• American Meteorological Society, 1910: Climatological districts of the United States. Mon. Wea. Rev., 38 (12), d.

  • Arndt, D. S., , M. O. Baringer & , and M. R. Johnson, Eds, 2010: State of the climate in 2009. Bull. Amer. Meteor. Soc., 91 (7), S1S222.

  • Bell, G. D., , E. Blake, , K. C. Mo, , C. W. Landsea, , R. Pasch, , M. Chelliah, , S. B. Goldenberg & , and H. J. Diamond, 2006: The record breaking 2005 Atlantic hurricane season [in “State of the Climate in 2005”]. Bull. Amer. Meteor. Soc., 87 (6), S44S45.

    • Search Google Scholar
    • Export Citation
  • Cerveny, R. S., , J. Lawrimore, , R. Edwards & , and C. Landsea, 2007: Extreme weather records: Compilation, adjudication, and publication. Bull. Amer. Meteor. Soc., 88, 853860.

    • Search Google Scholar
    • Export Citation
  • Easterling, D. R., , G. A. Meehl, , C. Parmesan, , S. A. Changnon, , T. R. Karl & , and L. O. Mearns, 2000: Climate extremes: Observations, modeling, and impacts. Science, 289, 20682074.

    • Search Google Scholar
    • Export Citation
  • Finneran, H. T., 1965: Preliminary inventory of operational and miscellaneous meteorological records of the Weather Bureau (record group 27). General Services Administration, National Archives and Records Service, National Archives, 160 pp.

    • Search Google Scholar
    • Export Citation
  • Harns, J. E., 1952: Storm study SA 5–8, 3–8 September 1950. U.S. Army Corps of Engineers, 201 pp.

  • Jennings, A. H., 1950: World's greatest observed point rainfalls. Mon. Wea. Rev., 78, 45.

  • Kerkam, R. E., 1895: Snowfall in February, 1895. La. Wea. J. Agriculturalist, 8, 4.

  • Lawson, T., 1851: Meteorological Register for Twelve Years, from 1831 to 1842 Inclusive: Compiled from Observations Made by the Officers of the Medical Department of the Army at the Military Posts of the United States. C. Alexander, 324 pp.

    • Search Google Scholar
    • Export Citation
  • National Climatic Data Center, cited 2006: US snow climatology background. [Available online at www.ncdc.noaa.gov/ussc/USSCAppController?action=scproject.]

  • National Weather Service, 1997: Evaluation of the reported January 11–12, 1997, Montague, New York, 77-inch, 24-hour lake-effect snowfall. U.S. Department of Commerce, 41 pp.

    • Search Google Scholar
    • Export Citation
  • National Weather Service, 2011: National Weather Service instruction 10- 1004. NWS Publ. NWSPD 10-10, 43 pp. [Available online at www.nws.noaa.gov/directives/sym/pd01010004curr.pdf.]

    • Search Google Scholar
    • Export Citation
  • New England Meteorological Society, 1885: Summary for January, 1885. Bull. New England Meteor. Soc., 3, 38.

  • Quetelard, H., , P. Bessemoulin, , R. S. Cerveny, , T. C. Peterson, , A. Burton & , and Y. Boodhoo, 2009: World record rainfalls during tropical cyclone Gamede. Bull. Amer. Meteor. Soc., 90, 603608.

    • Search Google Scholar
    • Export Citation
  • Ross, T. F. & , and R. T. Truesdell, 2010: NOAA's Climate Database Modernization Program—A decade of data rescue and modernization activities. Eighth Presidential History Symp., Atlanta, GA, Amer. Meteor. Soc., J7.2. [Available online at https://ams.confex.com/ams/90annual/techprogram/paper_159403.htm.]

    • Search Google Scholar
    • Export Citation
  • SCEC, 2009: Record minimum temperature in Maine. State Climate Extremes Committee Memo. to NCDC Director, Asheville, NC, 8 pp. [Available online at http://www1.ncdc.noaa.gov/pub/data/cmb/scec/reports/SCEC-ME-TMIN-20090116.pdf.]

    • Search Google Scholar
    • Export Citation
  • Shein, K. A., 2008: Interactive quality assurance practices. Preprints, 24th Conf. on Interactive Information Processing Systems (IIPS), New Orleans, LA, Amer. Meteor. Soc., 6A.9. [Available online at https://ams.confex.com/ams/88Annual/techprogram/paper_131217.htm.]

    • Search Google Scholar
    • Export Citation
  • Shein, K. A., , D. P. Todey, , F. A. Akyuz, , J. R. Angel, , T. M. Kearns & , and J. L. Zdrojewski, 2012: Evaluating statewide climate extremes for the United States. J. Appl. Meteor. Climatol., 51, 20472059.

