WeatherEater 2003
Introduction

WeatherEater 2003 is designed to read National Oceanic and Atmospheric (NOAA) weather observation data in different formats for different years and combine that data into a single output file. Data is output for a single NOAA weather station, with consistent units on an hourly, daily, monthly and annual basis. Data series can also be extracted for specific hours, based upon an input file specifying which hours are of interest. The output file is an ExcelÆ workbook, with several spreadsheets.

Please note: WeatherEater 2003 does not include the NOAA weather files that it is designed to read. These must be purchased separately from the National Climatic Data Center (NCDC). NCDCís phone number is 828-271-4800. The NCDC web address is http://www.ncdc.noaa.gov/.

Note, because not all NOAA weather files contain all of these data series, some data are missing for some periods. In addition WeatherEater 2003 calculates cooling and heating degree days and adjusted cooling and heating degree days for all temperature bases from 55 to 85, inclusive, and outputs them.

Hourly degree days are always calculated using the dry bulb temperature at that hour. Similarly, hourly cooling degree days adjusted for heat index are calculated using the heat index at that hour, hourly cooling degree days adjusted for temperature humidity index are calculated using the temperature humidity index at that hour and heating degree days adjusted for wind chill are calculated using the wind chill at that hour.

However, daily degree days can be calculated one of three different ways. The first is NOAAís definition. It calculates an average daily temperature by averaging the daily high and low temperatures and uses this as the temperature term in the degree day calculations. In the case of adjusted degree days, the high and low heat index, or the high and low temperature humidity index, are averaged for cooling degree day calculations and the high and low wind chill are averaged for heating degree day calculations.

Option 2, which is the default, is to obtain the true average temperature, heat index, temperature humidity index or wind chill by averaging all readings for the day. These results are then used in the degree day calculations.

The third option is to calculate daily degree days and daily adjusted degree days, by adding up the degree days and adjusted degree days from the hourly calculations.

Monthly and annual degree days are always summarized from daily degree days. So, by default, monthly and annual degree days use the same methodology as daily degree days.

Because WeatherEater 2003 makes temporary copies of some files on hard disks, and uses the temporary directory to store all files it creates, the user must have at least 500 MB of storage space available in the temporary directory. If not, WeatherEater 2003 will not run. If this occurs, point WeatherEater 2003 to any other drive and/or directory with at least this much free space.

WeatherEater 2003 starts by reading the raw NOAA data from four different sources (see NOAA Weather Files) into a temporary ACCESSÆ database that the program builds. After the data is read in, a series of queries are run against the ACCESSÆ database to extract the requested weather data.

The raw data is read from four different sets of NOAA weather files: Surface and Meteorological Surface Observations Network, Hourly US Weather Observations, Integrated Surface Hourly Observations (ISHO) and unedited data downloaded from the NOAA web site.

The ISHO and unedited data can contain multiple readings for any given hour. Therefore, WeatherEater 2003 first runs a summary query against the raw data to generate a table with one observation per hour. The individual weather data are averaged, or summed, as appropriate, across all hourly readings.

If a file containing peak times was specified (see Figure 1), WeatherEater 2003 reads the peak times into an ACCESSÆ table and creates a second table containing entries for each peak hour and the 72 prior hours, so that temperature, heat index, temperature humidity index and wind chill lags can be created.
WeatherEater 2003 queries the hourly weather table to retrieve and output hourly data that include degree days and adjusted degree days for any base from 55 to 85 degrees, inclusive, and the hourly for each weather variable. Next, WeatherEater 2003 daily data is calculated and output that include degree days and adjusted degree days for any base from 55 to 85 degrees, inclusive, and the minimum, average and maximum value for each weather variable. Once daily data is output, monthly and annual degree days and adjusted degree days for any base from 55 to 85 degrees, inclusive, are calculated and output.

If a peak times file was specified, WeatherEater 2003 outputs weather data for those specific hours as well as dry bulb temperatures, heat indices, temperature humidity indices and wind chills for the prior 72 hours. These allow the user to create his/her own heat build up terms.

After completion, WeatherEater 2003 cleans up after itself, and returns the user to the Main input screen (see Figure 1), where the program can be ended, or another NOAA weather station processed.

WeatherEater 2003 runs very efficiently. It only takes about 45 minutes to process NOAA weather data for a single weather station, from 1961 though 2001, with all NOAA data stored on CD-Rom, if hourly data is not output. The program will run even faster if the data files are transferred to disk before execut ing WeatherEater 2003?.

When it starts, WeatherEater 2003 assumes, by default, that all data files will be found on the drive identified as the CD-Rom drive on the Main input screen (see Figure 1). If the data has been moved to a hard drive, and the directory structure is maintained from the root down, just point the CD-Rom drive to the hard drive and WeatherEater 2003? will find the files.

If the directory structure has changed or WeatherEater 2003 cannot find the data files, it will prompt for their location. Once it finds the first file in a group (i.e. SAMSON, HUSWO, ISHO or unedited) it learns where to look for the remainder of the files in that group. It will only prompt if it cannot find files in the same relative location as the prior file, within the group.


Bakos & Richards Consulting
5016 Crane Court
Schererville, IN 46375
(219) 769-7188
info@bakosandrichards.com