A Machine Learning-Based Bias Correction Method for Global Forecast System Products

Accurate numerical weather forecasting is essential for forecasters to make predictions. However, operational forecast models often exhibit systematic biases. In this study, we present a machine learning-based approach called BC-Unet, a convolutional neural network (CNN) based on the renowned U-Net architecture, to correct biases in National Centers for Environmental Prediction (NCEP) operational Global Forecast System (GFS) v16 products over the contiguous United States (CONUS).