This data set gives 48-hour forecasts of 2-m surface temperature and the
associated observations for the US Pacific Northwest from January 1, 2004
to February 28, 2004. The ensemble forecasts come from an eight-member
version of the University of Washington Mesoscale Ensemble
(Grimit and Mass 2002; Eckel and Mass 2005).
Temperatures are measured in kelvins.
Note that forecasts are not available for some of the interim dates.
A data frame with 36826 rows and 15 columns:
forecasts from the 8 members of the ensemble (numeric).
observation the observed surface temperature (numeric).
date the date of each forecast/observation set,
in the format YYYYMMDDHH (categorical).
latitude the latitude of each forecast (numeric).
longitude the longitude of each forecast (numeric).
station weather station identifier (categorical).
type weather station type (categorical).
F. A. Eckel and C. F. Mass, Effective mesoscale, short-range ensemble forecasting, Weather and Forecasting 20:328–350, 2005.
E. P. Grimit and C. F. Mass, Initial results of a mesoscale short-range ensemble forecasting system over the Pacific Northwest, Weather and Forecasting 17:192–205, 2002.
V. J. Berrocal, A. E. Raftery and T. Gneiting, Combining spatial and ensemble information in probabilistic weather forecasts, Monthly Weather Review 133:1386–1402, 2007.
V. J. Berrocal, A. E. Raftery, T. Gneiting and R. C. Steed, Probabilistic Weather Forecasting for Winter Road Maintenance, Journal of the American Statistical Association, 2010 (to appear).
C. Fraley, A. E. Raftery, T. Gneiting and J. M. Sloughter,
R Package for Probabilistic Forecasting
using Ensembles and Bayesian Model Averaging,
Technical Report No. 516R, Department of Statistics, University of
Washington, 2007 (revised 2010).
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
## Not run: # R check data(srft) labels <- c("CMCG","ETA","GASP","GFS","JMA","NGPS","TCWB","UKMO") srftData <- ensembleData( forecasts = srft[ ,labels], dates = srft$date, observations = srft$obs, latitude = srft$lat, longitude = srft$lon, forecastHour = 48, initializationTime = "00") srftFit <- ensembleBMAnormal( srftData, date = "2004013100", trainingDays = 25) ## End(Not run)