Hi David:

I have been attempting to recreate your chart "A US Rural Data Set" in your presentation "Solar Cycle 24: Implications for the United States" in which the data is attributed to NASA GISS.

I have used the GISS GISTEMP station data site at the following url to download both the periurban adjusted and unadjusted data sets for the five stations listed: 
http://data.giss.nasa.gov/gistemp/station_data/
# Hawkinsville (32.3N, 83.5W),
# Glennville (31.3N, 89.1W)
# Calhoun Research Station (32.5N, 92.3W), 
# Highlands (35.0N, 82.3W)
# Talbotton(32.7N, 84.5W)

I have taken the simple mean of the five stations for each year excluding individual stations for years in which there is no annual data. The resulting average, even when smoothed in several ways including loess and spline, only roughly resembles your graph.

Which leads me to my questions: Can you supply the data or the methods (preferrably both) used in creating this chart? If not the original, do you have any recollection of how you may have done it, what tools or methods you might have used, or descriptions of how you create similar charts today? Perhaps the data you used was actually NOAA GHCN? Or USCHN?

I appreciate any help you can offer in replicating your work.

Thank You
Ron Broberg