paper
2022
Extremes & hazards
Observations & remote sensing New model or method
RNN / LSTM / ConvLSTM6 h–3 daysChina Meteorological Administration (CMA) Journal
2022-05-19
022
May 2022
Pingping Dong, Jie Lian, Hui Yu et al.
Weather and Forecasting
China Meteorological Administration (CMA)
A ConvLSTM encoder-decoder with spatial attention forecasts 24 h western North Pacific tropical cyclone tracks from satellite images, about 30 km better than ECMWF-HRES and GFS.
Tropical Cyclone Track Prediction with an Encoding-to-Forecasting Deep Learning Model
In one sentence. A ConvLSTM encoder-decoder with spatial attention forecasts 24 h western North Pacific tropical cyclone tracks from satellite images, about 30 km better than ECMWF-HRES and GFS.
Labels
- Topic
- Extremes & hazards, Observations & remote sensing
- Type
- New model or method
- Method
- RNN / LSTM / ConvLSTM
- Time scale
- 6 h–3 days
- Domain
- Regional
- Data
- Satellite, NWP analysis or forecasts (IFS, GFS, ...)
- Status
- Research
- Identifiers
- DOI:10.1175/waf-d-21-0116.1 OpenAlex:W4280638202
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.