Precipitation Nowcasting with Orographic Enhanced Stacked Generalization: Improving Deep Learning Predictions on Extreme Events

Abstract

One of the most crucial applications of radar-based precipitation nowcasting systems is the short-term forecast of extreme rainfall events such as flash floods and severe thunderstorms. While deep learning nowcasting models have recently shown to provide better overall skill than traditional echo extrapolation models, they suffer from conditional bias, sometimes reporting lower skill on extreme rain rates compared to Lagrangian persistence, due to excessive prediction smoothing. This work presents a novel method to improve deep learning prediction skills in particular for extreme rainfall regimes. The solution is based on model stacking, where a convolutional neural network is trained to combine an ensemble of deep learning models with orographic features, doubling the prediction skills with respect to the ensemble members and their average on extreme rain rates, and outperforming them on all rain regimes. The proposed architecture was applied on the recently released TAASRAD19 radar dataset: the initial ensemble was built by training four models with the same TrajGRU architecture over different rainfall thresholds on the first six years of the dataset, while the following three years of data were used for the stacked model. The stacked model can reach the same skill of Lagrangian persistence on extreme rain rates while retaining superior performance on lower rain regimes.Plain Positions Indicators (PPI) or Constant Altitude Plain Position Indicator (CAPPI), or the Maximum vertical reflectivity (CMAX or MAX(Z)). Sequences of reflectivity maps are used as input for prediction models. More formally, given a reflectivity field at time T 0 , radar-based nowcasting methods aim to extrapolate m future time steps T 1 , T 2 , ..., T m in the sequence, using as input the current and n previous observations T −n , ..., T −1 , T 0 .Traditional nowcasting models are manly based on Lagrangian echo extrapolation [7,8], with recent modification that try to infer precipitation growth and decay [9,10] or integrate with Numerical Weather Predictions to extend the time horizon of the prediction [11,12]. In the last few years, Deep Learning (DL) models based on combination of Recurrent Neural Networks (RNN) and Convolutional Neural Networks (CNN) have shown substantial improvement over nowcasting methods based on Lagrangian extrapolations for quantitative precipitation forecasting (QPF) [13]. Shi et al. [14] introduced the application of the Convolutional Long Short-Term Memory (Conv-LSTM) network architecture with the specific goal of improving precipitation nowcasting over extrapolation models, where LSTM is modified using a convolution operator in the state-to-state and input-to-state transitions. Subsequent work introduced dynamic recurrent connections [15] (TrajGRU) that allowed the improvement of prediction skills, spatial resolution, and temporal length of the forecast, with comparable number of parameters and memory requirements. Subsequent works introduced more complex memory blocks and architectures [16] and increased num...

Enhancing the Encoding-Forecasting Model for Precipitation Nowcasting by Putting High Emphasis on the Latest Data of the Time Step

“…Next, the false alarm rate (FAR), possibility of detection (POD), critical success index (CSI), and the Heidke skill score (HSS) [33] were obtained using the following formulas based on the confusion matrix presented in Table 3. A method of measuring prediction performance according to the level of rainfall rate based on the confusion matrix, which is used in a classification problem, is a common method in the meteorological field [16,18,20,21,24,25,27]. After predicting rainfall through the model, we generate five confusion matrices based on the five thresholds (i.e., the rainfall rate of 0.5, 2.0, 5.0, 10.0, and 30.0 mm/h).…”

“…In addition, they decreased the number of channels in high abstract recurrent layers, unlike the usual methods. Franch et al [27] improved the prediction performance of extreme events by creating an ensemble model based on the existing model. Tran and Song [16] improved the model performance by introducing a data augmentation technique to RNN as in CNN.…”

Monthly Rainfall Anomalies Forecasting for Southwestern Colombia Using Artificial Neural Networks Approaches

“…There are different architectures of ANN; however, the most common model is a Multi-Layer Perceptron (MLP) neural network, which has a structure with an input layer, single or multiple hidden layers, and an output layer. The MLP has been widely used to forecast several phenomena in meteorology and hydroclimatology [3,9,17,[54][55][56][57]. The typical mathematical expression of the ANN is:…”

TAASRAD19, a high-resolution weather radar reflectivity dataset for precipitation nowcasting

“…Events have been annotated by experts with precipitation classification tags extracted from daily weather summaries. As a technical validation of TAASRAD19, the annotated data are used to develop a deep learning solution for precipitation nowcasting 22 . Finally, the structure of each image data can be explored with an interactive data visualization of an Uniform Manifold Approximation and Projection (UMAP) embedding 23 .…”

“…Analog ensemble models 26 , 38 , 39 or extrapolation methods 12 are mainly used for probabilistic forecasting; however convolutional recurrent neural networks are now the state of the art for deterministic nowcasting 31 , 40 – 43 . In 22 we used TAASRAD19 to train a deep learning model that forecasts reflectivity up to 100 min ahead (i.e. 20 frames) at full spatial spatial resolution of the radar (0.5 × 0.5 km), based on 25 min (i.e.…”