Abstract
AbstractA one‐dimensional Kalman filter is proposed for the correction of maximum and minimum near surface (2 m) temperature forecasts obtained by a Numerical Weather Prediction model. In our study we used only one parameter (observed temperature), employing in this way scalar system and observation equations, and a limited time interval (7 days). As a result, our algorithm can be easily run on any PC with only minor technical support. The corresponding results are rather impressive, since the systematic error of our time series almost disappears, and they show the merit of post‐processing even when only a simple method is applied. Copyright © 2002 Royal Meteorological Society.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
