Search

Abstract 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.

Accurately capturing the impact of urban trees on temperature can help optimize urban heat mitigation strategies. Recently, there has been widespread use of remotely sensed land surface temperature ( T _s ) to quantify the cooling efficiency (CE) of urban trees. However, remotely sensed T _s reflects emitted radiation from the surface of an object seen from the point of view of the thermal sensor, which is not a good proxy for the air temperature ( T _a ) perceived by humans. The extent to which the CEs derived from T _s reflect the true experiences of urban residents is debatable. Therefore, this study systematically compared the T _s -based CE (CE _T _s ) with the T _a -based CE (CE _T _a ) in 392 European urban clusters. CE _T _s and CE _T _a were defined as the reductions in T _s and T _a , respectively, for every 1% increase in fractional tree cover (FTC). The results show that the increase in FTC has a substantial impact on reducing T _s and T _a in most cities during daytime. However, at night, the response of T _s and T _a to increased FTC appears to be much weaker and ambiguous. On average, for European cities, daytime CE _T _s reaches 0.075 °C % ^−1 , which is significantly higher (by an order of magnitude) than the corresponding CE _T _a of 0.006 °C % ^−1 . In contrast, the average nighttime CE _T _s and CE _T _a for European cities are similar, both approximating zero. Overall, urban trees can lower daytime temperatures, but the magnitude of their cooling effect is notably amplified when using remotely sensed T _s estimates compared to in situ T _a measurements, which is important to consider for accurately constraining public health benefits. Our findings provide critical insights into the realistic efficiencies of alleviating urban heat through tree planting.

Wetlands are important habitats for biodiversity and provide ecosystem services such as climate mitigation and carbon storage. The current wetland mapping techniques in Norway are tedious and costly, and remote sensing provides an opportunity for large-scale mapping and ecosystem accounting. We aimed to implement a deep learning approach to mapping wetlands with Sentinel-2 and LiDAR data over southern Norway. Our U-Net model, implemented through Google Earth Engine and TensorFlow, produced a wetland map with a balanced accuracy rate of 90.9% when validated against an independent ground-truth sample. This represents an improvement upon manually digitized land cover maps in Norway, which achieved accuracy rates of 46.8% (1:50,000 map) and 42.4% (1:5000 map). Using our map, we estimated a total wetland coverage area of 12.7% in southern Norway, which is double the previous benchmark estimates (5.6%). We followed an iterative model training and evaluation approach, which revealed that increasing the quantity and coverage of labeled wetlands greatly increases the model performance. We highlight the potential of satellite-based wetland maps for the ecosystem accounting of changes in wetland extents over time—something that is not feasible with traditional mapping methods.