    • Search Google Scholar
    • Export Citation
  • U.S. Weather Bureau, 1911: Climatological Service, District No. 9: Colorado Valley. U.S. Department of Agriculture, 11 pp.

  • U.S. Weather Bureau, 1942: Daily and hourly precipitation supplement storm of July 17–18, 1942 New York–Pennsylvania. U.S. Department of Commerce, 40 pp.

    • Search Google Scholar
    • Export Citation
  • Wilson, W. M., 1898: The snow storm of February 19–20, 1898. U.S. Weather Bureau Rep. for February 1898, Vol. III, No. 2 (Wisconsin section), 8 pp.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Map of the climatological districts of the United States used by the U.S. Weather Bureau in the early 1900s (American Meteorological Society 1910).

  • View in gallery

    U.S. Weather Bureau 1009 observation form from Perryville, Tennessee, for Aug 1930 showing a maximum temperature of 113°F on 9 Aug. In the remarks area, the reviewing official wrote, “The Perryville maximum record of July [also 113°F] and early August was questioned but the observer replied that the thermometer was always lowered before reading and he was sure readings were correct.”

  • View in gallery

    Narrative from the Louisiana Weather Journal and Agriculturalist, including comparative snow depth observations from various weather stations around southern Louisiana from the storm of 14–15 Feb 1895 supporting an observation of 24-in. snow depth at Rayne, Louisiana (Kerkam 1895).

  • View in gallery

    (a) U.S. Weather Bureau map from Mon. Wea. Rev. of Feb 1895 showing total snow depth and minimum isotherms for the month. (b) U.S. Weather Bureau surface weather map for 15 Apr 1921 in association with a record-setting 75.8-in. 24-h snowfall at Silver Lake, Colorado.

  • View in gallery

    (a) Photograph of the McIlroy family standing amidst heavy snow on 21 Dec 1929 in Vaughan, Texas. A storm on 20–21 Dec brought between 20 and 26 in. to the region. Many years later James R. “Mac” McIlroy (the boy in the photo) was still able to describe the storm in great detail to the NWS Fort Worth Climate Focal Point. (Courtesy of McIlroy family and NWS Fort Worth.) (b) Newspaper clipping from the Hillsboro Mirror (Texas) on 21 Dec 1929 describing the strong winter storm that brought over 20 in. of snow to the region. (Courtesy of NWS Fort Worth.)

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 7 7 1
PDF Downloads 4 4 1

Revisiting the Statewide Climate Extremes for the United States: Evaluating Existing Extremes, Archived Data, and New Observations

View More View Less
  • 1 NOAA/National Climatic Data Center, Asheville, North Carolina
  • 2 South Dakota Office of Climatology, South Dakota State University, Brookings, South Dakota
  • 3 North Dakota State University, Fargo, North Dakota
  • 4 Illinois State Climatologist Office, Illinois State Water Survey, Prairie Research Institute, University of Illinois at Urbana–Champaign, Urbana, Illinois
  • 5 NOAA/NWS/WFO Aberdeen, Aberdeen, South Dakota
  • 6 NOAA/NWS/Climate Services Division, Silver Spring, Maryland
© Get Permissions
Full access

New all-time extreme climate records have been set in several states over the past few years. These records highlighted a need to review the existing statewide climate extremes tables maintained by the NOAA National Climatic Data Center (NCDC). Also, since these tables were last up-dated, NCDC has greatly extended its digital data record into the past for many locations and has applied improved quality assurance processes to its archived data, revealing several potential new record values. To ensure the records maintained in the statewide climate extremes tables accurately reflect the most current and valid data available, the records were reevaluated. The all-time maximum and minimum temperature, all-time greatest 24-h precipitation and snowfall, and all-time greatest snow depth for each of the 50 states, Puerto Rico, and the U.S. Virgin Islands were manually examined to determine their validity, accuracy, accessibility, and provenance. NCDC's data holdings were scoured for values that might exceed established records, and the validity of each potentially record-breaking observation was evaluated. The revised extremes tables were vetted by the National Weather Service, regional climate centers, and state climatologists to ensure agreement. In conjunction with this revision, a new state climate extremes evaluation process has been established to formally consider any potential challenges to the existing records and update the records tables as necessary.

