Temperature forecasting by deep learning methods

Abstract

Abstract. Numerical weather prediction (NWP) models solve a system of partial differential equations based on physical laws to forecast the future state of the atmosphere. These models are deployed operationally, but they are computationally very expensive. Recently, the potential of deep neural networks to generate bespoken weather forecasts has been explored in a couple of scientific studies inspired by the success of video frame prediction models in computer vision. In this study, a simple recurrent neural network with convolutional filters, called ConvLSTM, and an advanced generative network, the Stochastic Adversarial Video Prediction (SAVP) model, are applied to create hourly forecasts of the 2 m temperature for the next 12 hours over Europe. We make use of 13 years of data from the ERA5 reanalysis, of which 11 years are utilized for training and one year each is used for validating and testing. We choose the 2 m temperature, total cloud cover and the 850 hPa temperature as predictors and show that both models attain predictive skill by outperforming persistence forecasts. SAVP is superior to ConvLSTM in terms of several evaluation metrics, confirming previous results from computer vision that larger, more complex networks are better suited to learn complex features and to generate better predictions. The 12-hour forecasts of SAVP attain a mean squared error (MSE) of about 2.3 K2, an anomaly correlation coefficient (ACC) larger than 0.85, a Structural Similarity Index (SSIM) of around 0.72, and a gradient ratio (rG) of about 0.82. The ConvLSTM yields a higher MSE (3.6 K2), a smaller ACC (0.80), and SSIM (0.65), but a slightly larger rG (0.84). The superior performance of SAVP in terms of MSE, ACC, and SSIM can be largely attributed to the generator. A sensitivity study shows that a larger weight of the GAN component in the SAVP loss leads to even better preservation of spatial variability at the cost of a somewhat increased MSE (2.5 K2). Including the 850 hPa temperature as an additional predictor enhances the forecast quality and the model also benefits from a larger spatial domain. By contrast, adding the total cloud cover as predictor or reducing the amount of training data to eight years has only small effects. Although the temperature forecasts obtained in this way are still less powerful than contemporary NWP models, this study demonstrates that sophisticated deep neural networks may achieve considerable forecast quality beyond the nowcasting range in a purely data-driven way.

Deep-learning-based post-processing for probabilistic precipitation forecasting

“…This indicates that the CNN model is able to capture the statistical dependence between the raw ensembles and the observations in our study using ten-year training samples. Similar results can be seen in the recent work of Gong et al (2022).…”

A Generative Deep Learning Approach to Stochastic Downscaling of Precipitation Forecasts

“…Gong et al. ( 2022 ) forecasts the evolution of 2 m temperature over 12 hr, trained on ERA5 data, using an existing adversarial video‐prediction architecture.…”

“…Ravuri et al (2021) successfully tackled the precipitation nowcasting problem, producing high-resolution 90-min forecasts over the UK. Gong et al (2022) forecasts the evolution of 2 m temperature over 12 hr, trained on ERA5 data, using an existing adversarial video-prediction architecture.…”

Deep Tower Networks for Efficient Temperature Forecasting from Multiple Data Sources

“…In this continuation of the work, we also tested the use of input data related to topography and time of year. It is important to point out that we are not aiming to compare specific variations of basic architectures like [ 16 , 24 , 25 , 26 ] on different datasets but rather want to compare to the basic architectures themselves. In addition, these methods usually have a lot of parameters and will result in over-fitting on our dataset with a short time horizon.…”