Ensemble Forecasting for the CWFC*

Here are a few ensemble-based probabilistic forecast guidance products to glance at for the forecaster in a hurry.  Explore the links at the bottom, if you have the time, as the basic links below only scratch the surface of what is available.


Current Station: KBOS: Boston, MA (off-season default)


Precipitation Forecasting:  At least look at SREF (Short-Range Ensemble Forecast)

SREF Guidance:  The SREF is run at 09Z out to +87 hours at roughly 40 km resolution.  It is a multimodel ensemble, with 10 ETA members, 6 WRF members, and 5 RSM (regional GFS, essentially) members.  Initial conditions, lateral boundary conditions, and model physics are all perturbed.  For comparison, the operational deterministic NAM is run at 12 km resolution.

    Probability of accumulated precipitation above given threshold during the 24 hour forecast period
    > 0.01 inches        > 0.1 inches        > 0.25 inches       > 0.5 inches        > 1.00 inches        > 2.00 inches

    Stamp maps of accumulated precipitation during the 24 hour forecast period for the individual members

Canadian Ensemble Guidance:  The Canadian ensemble is run at 00z, with pertubations to both the model physics and the initial conditions.

    Calibrated probability of precipitation:  metric conversions: 2 mm ~ 0.08 inches; 5 mm ~ 0.2 inches; 10 mm ~ 0.4 inches, 25 mm ~ 1 inch
    Stamp maps:   See bottom for individual member accumulated precipitation.  Top panel shows the ensemble mean and member centers

GFS Ensemble Guidance:  The GFS ensemble is run at 12z daily, with 14 perturbed members and a control run at T126 resolution (the operational deterministic GFS is run at T382, for comparision).   Only the inital conditions are perturbed for the GFS ensemble.  Click here to get to the display applet.   Pick the 12z cycle and one of the USPOP or EASTPOP variables.  The numbers after the USPOP or EASTPOP are the duration of the inverval and the threshold.  For example, USPOP240.10 shows the probability that there will be at least 0.10 inches of precipitation during a given 24 hour window.


Temperature Forecasting:  GFS Ensemble MOS

GFS Ensemble MOS:  For some inexplicable reason the MOS calculations are only run for the 00z GFS ensemble, so for forecast contest purposes be wary of the actual numbers it puts out, as the latest deterministic GFS MOS is 18 hours younger.  Information about the forecast probability distributions from the 00z GFS ensemble is, nonetheless, still very useful.  Click here for the current CWFC station.

** Forecast probability distributions based on GFS ensemble MOS **   Click here for an explanation/interpretation (updated 8/24/2006)
Tonight's low at CWFC station
Tomorrow's high at CWFC station

** New: 4/27/06 **   Click here to see a verification study of the 2005-2006 season's GFS ensemble MOS guidance, with a comparison to conventional MOS
Verification study for the 2004-2005 season.

SREF Meteograms:  Experimental interface from SPC   To produce a 2m temperature meteogram, hit 'submit' in the uppermost frame, and 'submit' in the second frame down.  Then, choose a station from the list in the lower-left frame, above the map (or choose a location by clicking it on the map) and hit submit again. An 2m-temperature ensemble meteogram will appear in the lower-right frame. A 10m wind speed ensemble meteogram can also be created, by choosing that variable in the uppermost frame.


If that isn't enough for you ... links to the 'portal' sites for ensemble output

SREF:   NCEP    PSU    SPC
GFS:   NCEP Spaghetti    NCEP PQPF    PSU
Canadian
Pacific Northwest MM5
HPC Ensemble Training Page


Jon Moskaitis- Grad Student
Program in Atmospheres, Oceans and Climate
Massachusetts Institute of Technology
jonmosk@mit.edu
Other pages:   Research page         Boston GFS Ensemble MOS: High Temp  Low Temp         Boston climate
*this site is for entertainment purposes only.  Use this data at your own risk.  Official forecasts are at nws.noaa.gov :-)