Leveraging Deep Spatiotemporal Sequence Prediction Network with Self-Attention for Ground-Based Cloud Dynamics Forecasting

• Published in: Remote sensing (Basel, Switzerland) Vol. 17; no. 1; p. 18
• Main Authors: Li, Sheng, Wang, Min, Shi, Minghang, Wang, Jiafeng, Cao, Ran
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.01.2025
• Subjects: Accuracy; Algorithms; Attention; Clouds; Decoupling; Effectiveness; Embedding; Feature extraction; Forecasting; ground-based cloud prediction; Highway construction; Measurement; Meteorological data; Methods; Modelling; Neural networks; Precipitation; Prediction models; Predictions; Radiation; recurrent neural network; Remote sensing; self-attention mechanism; Spatiotemporal data; spatiotemporal prediction network; Velocity; Weather forecasting
• ISSN: 2072-4292; 2072-4292
• DOI: 10.3390/rs17010018

Ground-based cloud image features high-spatiotemporal resolution, presenting detailed local cloud structures and valuable weather information, which are crucial for meteorological forecasting. However, the inherent fuzziness and dynamism of ground-based clouds have hindered the development of effective prediction algorithms, resulting in low accuracy. This paper presents CloudPredRNN++, a novel method for predicting ground-based cloud dynamics, leveraging a deep spatiotemporal sequence prediction network enhanced with a self-attention mechanism. Initially, a Cascaded Causal LSTM (CCLSTM) with a dual-memory group decoupling structure is designed to enhance the representation of short-term cloud changes. Next, self-attention memory units are incorporated to capture the long-term dependencies and emphasize the non-stationary characteristics of cloud movements. These components are integrated into cloud dynamic feature mining units, which concurrently extract spatiotemporal features to strengthen unified spatiotemporal modeling. Finally, by embedding gradient highway units and adding skip connection, CloudPredRNN++ is constructed into a hierarchical recursive structure, mitigating the gradient vanishing and enhancing the uniform modeling of temporal–spatial features. Experiments on the sequence ground-based cloud dataset demonstrate that CloudPredRNN++ can predict the future cloud state more accurately and quickly. Compared with other spatiotemporal sequence prediction models, CloudPredRNN++ shows significant improvements in evaluation metrics, improving the accuracy of cloud dynamics forecasting and alleviating long-term dependency decay, thus confirming the effectiveness in ground-based cloud prediction tasks.