Paciorek (2013) Electronic Journal of Statistics

This archive contains the data (and code) used in Paciorek (2013), in EJS. I include the PM data for the first example and R code for generating the simulations. The LA cancer analysis is based on data that is not in the public domain. April 2013.

Paciorek (2012) JRSSC (Applied Statistics)

This data archive contains the data used in Paciorek (2012), in JRSSC. I include the simulation data and the monthly data for the mid-Atlantic analyses for 2004. The eastern U.S. analyses are based on proprietary CMAQ model output provided by AER with funding from EPRI and I have not explored whether they are willing to release the data. November 2011.

Paciorek et al. (2009) Annals of Applied Statistics

This data archive contains the data used in Paciorek et al. (2009), in the Annals of Applied Statistics. The data are monthly average PM10 (for 1988-2002) and PM2.5 (for 1999-2002) concentrations in the northeastern U.S. from AQS, IMPROVE, CASTNET and a few other sources, as collated by Jeff Yanosky. The data can fit with a spatio-temporal model using this code, used for our paper. August 2008.

Paciorek (2007) Computational Statistics and Data Analysis

This data archive contains the simulation data used in Paciorek (2007), in Computational Statistics and Data Analysis. Note that the real data on Taiwan cancer cases is not available here for confidentiality reasons, but may be available upon request. The archive contains files for 51 simulated data samples from each of 4 datasets: flat, aniso, iso, and cohort, described in the paper.

Paciorek et al. (2002), Journal of Climate

This data archive currently contains the six indices of storm activity calculated in Paciorek et al. (2002), Journal of Climate, from the NCEP-NCAR Reanalysis. We thank the Reanalysis project and the NOAA Climate Diagnostics Center for the data. We have calculated values at each of the NCEP-NCAR grid points in the Northern Hemisphere, 0-90N, 0-357.5E, at intervals of 2.5 degrees. Values are available for 1949-1999. Data are available at the winter level (one value per winter) as flat ascii data files (.dat) or in netCDF format (.nc). There are library routines for reading and writing netCDF format that can be obtained by anonymous FTP or use of a web browser. You can check out this example C++ code for reading from and writing to netCDF data files.

The indices (stem file names) are

  1. Eady growth rate (eady)
  2. Meridional temperature gradient (mtg)
  3. Temperature variance (tempvar)
  4. Extreme wind speed - sigma995 level (wind95,wind99)
  5. Extreme wind speed - geostrophic (wind95.geostr,wind99.geostr)
  6. Storm count (storms)
  7. Intense storm count (storms980,storms970)
The wind99 and storms970 indices are alternatives to wind95 and storms980, with different values for the cutoffs of the indices (99th percentile and 970 hPa).

Get the files here.

Last updated: August 2008.