StefaLand: An Efficient Geoscience Foundation Model That Improves Dynamic Land-Surface Predictions

• URL: http://arxiv.org/abs/2509.17942v2
• Date: Sun, 28 Sep 2025 00:59:45 GMT
• Title: StefaLand: An Efficient Geoscience Foundation Model That Improves Dynamic Land-Surface Predictions
• Authors: Nicholas Kraabel, Jiangtao Liu, Yuchen Bian, Daniel Kifer, Chaopeng Shen,
• Abstract summary: Traditional impact models struggle with spatial generalization due to limited observations and concept drift.<n>We introduce StefaLand, a generative earth foundation model centered on landscape interactions.<n>To our knowledge, this is the first geoscience land-surface foundation model that demonstrably improves dynamic land-surface interaction predictions.
• Score: 8.261844935991348
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Stewarding natural resources, mitigating floods, droughts, wildfires, and landslides, and meeting growing demands require models that can predict climate-driven land-surface responses and human feedback with high accuracy. Traditional impact models, whether process-based, statistical, or machine learning, struggle with spatial generalization due to limited observations and concept drift. Recently proposed vision foundation models trained on satellite imagery demand massive compute and are ill-suited for dynamic land-surface prediction. We introduce StefaLand, a generative spatiotemporal earth foundation model centered on landscape interactions. StefaLand improves predictions on four tasks and five datasets: streamflow, soil moisture, and soil composition, compared to prior state-of-the-art. Results highlight its ability to generalize across diverse, data-scarce regions and support broad land-surface applications. The model builds on a masked autoencoder backbone that learns deep joint representations of landscape attributes, with a location-aware architecture fusing static and time-series inputs, attribute-based representations that drastically reduce compute, and residual fine-tuning adapters that enhance transfer. While inspired by prior methods, their alignment with geoscience and integration in one model enables robust performance on dynamic land-surface tasks. StefaLand can be pretrained and finetuned on academic compute yet outperforms state-of-the-art baselines and even fine-tuned vision foundation models. To our knowledge, this is the first geoscience land-surface foundation model that demonstrably improves dynamic land-surface interaction predictions and supports diverse downstream applications.

