Can deep learning beat numerical weather prediction?

The recent hype about artificial intelligence has sparked renewed interest in applying the successful deep learning (DL) methods for image recognition, speech recognition, robotics, strategic games and other application areas to the field of meteorology. There is some evidence that better weather forecasts can be produced by introducing big data mining and neural networks into the weather prediction workflow. Here, we discuss the question of whether it is possible to completely replace the current numerical weather models and data assimilation systems with DL approaches. This discussion entails a review of state-of-the-art machine learning concepts and their applicability to weather data with its pertinent statistical properties. We think that it is not inconceivable that numerical weather models may one day become obsolete, but a number of fundamental breakthroughs are needed before this goal comes into reach. This article is part of the theme issue 'Machine learning for weather and climate modelling'.

No effect of prenatal vitamin D deficiency on autism-relevant behaviours in multiple inbred strains of mice.

Autism Spectrum Disorders (ASD) is a group of neurodevelopmental disorders commonly characterised by verbal and non-verbal communication deficits, impaired social interaction and repetitive, stereotypic behaviours. The aetiology of ASD is most likely a combination of genetic and environmental factors. Epidemiological evidence suggests that prenatal vitamin D deficiency is associated with an increased incidence of ASD. The overall aim of this study was to investigate prenatal vitamin D deficiency on ASD-related behavioural phenotypes in multiple inbred strains of mice. We included two commonly used inbred mouse strains (C57BL/6J and BALB/c) as well as inbred BTBR mice, which show ASD-related behaviours, such as excessive self-grooming, hyperlocomotion, social interaction deficits and altered communication. We also studied the effect of prenatal vitamin D deficiency in a fourth strain; an F1 cross of C57BL/6J x BTBR mice, which have a partial BTBR phenotype. To implement prenatal vitamin D deficiency, female mice were placed on vitamin D deplete diets for ten weeks, including mating and gestation, until littering, when all dams were switched to the control diet. Behavioural symptoms related to ASD were measured, including isolation-induced ultrasonic vocalisations to measure communication, the three-chambered social interaction task to observe social interaction, the open field test to examine hyperlocomotion, assessment of grooming and rearing behaviour and finally the active place avoidance task to observe spatial learning and memory in response to a mild foot shock. Prenatal vitamin D deficiency had a negative impact on preference for social novelty in C57BL/6J mice, despite similar vocalisation phenotypes, and prenatal vitamin D-deficient F1 mice were found to be hypolocomotive in the open field test yet performed better on the active place avoidance task. Despite clear differences between strains, there were no other consistent significant main effects of maternal diet on the behaviour of the offspring. Vitamin D deficiency has been implicated as a risk factor for ASD and these data show that there is greater variation between different inbred strains in ASD-related behaviour, suggesting that prenatal vitamin D deficiency is not sufficient to recapitulate an ASD phenotype in multiple inbred strains of mice.