STATION DATA USED BY PARKER, D.E., "A DEMONSTRATION THAT LARGE-SCALE WARMING IS NOT URBAN", J. CLIMATE, 19, 2882-2895 (2006). In the data files on this folder, Jan, Feb...Dec for a given year and day of the month runs along a row. Values are in tenths deg C with -999 = missing (generally denoting nonexistent days). Station details are in Parker_JClimate2006_StationDetails.doc Lists of stations by sources of data are in Station_list1.txt, Station_list2.txt, Station_list3.txt and Station_list4.txt The first row always constitutes a header. The format for an individual data row is '(I5,I5,12(I5))' (calendar year, calendar day, monthly values). The values are degrees C multiplied by ten, so to retrieve the recorded value divide these numbers by ten. This convention is a throwback to the days when the smaller memory space from storing integers rather than real numbers was a significant requirement. The convention has absolutely no impact on the underlying data quality.