Interpolation of GEDI Biomass Estimates with Calibrated Uncertainty Quantification

2026-01-23

AI Summary

This study addresses the limitations of existing machine learning approaches in interpolating sparse GEDI LiDAR observations to produce continuous biomass maps, which often fail to yield well-calibrated prediction intervals and overlook spatial heterogeneity in local context and uncertainty. To overcome these challenges, the work introduces Attention Neural Processes (ANPs) into remote sensing–based biomass mapping for the first time, integrating geospatial foundation model embeddings with probabilistic meta-learning to dynamically model spatial covariance and disentangle epistemic from aleatoric uncertainty. Evaluated across five major biomes, the proposed method achieves accuracy comparable to state-of-the-art models while delivering near-perfect uncertainty calibration. Furthermore, it enables few-shot cross-regional adaptation, rapidly recovering high predictive performance with only a handful of local observations.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Remote Sensing / Geospatial AI
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Reliable wall-to-wall biomass density estimation from NASA's GEDI mission requires interpolating sparse LIDAR observations across heterogeneous landscapes. While machine learning approaches like Random Forest and XGBoost are widely used, they treat spatial predictions of GEDI observations from multispectral or SAR remote sensing data as independent without adapting to the varying difficulty of heterogeneous landscapes. We demonstrate these approaches generally fail to produce calibrated prediction intervals. We show that this stems from conflating ensemble variance with aleatoric uncertainty and ignoring local spatial context. To resolve this, we introduce Attentive Neural Processes (ANPs), a probabilistic meta-learning architecture that explicitly conditions predictions on local observation sets and exploits geospatial foundation model embeddings. Unlike static ensembles, ANPs learn a flexible spatial covariance function, allowing estimates to be more uncertain in complex landscapes and less in homogeneous areas. We validate this approach across five distinct biomes ranging from tropical Amazonian forests to boreal, temperate, and alpine ecosystems, demonstrating that ANPs achieve competitive accuracy while maintaining near-ideal uncertainty calibration. We demonstrate the operational utility of the method through few-shot adaptation, where the model recovers most of the performance gap in cross-region transfer using minimal local data. This work provides a scalable, theoretically rigorous alternative to ensemble variance for continental scale earth observation.

Problem

• biomass mapping
• uncertainty calibration
• spatial interpolation
• heterogeneous landscapes
• GEDI

Innovation

• Attentive Neural Processes
• uncertainty calibration
• spatial interpolation
• geospatial foundation models
• probabilistic meta-learning