Hybrid LSTM-ViT model using mesonet surface data and profiler vertical profiles improves HRRR forecast error prediction for precipitation, wind speed, and temperature, with roughly twofold skill gain for precipitation over baseline LSTM.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A Hybrid LSTM--Vision Transformer Architecture for Predicting HRRR Forecast Errors
Hybrid LSTM-ViT model using mesonet surface data and profiler vertical profiles improves HRRR forecast error prediction for precipitation, wind speed, and temperature, with roughly twofold skill gain for precipitation over baseline LSTM.