Precipitation Nowcasting Based on Deep Learning over Guizhou, China
نویسندگان
چکیده
Accurate precipitation nowcasting (lead time: 0–2 h), which requires high spatiotemporal resolution data, is of great relevance in many weather-dependent social and operational activities. In this study, we are aiming to construct highly accurate deep learning (DL) models directly obtain at 6-min intervals for the lead time h. The Convolutional Long Short-Term Memory (ConvLSTM) Predictive Recurrent Neural Network (PredRNN) were used as comparative DL models, Lucas–Kanade (LK) Optical Flow method was selected a traditional extrapolation baseline. trained with high-quality datasets (resolution: 1 min) created from observations recorded by automatic weather stations Guizhou Province (China). A comprehensive evaluation performed, included consideration root mean square error, equitable threat score (ETS), probability detection (POD). indicated that reduction number missing values data normalization boosted training efficiency improved forecasting skill models. Increasing series length set samples both POD ETS enhanced stability time. Training Hea-P dataset further sharply increased thresholds 2.5, 8, 15 mm, especially 1-h PredRNN model (time length: 8 years) outperformed LK all (0.1, 1, mm) obtained best performance considered study terms ETS. Moreover, Method Object-Based Diagnostic Evaluation on rainstorm case revealed model, well observation could capture complex nonlinear characteristics more accurately than achievable using establish reasonable mapping network during drastic changes precipitation. Thus, its results closely matched observations, exceeding mm substantially.
منابع مشابه
Deep Learning for Precipitation Nowcasting: A Benchmark and A New Model
With the goal of making high-resolution forecasts of regional rainfall, precipitation nowcasting has become an important and fundamental technology underlying various public services ranging from rainstorm warnings to flight safety. Recently, the Convolutional LSTM (ConvLSTM) model has been shown to outperform traditional optical flow based methods for precipitation nowcasting, suggesting that ...
متن کاملConvolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting
The goal of precipitation nowcasting is to predict the future rainfall intensity in a local region over a relatively short period of time. Very few previous studies have examined this crucial and challenging weather forecasting problem from the machine learning perspective. In this paper, we formulate precipitation nowcasting as a spatiotemporal sequence forecasting problem in which both the in...
متن کاملQuantitative Precipitation Nowcasting: A Lagrangian Pixel-Based Approach
Short-term high-resolution precipitation forecasting has important implications for navigation, flood forecasting, and other hydrological and meteorological concerns. This article introduces a pixel-based algorithm for Short-term Quantitative Precipitation Forecasting (SQPF) using radar-based rainfall data. The proposed algorithm called PixelBased Nowcasting (PBN) tracks severe storms with a hi...
متن کاملReconstructing Satellite-Based Monthly Precipitation over Northeast China Using Machine Learning Algorithms
Attaining accurate precipitation data is critical to understanding land surface processes and global climate change. The development of satellite sensors and remote sensing technology has resulted in multi-source precipitation datasets that provide reliable estimates of precipitation over un-gauged areas. However, gaps exist over high latitude areas due to the limited spatial extent of several ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Atmosphere
سال: 2023
ISSN: ['2073-4433']
DOI: https://doi.org/10.3390/atmos14050807