Abstract:This study explores the application of the fully convolutional network (FCN) algorithm to the field of meteorology, specifically for the short-term nowcasting of severe convective weather events such as hail, convective wind gust (CG), thunderstorms, and short-term heavy rain (STHR) in Gansu. The training data come from the European Center for Medium-Range Weather Forecasts (ECMWF) and real-time ground observations. The performance of the proposed FCN model, based on 2017 to 2021 training datasets, demonstrated a high prediction accuracy, with an overall error rate of 16.6%. Furthermore, the model exhibited an error rate of 18.6% across both severe and non-severe weather conditions when tested against the 2022 dataset. Operational deployment in 2023 yielded an average critical success index (CSI) of 24.3%, a probability of detection (POD) of 62.6%, and a false alarm ratio (FAR) of 71.2% for these convective events. It is noteworthy that the predicting performance for STHR was particularly effective with the highest POD and CSI, as well as the lowest FAR. CG and hail predictions had comparable CSI and FAR scores, although the POD for CG surpassed that for hail. The FCN model's optimal performances in terms of hail prediction occurred at the 4th, 8th, and 10th forecast hours, while for CG, the 6th hour was most accurate, and for STHR, the 2nd and 4th hours were most effective. These findings underscore the FCN model's ideal suitability for short-term forecasting of severe convective weather, presenting extensive prospects for the automation of meteorological operations in the future.
What problem does this paper attempt to address?
The paper mainly discusses how to apply the Fully Convolutional Network (FCN) algorithm to improve the short-term near forecasting of severe convective weather in different types in Gansu Province, China (such as hail, strong gusts, thunderstorms, and short-term heavy rainfall). In the study, the training data is from the European Centre for Medium-Range Weather Forecasts (ECMWF) data and real-time surface observations. The FCN model performs well on the training data from 2017 to 2021, with a prediction error rate of 16.6%, and a prediction error rate of 18.6% on the severe and non-severe weather test set in 2022. In actual operation in 2023, the model has an average Critical Success Index (CSI) of 24.3%, Probability of Detection (POD) of 62.6%, and False Alarm Rate (FAR) of 71.2% for these convective events. In particular, the model performs best in predicting short-term heavy rainfall.
The paper points out that accurate forecasting of severe convective weather is challenging due to its rapid changes and local characteristics. The study improves the short-term forecast accuracy of classifying severe convective weather by using the FCN model to process multi-source observation data and high-resolution numerical model prediction data. The FCN model achieves pixel-level semantic segmentation through techniques such as convolution, upsampling, and skip connections, which is suitable for the meteorological conditions in Gansu Province.
In the study, the authors collected 24 meteorological features, including vertical wind speed, potential vorticity, temperature, topographic height, horizontal wind speed, specific humidity, etc., and used ECMWF's numerical forecast products for temporal and spatial downscaling. The input data of the model is standardized or normalized to meet the requirements of the model.
In addition, the paper also introduces modeling methods for dependent variables and samples, as well as model validation metrics such as Probability of Detection (POD), False Alarm Rate (FAR), and Critical Success Index (CSI), to evaluate the model's identification and prediction of short-term heavy rainfall and severe convective weather. Through these methods, the researchers evaluated the performance of the model on independent testing and validation sets in 2022 and 2023.