A detailed description of data and data reduction proceedures is given in
Great Lakes States Average Monthly Temperature Data - Begining of Record
to 1990 report.   

                            TEMPERATURE FILE LIST

ILLINOIS                   T_IL_DAT.ZIP   IL-HST.ZIP

INDIANA                    T_IN_DAT.ZIP   IN-HST.ZIP

MICHIGAN                   T_MI_DAT.ZIP   MI-HST.ZIP

MINNESOTA                  T_MN_DAT.ZIP   MN-HST.ZIP

NEW YORK                   T_NY_DAT.ZIP   NY-HST.ZIP

OHIO                       T_OH_DAT.ZIP   OH-HST.ZIP

PENNSYLVANIA               T_PA_DAT.ZIP   PA-HST.ZIP

WISCONSIN                  T_WI_DAT.ZIP   WI-HST.ZIP


For each state there are two files having similar file names with the 
exception of the first two letters.  All of these files are ZIPPED
files, using PKZIP Version 2.04g.  The ZIPPED files must be transfered
using BINARY transfer mode to preserve the file integrity.


T_*_DAT files contain the monthly temperature data for each station in
that state.  These are ZIPPED files, with one file for each station.

*-HST files contain the station history information extracted from the
WB 530-1 forms by the National Climatic Data Center (NCDC).  These data
show the station number, latitude, longitude, elevation, station name,
and period of record for each station location.  These are ZIPPED files,
with one file for each state.  

The format of the unzipped files is as follows

T*.DAT

The name of a station's data file is based on that station's number. 
All file names follow the format of T<state #><station #>.DAT.  For 
example, Mansfield WSO AP is station number 4865 in Ohio, state number 33,
thus giving the file name T334865.DAT.  The first four lines of each data
file contain the header information.  The first line tells the type of
data (temperature).  The second line gives the NCDC station number,
latitude, longitude, station name, and elevation in feet above sea level.
The third line shows what period of years is covered.  The fourth line is
blank.  Following the header are data for each year covered.  The format
for these data are the year followed by the data for each month.  The
FORTRAN format is (1X,I4,2X,12F6.2).  Following the data is a blank line
followed by a footer. The footer contains notes regarding the data.  The
first line of the footer gives the data codes for missing values.  Some
files may also have information regarding combined data from nearby
locations, or station name changes.

*-HST.DAT

The file naming convention is state two letter abbreviation followed by
-HST.DAT.  The following table gives the column, item, and missing value 
identifier, if any, for the *_HST.DAT files.

COLUMNS  ITEM                       MISSING IDENTIFIER

01-06    STATION NUMBER             NONE

08-13    LATITUDE                   99.999

15-21    LONGITUDE                  -99.999

23-26    ELEVATION                  -999

28-48    STATION NAME               NONE

50-71    STATION NAME QUALIFIER     NONE

73-82    BEGINING DATE (MM/DD/YYYY) ALL 9's IN ANY FIELD

84-93    ENDING DATE (MM/DD/YYYY)   ALL 9's IN ANY FIELD, 99/99/9999 FOR
                                    EITHER UNKNOWN ENDING DATE, OR STATION
                                    STILL IN OPERATION AS OF 12/31/1990