Spatial Forecast Verification Methods Inter-Comparison Project
Information
We currently have three sets of test cases picked out for this project. Thanks to Mike Baldwin, Beth Ebert, Barbara Casati and others for helping to obtain them, and to David Ahijevych for creating the fake test cases. All of the cases are included for download by clicking on the link to this file: Cases20081023.tar.gz
Once extracted you will have the following folders inside the folder called Cases: 'sp2005,' 'nimrod,' 'MeteoSwiss,' and 'Fake.' Inside 'Fake' are further folders for the geometric and perturbed cases.
Initial Test Cases from the Storm Prediction Center/National Severe Storms Laboratory (SPC/NSSL) Spring 2005 Program (sp2005)
We have cases from the WRF Spring 2005 program (see Kain et al., 2008 Wea. Forecasting, 23:931--952) including stage II analyses and three versions of WRF referred to as 'wrf2,' 'wrf4ncar' and 'wrf4ncep.' The fields are hourly accumulated precipitation in hundredths-of-an-inch. Please convert to mm by multiplying by 0.254 so that we all use the same units. All of these models have been interpolated onto the same 501 x 601 grid. Specifically, the NCEP g240 grid, which is a "4-km grid," but note that the grid is irregular (i.e., not always 4-km at each grid square). The longitude and latitude information is provided in the sp2005 folder as 'g240LatNorth.txt' and 'g240LonEast.txt.'
In the 'figures' folder you will find postscript and pdf graphs in the 'ps' and 'pdf' folders, respectively, of the first 9 cases chosen from this Spring program for this inter-comparison. Note that there are a couple extra cases in the 'st2,' 'wrf2' and 'wrf4ncar' folders. We did not have these dates for the NCEP model, so we will exclude them, but they are still provided as they may make for nice cases. All files are in ASCII format. See the README file in this folder for more information.
The cases we will start with are for verification times (00Z issue times): 4/26, 5/13, 5/14, 5/18, 5/19, 5/25, 6/1, 6/3 and 6/4.
The entire 32 Spring 2005 ARW and NMM cases studied in Davis et al (2009) and Gilleland et al (2010) are available here (spc2005agg24hLead00UTC.tar.gz).
Nimrod Test Cases
Please acknowledge the UK Met Office (see the README file in the 'Nimrod' folder. These are the same cases used in the Casati et al (2004) paper (Meteorol. Appl. 11:141--154), and are described therein. We do not presently have lon/lat info for these cases (sorry).
MeteoSwiss Test Cases
Daniel Leuenberger at MeteoSwiss has offered two cases we could consider using. He asks that we acknowledge their source, and use them only for research purposes (see the 'README' file in the 'MeteoSwiss' folder). The cases give high-resolution aLMo radar data over Switzerland, along with hourly forecasts for a 24-hour period. One case also has results on a 7-km grid corresponding to a lower resolution version of the model. The files are in netCDF format, and IDL code to read them is included. One case has a postscript file that shows the hourly fields and basic verification statistics. The other case has a postscript file for the 24-hr totals, plus gif files for 6h totals.
Fake Test Cases
Fake test cases are based on the same grid as the sp2005 cases. There are both idealized geometric cases as well as perturbed real cases. See the README files for each case for more information.
The National Center for Atmospheric Research is sponsored by the National Science Foundation. Any opinions, findings, and conclusions or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
