ESA’s groundbreaking Biomass satellite, designed to provide unprecedented insights into the world’s forests and their crucial role in Earth’s carbon cycle, has been launched. The satellite lifted off aboard a Vega-C rocket from Europe’s Spaceport in Kourou, French Guiana, on 29 April at 11:15 CEST (06:15 local time).“

https://www.esa.int/Applications/Observing_the_Earth/FutureEO/Biomass/Biomass_launched_to_count_forest_carbon#msdynmkt_trackingcontext=ef7541a4-9cec-46bd-9b0d-a56c04ec0200