SUPPLEMENTAL INFORMATION...IN GREATER DEPTH

DataStreme ECS Week Two: 8-12 February 2010

MEANS, EXTREMES IN CLIMATE RECORDS - A STATISTICIAN'S DELIGHT


The descriptive definition of climate involves the statistical summary of weather over a period of time at a location, a state, the nation or the planet. The development of a statistical summary represents the empirical definition of climate. This "climatography" or statistical description of the climate of a locale involves the determination of means (normals) and extremes of a compilation of weather data over an extended time interval at a station or group of stations. The averages and variability that comprise the climatography also provide a baseline for detecting when a change in the climate occurs.

WHAT IS RECORDED

Detailed instrumental weather records have been collected and compiled in this country at many stations for more than a century. The National Weather Service and its predecessor, the U.S. Weather Bureau, have operated a network of weather observation stations and offices in or near many of the large cities in every state, commonwealth and territory under its jurisdiction. During the last two decades, over 1000 automatic weather stations have been installed around the nation to make systematic measurements of numerous weather elements, such as temperature, precipitation, humidity, air pressure, cloud cover, visibility, wind speed and wind direction. Some of these weather data are recorded hourly, while other data are recorded once a day at some fixed time. An additional cooperative observer network of approximately 8000 volunteer observers provides daily readings of such weather elements as daily maximum and minimum temperatures, and 24-hour precipitation totals. Last week's Supplemental Information...In Great Depth introduced us to where we could access the current Preliminary Local Climatological Data for nearly 300 "first-order stations" around the nation from off the Web. Recall that the following weather information, which pertain to these first-order stations", is collected at essentially all types of stations, including the cooperative observer stations:

These daily data are used to compile monthly and annual averages of temperature, together with monthly and annual precipitation totals during that year. Ultimately, these data help define the location's empirical average climate. For example, the daily high temperatures observed for each of the 31 days at a particular weather station during the month of January are averaged to determine a monthly average high temperature for January. Similarly, a January monthly average minimum temperature is also calculated from the arithmetic average of the 31 daily low temperatures. The average monthly temperature is then determined as the average between the average monthly maximum and minimum temperatures for the given month. The annual average maximum, minimum and daily average temperatures are determined from the corresponding12 monthly averages. The monthly precipitation and snowfall totals are calculated from the sum of the daily values; likewise, the annual snowfall totals are calculated.

The monthly and annual data are published by NOAA's National Climatic Data Center (NCDC). As part of your investigations this week, you will become familiar with the wealth of climate information that is available in the annual issue of the Local Climatic Data that NCDC publishes for each of the nearly 300 cities around the United States and its territories.

"NORMALS" AND COMPARATIVE DATA

The annual summaries of the Local Climatic Data for the first-order stations include the normals and the extremes for the various weather elements at that particular location. While the weather is never truly normal, the term usually refers to a long-term average condition, typically of temperature or precipitation. By international convention, normals are computed for a standard three-decade (30-year) interval. Every 10 years this interval is shifted forward in time by a decade, and a new set of 30-year averages are computed. During the 1980s, the normals referred to the 1951 to 1980 interval, whereas during the 1990s, the 1961-90 interval was used. Currently, the normals for each station encompass the 1971-2000 interval. Climate normals for the 1981 to 2010 interval will be determined by NCDC and made available to the public by 2012.

Typically, the term normal is applied to the 30-year averages of temperature, precipitation, snow and heating/cooling degree-day unit data made for the reference interval, such as 1971-2000. Other weather elements, such as humidity measures (wet-bulb and dewpoint temperatures) and wind speed/direction data are averaged for the period of record and are included in the tabulation of "Normals and Comparative Data."

EXTREMES IN THE RECORD

Annual summaries also include the extremes in temperature, precipitation, and snow that have been observed at that current observing site. On each day, the observed daily weather data are compared with the long-term records that have been maintained at the weather station. At some stations, these daily weather records may extend back for more than a century. Since many of the weather stations have moved from city offices to airport locations, the noted extremes may not be the all-time extremes for that city.

For example, a record event report is made to the public if the day's temperature tied or exceeded the long-term record for that calendar date or season:

DAILY RECORD TEMPERATURES

The first and last records -- or simply, the daily record highs and lows -- are the most often reported, and are relatively obvious. The other records are reported less often and usually require some explanation. A record low maximum indicates that exceptionally cold conditions prevailed through the daylight hours, either because of a very cold air mass or because extensive cloud cover blocked solar heating. Record high minimums typically occur when a hot, humid air mass prevents overnight cooling, thereby keeping an elevated overnight minimum temperature.

MONTHLY AND SEASONAL RECORD TEMPERATURE EVENTS

Attention is given to those times when the maximum or minimum temperature exceeds the respective monthly temperature record. In addition, during meteorological spring (March, April, May), a low temperature record may be classified as "Low so late" or a record high temperature may be identified as "High so early". Similarly, during meteorological autumn (September, October, November) record temperatures could have the designations, "Low so early" or "High so late". No such designations are used during meteorological summer or winter other than monthly record exceeded.

ANNUAL RECORD TEMPERATURE EVENTS

Anytime the absolute lowest or highest temperature that was ever observed at the weather office is exceeded, the event would be identified as the "all time record low" or "all time high".

OTHER RECORD EVENTS

The public is usually informed when an exceptionally heavy precipitation event sets a daily, monthly or annual precipitation record. Such a record means that the amount of rain that has fallen on a given day, month or year is the greatest ever recorded for that given time interval. Snowfall records are similarly noted when applicable. However, in the latter case, the snowfall season used is from 1 July to 30 June rather than the calendar year as for total liquid equivalent precipitation.

On occasion, a station's record sea level pressure or record peak wind gust may be reported. Usually sufficient explanation is provided in these record event reports.

A WORD OF CAUTION

While a record event that eclipses a century old record may be newsworthy, one should realize that many of the first order weather stations may have moved from a city office to an airport office, resulting in a slightly different weather record. Changes in the types of thermometers also may pose a serious problem.

During the last several years, NOAA's National Climatic Data Center has undertaken a project called ThreadEX, for "threaded data extremes", in which climate bases have been assembled from the earlier 19th and 20th century data sets collected by the US Army Signal Service and the US Weather Bureau, predecessors to the current National Weather Service. The new ThreadEX records can be used by the public and researchers to study the longer-term climate extremes at a number of locations across the nation, especially where changes have occurred in station location . For access to the ThreadEX records for selected stations in your area, check http://threadex.rcc-acis.org/ for the daily temperature and precipitation records.

THE SIGNIFICANCE

What do these record weather events signify? One may think that some of these weather statistics are as obscure, detailed or complicated as some of the detailed baseball statistics that die-hard fans keep on their favorite player or club. While some of the weather and climate records may appear to have little meaning, they still provide benchmarks to be used to gauge the unusual behavior or severity of a particular weather event, just as the baseball statistics are used to judge athletic achievement. Furthermore, they can be used to judge how the climate variability or how the climate in a particular locale has been changing of the period of the instrumental record at that location.


Return to DataStreme ECS website

Prepared by Edward J. Hopkins, Ph.D., email hopkins@meteor.wisc.edu
© Copyright, 2010, The American Meteorological Society.