ensBMAtest: Ensemble BMA Test Data Set

In ensembleBMA: Probabilistic Forecasting using Ensembles and Bayesian Model Averaging

Ensemble BMA Test Data Set

Description

This data set gives 48-hour forecasts for 2-m temperature, precipitation accumulated over the last 24 hours, and maximum wind speed at SeaTac (KSEA) and Portland (PDX) ariports in 2007/2008 initialized at 00 hours UTC using a 12km grid. The forecasts are based on an 8 member version of the University of Washington mesoscale ensemble (Grimit and Mass 2002; Eckel and Mass 2005).

Format

A data frame with 66 rows and 34 columns:
idate the initialization date of each forecast/observation, format YYYYMMDDHH (categorical).
vdate the validation date of each forecast/observation, format YYYYMMDDHH (categorical).
latitude the latitude of each forecast/observation (numeric).
longitude the longitude of each forecast/observation (numeric).
longitude the elevation (in meters) above sea level (numeric).
station weather station identifier (categorical).
network weather network identifier (categorical). *.gfs,*.cmcg,*.eta,*.gasp,*.jma,*.ngps,*.tcwb forecasts from the 8 members of the ensemble (numeric). *.obs observed values for the weather parameters. The prefix * is one of T2 for temperature, PCP24 for precipitation, MAXWSP10 for wind speed.

Details

Temperature is given in Kelvin.
Precipitation amounts are quantized to hundredths of an inch.
Maximum wind speed is defined as the maximum of the hourly 'instantaneous' wind speeds over the previous 18 hours, where an hourly 'instantaneous' wind speed is a 2-minute average from the period of two minutes before the hour to on the hour.
The wind speed observations are measured at 10-m above the ground and discretized when recorded by rounding to the nearest meter per second.
This is a small dataset provided for the purposes of testing. Typically forecasting would be performed on much larger datasets.

References

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.

Examples

## Not run: # R check

  data(ensBMAtest)

  ensMemNames <- c("gfs","cmcg","eta","gasp","jma","ngps","tcwb","ukmo")

#----------------------------------------------------------------------------

  obs <- paste("T2","obs", sep = ".")
  ens <- paste("T2", ensMemNames, sep = ".")

  tempTestData <- ensembleData( forecasts = ensBMAtest[,ens],
                                dates = ensBMAtest[,"vdate"],
                                observations = ensBMAtest[,obs],
                                station = ensBMAtest[,"station"],
                                forecastHour = 48,
                                initializationTime = "00")

  tempTestFit <- ensembleBMAnormal( tempTestData, trainingDays = 30)

  MAE( tempFit, tempTestData)
  CRPS( tempFit, tempTestData)

#----------------------------------------------------------------------------

  obs <- paste("PCP24","obs", sep = ".")
  ens <- paste("PCP24", ensMemNames, sep = ".")

  prcpTestData <- ensembleData( forecasts = ensBMAtest[,ens],
                                dates = ensBMAtest[,"vdate"], 
                                observations = ensBMAtest[,obs], 
                                station = ensBMAtest[,"station"],
                                forecastHour = 48,
                                initializationTime = "00")

  prcpTestFit <- ensembleBMAgamma0( prcpTestData, trainingDays = 30)

  MAE( prcpTestFit, prcpTestData)
  CRPS( prcpTestFit, prcpTestData)

#----------------------------------------------------------------------------

  obs <- paste("MAXWSP10","obs", sep = ".")
  ens <- paste("MAXWSP10", ensMemNames, sep = ".")

  winsTestData <- ensembleData( forecasts = ensBMAtest[,ens],
                                dates = ensBMAtest[,"vdate"], 
                                observations = ensBMAtest[,obs],
                                station = ensBMAtest[,"station"],
                                forecastHour = 48,
                                initializationTime = "00")

   winsTestFit <- ensembleBMAgamma(winsTestData, trainingDays = 30)

   MAE( winsTestFit, winsTestData)
   CRPS( winsTestFit, winsTestData)

## End(Not run)

ensembleBMA documentation built on Sept. 2, 2022, 9:05 a.m.