Here, we analyse high-frequency (1 min) surface air temperature, mean sea-level pressure (MSLP), wind speed and direction and cloud-cover data acquired during the solar eclipse of 20 March 2015 from 76 UK Met Office weather stations, and compare the results with those from 30 weather stations in the Faroe Islands and 148 stations in Iceland. There was a statistically significant mean UK temperature drop of 0.83±0.63°C, which occurred over 39 min on average, and the minimum temperature lagged the peak of the eclipse by about 10 min. For a subset of 14 (16) relatively clear (cloudy) stations, the mean temperature drop was 0.91±0.78 (0.31±0.40)°C but the mean temperature drops for relatively calm and windy stations were almost identical. Mean wind speed dropped significantly by 9% on average during the first half of the eclipse. There was no discernible effect of the eclipse on the wind-direction or MSLP time series, and therefore we can discount any localized eclipse cyclone effect over Britain during this event. Similar changes in air temperature and wind speed are observed for Iceland, where conditions were generally clearer, but here too there was no evidence of an eclipse cyclone; in the Faroes, there was a much more muted meteorological signature.This article is part of the themed issue ‘Atmospheric effects of solar eclipses stimulated by the 2015 UK eclipse’.

Image translation with convolutional autoencoders has recently been used as an approach to multimodal change detection (CD) in bitemporal satellite images. A main challenge is the alignment of the code spaces by reducing the contribution of change pixels to the learning of the translation function. Many existing approaches train the networks by exploiting supervised information of the change areas, which, however, is not always available. We propose to extract relational pixel information captured by domain-specific affinity matrices at the input and use this to enforce alignment of the code spaces and reduce the impact of change pixels on the learning objective. A change prior is derived in an unsupervised fashion from pixel pair affinities that are comparable across domains. To achieve code space alignment, we enforce pixels with similar affinity relations in the input domains to be correlated also in code space. We demonstrate the utility of this procedure in combination with cycle consistency. The proposed approach is compared with the state-of-the-art machine learning and deep learning algorithms. Experiments conducted on four real and representative datasets show the effectiveness of our methodology.

No abstract

This paper proposes a new method for bitemporal change detection in heterogeneous remote sensing images. A modified canonical correlation analysis is used to align the code layers of two deep convolutional autoencoders, one for each image domain. It weights the input with a new affinity-based prior, which measures changes in pixel relations across the image domains and is used to reduce the influence of data points prone to change. By this procedure of self-supervision, we adapt the intrinsically supervised architecture to the unsupervised case, noting that the censoring of change pixels is key to efficiently learning the required data transformations. The result is an unsupervised algorithm which allows change detection in either of the image domains, or a combination of those, since efficient domain translation is obtained by coupling cross-domain encoders and decoders. We demonstrate state-of-the-art performance on real test datasets.

No abstract