Related papers

• Utilizing Earth Foundation Models to Enhance the Simulation Performance of Hydrological Models with AlphaEarth Embeddings [48.54230513143053]
  This study examines whether AlphaEarth Foundation embeddings offer a more informative way to describe basin characteristics.<n>We find that models using them achieve higher accuracy when predicting flows in basins not used for training.
  arXiv  Detail & Related papers  (2026-01-04T15:14:16Z)
• Towards Scalable and Generalizable Earth Observation Data Mining via Foundation Model Composition [0.0]
  We investigate whether foundation models pretrained on remote sensing and general vision datasets can be effectively combined to improve performance.<n>The results show that feature-level ensembling of smaller pretrained models can match or exceed the performance of much larger models.<n>The study highlights the potential of applying knowledge distillation to transfer the strengths of ensembles into more compact models.
  arXiv  Detail & Related papers  (2025-06-25T07:02:42Z)
• Synthetic Geology -- Structural Geology Meets Deep Learning [3.216132991084434]
  Building on techniques of generative artificial intelligence applied to voxelated images, we demonstrate a method that extends surface geological data to a three-dimensional subsurface region by training a neural network.<n>We close this data gap in the development of subsurface deep learning by designing a synthetic data-generator process that mimics eons of geological activity.<n>A foundation model trained on such synthetic data is able to generate a 3D image of the subsurface from a previously unseen map of surface topography and geology.
  arXiv  Detail & Related papers  (2025-06-11T20:42:28Z)
• Efficient Self-Supervised Learning for Earth Observation via Dynamic Dataset Curation [67.23953699167274]
  Self-supervised learning (SSL) has enabled the development of vision foundation models for Earth Observation (EO)<n>In EO, this challenge is amplified by the redundancy and heavy-tailed distributions common in satellite imagery.<n>We propose a dynamic dataset pruning strategy designed to improve SSL pre-training by maximizing dataset diversity and balance.
  arXiv  Detail & Related papers  (2025-04-09T15:13:26Z)
• ImplicitTerrain: a Continuous Surface Model for Terrain Data Analysis [14.013976303831313]
  ImplicitTerrain is an implicit neural representation (INR) approach for modeling high-resolution terrain continuously and differentiably. Our experiments demonstrate superior surface fitting accuracy, effective topological feature retrieval, and various topographical feature extraction.
  arXiv  Detail & Related papers  (2024-05-31T23:05:34Z)
• Deep autoregressive modeling for land use land cover [0.0]
  Land use / land cover (LULC) modeling is a challenging task due to long-range dependencies between geographic features and distinct spatial patterns related to topography, ecology, and human development. We identify a close connection between modeling of spatial patterns of land use and the task of image inpainting from computer vision and conduct a study of a modified PixelCNN architecture with approximately 19 million parameters for modeling LULC.
  arXiv  Detail & Related papers  (2024-01-02T18:03:57Z)
• Assessment of a new GeoAI foundation model for flood inundation mapping [4.312965283062856]
  This paper evaluates the performance of the first-of-its-kind geospatial foundation model, IBM-NASA's Prithvi, to support a crucial geospatial analysis task: flood inundation mapping. A benchmark dataset, Sen1Floods11, is used in the experiments, and the models' predictability, generalizability, and transferability are evaluated. Results show the good transferability of the Prithvi model, highlighting its performance advantages in segmenting flooded areas in previously unseen regions.
  arXiv  Detail & Related papers  (2023-09-25T19:50:47Z)
• An evaluation of deep learning models for predicting water depth evolution in urban floods [59.31940764426359]
  We compare different deep learning models for prediction of water depth at high spatial resolution. Deep learning models are trained to reproduce the data simulated by the CADDIES cellular-automata flood model. Our results show that the deep learning models present in general lower errors compared to the other methods.
  arXiv  Detail & Related papers  (2023-02-20T16:08:54Z)
• LOPR: Latent Occupancy PRediction using Generative Models [49.15687400958916]
  LiDAR generated occupancy grid maps (L-OGMs) offer a robust bird's eye-view scene representation. We propose a framework that decouples occupancy prediction into: representation learning and prediction within the learned latent space.
  arXiv  Detail & Related papers  (2022-10-03T22:04:00Z)
• Conditioned Human Trajectory Prediction using Iterative Attention Blocks [70.36888514074022]
  We present a simple yet effective pedestrian trajectory prediction model aimed at pedestrians positions prediction in urban-like environments. Our model is a neural-based architecture that can run several layers of attention blocks and transformers in an iterative sequential fashion. We show that without explicit introduction of social masks, dynamical models, social pooling layers, or complicated graph-like structures, it is possible to produce on par results with SoTA models.
  arXiv  Detail & Related papers  (2022-06-29T07:49:48Z)
• Activation Regression for Continuous Domain Generalization with Applications to Crop Classification [48.795866501365694]
  Geographic variance in satellite imagery impacts the ability of machine learning models to generalise to new regions. We model geographic generalisation in medium resolution Landsat-8 satellite imagery as a continuous domain adaptation problem. We develop a dataset spatially distributed across the entire continental United States.
  arXiv  Detail & Related papers  (2022-04-14T15:41:39Z)
• Embedding Earth: Self-supervised contrastive pre-training for dense land cover classification [61.44538721707377]
  We present Embedding Earth a self-supervised contrastive pre-training method for leveraging the large availability of satellite imagery. We observe significant improvements up to 25% absolute mIoU when pre-trained with our proposed method. We find that learnt features can generalize between disparate regions opening up the possibility of using the proposed pre-training scheme.
  arXiv  Detail & Related papers  (2022-03-11T16:14:14Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.