no code implementations • 25 Jan 2022 • Charlie Kirkwood, Theo Economou, Henry Odbert, Nicolas Pugeault
However, as the number of available observation sites increases, so too does the opportunity for data quality issues to emerge, particularly given that many of these sensors do not have the benefit of official maintenance teams.
1 code implementation • 6 May 2020 • Charlie Kirkwood, Theo Economou, Henry Odbert, Nicolas Pugeault
In this paper, we use a road surface temperature example to demonstrate a three-stage framework that uses machine learning to bridge the gap between sets of separate forecasts from NWP models and the 'ideal' forecast for decision support: probabilities of future weather outcomes.