CORRESPONDING AUTHOR: Dr. Karsten A. Shein, NOAA/National Climatic Data Center, 151 Patton Ave., Asheville, NC 28801-5111, E-mail:karsten.shein@noaa.gov

New all-time extreme climate records have been set in several states over the past few years. These records highlighted a need to review the existing statewide climate extremes tables maintained by the NOAA National Climatic Data Center (NCDC). Also, since these tables were last up-dated, NCDC has greatly extended its digital data record into the past for many locations and has applied improved quality assurance processes to its archived data, revealing several potential new record values. To ensure the records maintained in the statewide climate extremes tables accurately reflect the most current and valid data available, the records were reevaluated. The all-time maximum and minimum temperature, all-time greatest 24-h precipitation and snowfall, and all-time greatest snow depth for each of the 50 states, Puerto Rico, and the U.S. Virgin Islands were manually examined to determine their validity, accuracy, accessibility, and provenance. NCDC's data holdings were scoured for values that might exceed established records, and the validity of each potentially record-breaking observation was evaluated. The revised extremes tables were vetted by the National Weather Service, regional climate centers, and state climatologists to ensure agreement. In conjunction with this revision, a new state climate extremes evaluation process has been established to formally consider any potential challenges to the existing records and update the records tables as necessary.

CORRESPONDING AUTHOR: Dr. Karsten A. Shein, NOAA/National Climatic Data Center, 151 Patton Ave., Asheville, NC 28801-5111, E-mail:karsten.shein@noaa.gov

Ensuring valid climate extremes for each state requires digging into the archives as well as establishing a new, objective methodology for evaluating challengers.

Extreme weather and climate conditions hold an important place in both the atmospheric sciences as well as the public psyche. Severe and extreme weather events, such as blizzards, heat waves, and hurricanes, often have a profound impact on society and spur a great deal of interest by scientists, policy makers, and the public. We want to know what was the hottest, the coldest, the wettest, and the driest. Each year, this journal publishes the “State of the Climate” (e.g., Arndt et al. 2010), which not only describes Earth's climate, but ranks that year's conditions relative to the historical record, making special note of newly established records, such as the record-setting Atlantic tropical cyclone season of 2005 (Bell et al. 2006).

Beyond simple human interest, however, the quantification of extreme conditions has tangible benefits to society and to science. Record values, along with lesser but still extreme meteorological values, establish exceedance levels and return intervals that are, for example, used to plan infrastructure design limits, set development restrictions, and develop public safety and disaster recovery plans. It is verifiable, observed extremes that provide a conceptual understanding of the range of conditions that can be expected at a given location or over a certain area. Extremes, or rather changes in their frequency and magnitude, also are of great interest to those investigating climate changes and associated consequences (e.g., Easterling et al. 2000). Biased or incorrect observations of extremes can lead to erroneous conclusions regarding Earth's climate. For these reasons, it is critical that the public have the utmost confidence in observations of extreme values, and why a great deal of effort must be placed on ensuring each stated extreme is accurate and defensible.

A HISTORY OF STATEWIDE EXTREMES.

Fortunately, the substantial public interest in the extremes of our climate is not recent, and has resulted in the identification and tracking of record values almost since the inception of formal weather monitoring. Indeed, under the auspices of the World Meteorological Organization, the international meteorological community has tracked global weather extremes for many years (e.g., Quetelard et al. 2009). Although individuals throughout the United States were independently making weather observations as early as the mid-1700s, systematic weather observations as part of a national observing network began in 1814. The preservation of these observations allowed the nation's climatic conditions to be characterized. Summary statistics from observational records made by U.S. Army surgeons at various military posts across the expanding United States first appeared in 1826 and covered the period 1822–25 (Lawson 1851). However, it was not until around 1890, when the U.S. Weather Bureau was transferred from the War Department to the Department of Agriculture, that most individual state weather offices began to routinely issue monthly climate reports in which averages and extremes of some meteorological elements (e.g., maximum temperature and total precipitation) from preceding months and years were regularly published as comparative data, facilitating the tracking of extremes.

Until the 1960s, these reports also included descriptive narratives of the month's weather, often mentioning the extremes for the month and whether they surpassed any historical records. The comparative data tables only extend back to the early 1880s, coinciding with improved centralization and reporting by the U.S. Signal Corps (Finneran 1965). As a consequence, references to earlier extreme events are largely anecdotal [e.g., “Not since the memorable snowstorm of March 2–4, 1881, has there been a snowstorm of equal severity in this locality,” stated in Wilson (1898, p. 3)]. Furthermore, climatological reports that were published in the early part of the twentieth century were divided by Weather Bureau district (Fig. 1). Some states contained multiple districts, while some districts contained parts of several states (e.g., U.S. Weather Bureau 1911). Although observational divisions that adhere to state boundaries may not be optimal from a climatological analysis standpoint, this incongruity between states and districts precluded effective tracking of individual state weather records during that period as comparative statistics were supplied for the district, rather than for individual states.

Fig. 1.
Fig. 1.

Map of the climatological districts of the United States used by the U.S. Weather Bureau in the early 1900s (American Meteorological Society 1910).

