Cloud cover nowcasting is an attempt to forecast the position of clouds on a short time scale. Contrary to physical-modeling-based methods employed in traditional meteorology, it extrapolates pixels from wind vectors. A recent study applies deep convolutional networks to forecast the position of clouds on a 1h30’ time scale.
Models based on different neural network architectures were able to learn to predict the movement of cloud covers. The results show that the suggested technique outperforms both traditional image extrapolation and physical modeling methods and requires less computational time. The main limitation of the neural network architectures is that it cannot predict the emergence of new clouds. The cloud cover nowcasting can be used to forecast the electricity production of solar panels or to optimize satellite image shots.
Nowcasting is a field of meteorology which aims at forecasting weather on a short term of up to a few hours. In the meteorology landscape, this field is rather specific as it requires particular techniques, such as data extrapolation, where conventional meteorology is generally based on physical modeling. In this paper, we focus on cloud cover nowcasting, which has various application areas such as satellite shots optimisation and photovoltaic energy production forecast.
Following recent deep learning successes on multiple imagery tasks, we applied deep convolutionnal neural networks on Meteosat satellite images for cloud cover nowcasting. We present the results of several architectures specialized in image segmentation and time series prediction. We selected the best models according to machine learning metrics as well as meteorological metrics. All selected architectures showed significant improvements over persistence and the well-known U-Net surpasses AROME physical model.