article · Scientific Reports
Machine learning models applied to Earth observation data face inherent uncertainties that can compromise decisions if left unquantified. A review of Earth observation datasets indicates that only 22.5 percent include any uncertainty information, with unreliable estimation methods widely prevalent. Conformal prediction addresses these shortcomings by delivering statistically valid prediction regions across arbitrary machine learning models and data distributions. To overcome the bottleneck of moving massive geospatial datasets to external algorithms, native modules were developed directly within Google Earth Engine. This allows seamless integration of uncertainty quantification into existing traditional and deep learning workflows. The framework has been successfully tested on both classification and regression tasks across local to global scales, facilitating reliable uncertainty reporting for critical environmental analyses.
Earth observation datasets increasingly inform international accords, operational monitoring, and risk management. Without dependable uncertainty measurements, automated machine learning models can yield misleading predictions with adverse consequences. Providing accessible, mathematically sound uncertainty bounds enables analysts, environmental agencies, and policymakers to understand the confidence limits of automated geospatial predictions before making high-stakes decisions.
The work delivers applied and tested software modules natively integrated into Google Earth Engine, allowing immediate uptake by Earth observation platforms and remote sensing analysts. Potential commercial users include environmental consultancies, agricultural monitoring firms, and geospatial intelligence providers seeking rigorous confidence intervals for their operational mapping products without needing to transfer vast volumes of satellite data to external servers.
AI-generated from the published abstract. Always read the original work before citing.
Machine learning is increasingly applied to Earth Observation (EO) data to obtain datasets that contribute towards international accords. However, these datasets contain inherent uncertainty that needs to be quantified reliably to avoid negative consequences. In response to the increased need to report uncertainty, we bring attention to the promise of conformal prediction within the domain of EO. Unlike previous uncertainty quantification methods, conformal prediction offers statistically valid prediction regions while concurrently supporting any machine learning model and data distribution. To support the need for conformal prediction, we reviewed EO datasets and found that only 22.5% of the datasets incorporated a degree of uncertainty information, with unreliable methods prevalent. Current open implementations require moving large amounts of EO data to the algorithms. We introduced Google Earth Engine native modules that bring conformal prediction to the data and compute, facilitating the integration of uncertainty quantification into existing traditional and deep learning modelling workflows. To demonstrate the versatility and scalability of these tools we apply them to valued EO applications spanning local to global extents, regression, and classification tasks. Subsequently, we discuss the opportunities arising from the use of conformal prediction in EO. We anticipate that accessible and easy-to-use tools, such as those provided here, will drive wider adoption of rigorous uncertainty quantification in EO, thereby enhancing the reliability of downstream uses such as operational monitoring and decision-making.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1038/s41598-024-65954-w
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
