DroughtFormer predicts soil moisture, vegetation health, and related variables in Africa with skill out to 90 days that matches or exceeds climatology for most targets, but shows lower accuracy for precipitation and flash drought indices.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
verdicts
UNVERDICTED 2representative citing papers
CNN post-processing applied member-wise to a 51-member 40-km NWP ensemble creates a 5-km high-resolution ensemble forecast system with improved deterministic accuracy and probabilistic reliability for surface temperatures.
citing papers explorer
-
Prediction of Drought and Flash Drought in Africa at the Seasonal-to-Subseasonal Scale using the Community Research Earth Digital Intelligence Twin Framework
DroughtFormer predicts soil moisture, vegetation health, and related variables in Africa with skill out to 90 days that matches or exceeds climatology for most targets, but shows lower accuracy for precipitation and flash drought indices.
-
CNN-based Surface Temperature Forecasts with Ensemble Numerical Weather Prediction
CNN post-processing applied member-wise to a 51-member 40-km NWP ensemble creates a 5-km high-resolution ensemble forecast system with improved deterministic accuracy and probabilistic reliability for surface temperatures.