EcoMapper: Generative Modeling for Climate-Aware Satellite Imagery

Muhammed Goktepe, Amir hossein Shamseddin, Erencan Uysal, Javier Muinelo Monteagudo, Lukas Drees, Aysim Toker, Senthold Asseng, Malte von Bloh

TL;DR: We present a novel dataset and two generative approaches using fine-tuned Stable Diffusion 3 models to create realistic satellite imagery conditioned on climate and land cover data, enabling single-image and conditional time-series generation.

Abstract:

Satellite imagery is essential for Earth observation, enabling applications like crop yield prediction, environmental monitoring, and climate
change assessment. However, integrating satellite imagery with climate data remains a challenge, limiting its utility for forecasting and scenario analysis. We introduce a novel dataset of 2.9 million Sentinel-2 images spanning 15 land cover types with corresponding climate records, forming the foundation for two satellite image generation approaches using fine-tuned Stable Diffusion 3 models. The first is a text-to-image generation model that uses textual prompts with climate and land cover details to produce realistic synthetic imagery for specific regions. The second leverages ControlNet for multi-conditional image generation, preserving spatial structures while mapping climate data or generating time-series to simulate landscape evolution. By combining synthetic image generation with climate and land cover data, our work advances generative modeling in remote sensing, offering realistic inputs for environmental forecasting and new possibilities for climate adaptation and geospatial analysis.

Lay Summary:

Our work introduces EcoMapper, a generative machine learning model that creates realistic satellite images based on climate data. Trained on over 2.9 million satellite images linked with weather records, EcoMapper can simulate how specific landscapes (e.g. cropland or forest) might look under different weather or climate scenarios - whether next week's forecast or the conditions expected in 2050.
EcoMapper generates synthetic satellite images that reflect expected changes in vegetation, land cover, or surface conditions. It can also fill data gaps caused by clouds and extend satellite monitoring into the future - without needing new observations.
By connecting climate projections with satellite imagery, EcoMapper helps scientists, policymakers, and planners explore the effects of both short-term weather events and long-term climate change, improving environmental forecasting, disaster preparedness, and sustainable land use strategies.

Application-Driven Machine Learning: This submission is on Application-Driven Machine Learning.

Link To Code: https://github.com/maltevb/ecomapper

Primary Area: Applications->Chemistry, Physics, and Earth Sciences

Keywords: remote sensing, generative modeling, stable diffusion, climate, satellite imagery, computer vision

Submission Number: 16158