Transcript Furthermore…

Furthermore…
• References
Katz, R.W. and A.H. Murphy (eds), 1997: Economic Value of Weather
and Climate Forecasts. Cambridge University Press, Cambridge.
Jolliffe, I.T., and D.B. Stephenson, 2003: Forecast Verification. A
Practitioner's Guide in Atmospheric Science. Wiley and Sons Ltd, 240
pp. Click here to see the Table of Contents.
von Storch, H. and F.W. Zwiers, 1999: Statistical Analysis in Climate
Research. Cambridge University Press, Cambridge.
Wilks, D.S., 2006: Statistical Methods in the Atmospheric Sciences. An
Introduction. Academic Press, San Diego, 467 pp.
Challenges in Spatial Forecasts
O
F
O
O
F
F
O
O F
F
• Forecasts 1-4 have POD=0; FAR=1;
CSI=0
• Fifth forecast has POD>0, FAR<1, CSI>1
Verifying Spatial Forecasts
Verifying Extreme
Forecasts
• What are extremes?
– Events large in magnitude
– Events rare in occurrence
• Traditional skill scores become unstable
as the probability of the event becomes
increasingly small.
• Extreme value statistics. Different
distributions are more appropriate for
use with extreme events.
Confidence Intervals
• Skill scores are statistics and as such it is reasonable
to ask for confidence intervals.
• Some techniques for estimating confidence intervals
– Parametric assumptions and inference
– Bootstrapping (re-sampling original data)
– Use of holdout data in creating model
• Difficulties include
– Dependent observations
– Huge numbers of observation – everything is
significant
– Small number of data