Operational Forecasts of Astronomical Observing Conditions
Goals of the talk
Two types of product
The path followed by the data
The CIRA products
The satellite images
The ECMWF products
The two types of graphs
Biases in the ECMWF products
The Kalman filter
The Kalman applet
A need for verification
The verification graphs
Kalman filter: some good results
Kalman filter: results for temperature
Some statistics
The use of those forecasts
Conclusion
Future developments