Citation: Bulletin of the American Meteorological Society 94, 3; 10.1175/BAMS-D-11-00013.1

More recently, efforts have been made to continually report and track both national and statewide weather records. Tables of record extremes by state are maintained at the National Oceanic and Atmospheric Administration (NOAA) National Climatic Data Center (NCDC), and are tracked by both the National Weather Service (NWS) local and regional offices, as well as individual state climatologist offices. But while national records have traditionally undergone extensive review and evaluation by a panel of experts (National Weather Service 1997; Cerveny et al. 2007), the verification of statewide records has, in many cases, been less rigorous. Some notable record-setting meteorological values have been subsequently reviewed by climate experts (e.g., Harns 1952), but most have simply been handed down to posterity as valid. Additionally, late-arriving observational reports sometimes ensured that a potential record was left undiscovered—buried in the archives.

In 2006, a maximum temperature observation from an NWS Cooperative Observer (COOP) network observer challenged the all-time maximum temperature for South Dakota. While the value was eventually declared invalid, the investigation revealed that several years had passed since tables of statewide climatological records had been reviewed. Statewide all-time maximum and minimum temperature extremes tables were last updated in December 2003, while the all-time greatest 24-h precipitation and all-time maximum and minimum annual precipitation tables had not been revised since 1998. Since that time, the digitization and transcription efforts of the Climate Database Modernization Program (CDMP) added millions of historical observations from paper archives to NCDC's digital datasets (Ross and Truesdell 2010). Statewide snow records (e.g., 24-h snowfall and maximum snow depth) had been incorporated into the U.S. Snow Climatology (USSC) project and statewide extremes were therefore available through its most recent update in 2006, but were limited to automatic quality-controlled (QC) data residing in NCDC's digital daily datasets (National Climatic Data Center 2006).

The lack of a formalized methodology for adjudicating extreme values and the need to update the statewide climate extremes tables maintained at NCDC were taken up by the National Data Stewardship Team (NDST). The NDST was formed in 2004 to address national-level issues related to climate data handling between NOAA and its climate service partners, and consists of representatives from the NWS, NCDC, regional climate centers, state climatologists, several federal agencies, and academia. As a result of the South Dakota record challenge, the NDST convened an extremes subcommittee (i.e., the authors) to address the issues. Based on the subcommittee's suggestions, the NDST established the framework for a State Climate Extremes Committee (SCEC) to evaluate future potential records. The NDST also recommended that NCDC review and revise the existing statewide records tables to bring them up to date.

STATE CLIMATE EXTREMES COMMITTEE.

The SCEC is an ad hoc committee that is convened at the discretion of a state climatologist or local NWS official when the person believes an observation may have tied or exceeded an existing climate record. The committee consists of voting representatives from NCDC, the relevant regional climate center, the regional NWS headquarters, the local NWS Weather Forecast Office (WFO), and the state climatologist. These members review all available information surrounding the existing record extreme value and the observation challenging that record. Based on the evidence, the SCEC makes a determination regarding the validity of the observation and whether the observation can be recognized as a new climate extreme for the state. In addition, an SCEC may be convened if the validity of an existing record extreme is challenged by new information.

To ensure openness and objectivity, the operational guidelines for the SCEC are codified in NWS Instruction 10-1004 (National Weather Service 2011) and on the SCEC website (www.ncdc.noaa.gov/extremes/scec). Any observations, regardless of observation network, may be considered by the SCEC, with the caveat that they be evaluated by either the state climatologist or the local NWS WFO before the committee is convened. Challenger observations must meet certain minimum observational, archival, and access standards to be considered as “official” in the context of extremes recognition (National Weather Service 2011).

Since 2006, the SCEC has been convened 14 times, resulting in 12 newly established statewide climate records (one—a hail stone that fell at Vivian, South Dakota, in July 2010—also set a national record for diameter and mass). An SCEC, in establishing a new hail stone size record for Vermont discovered that no formal guidelines existed for the collection, measurement, and preservation of hail. Rather, there were only scattered references from the NWS and World Meteorological Organization regarding reporting of hail size. As a result, in 2009 the SCEC produced hail collection and measurement guidelines that it uses when evaluating hail challenges (http://www1.ncdc.noaa.gov/pub/data/cmb/extremes/scec/reports/SCEC-Hail-Guide.pdf).

REVISING THE STATEWIDE EXTREMES TABLES.

Based on the NDST suggestion, NCDC began to reexamine each value listed in its existing climate extremes tables for all-time maximum and all-time minimum temperature, 24-h precipitation, 24-h snowfall, and maximum snow depth. These five variables are the most widely observed daily meteorological parameters for the United States, observed at thousands of weather stations.

The revision process was a multistep process that is more fully described in Shein et al. (2012). We began by manually evaluating the existing record values for each state. Next, we extracted all possible challenger values from NCDC's daily and hourly digital datasets. We also looked for extreme observations in undigitized forms from notable historical extreme events (e.g., blizzards and hurricanes). All observations that could be considered a record extreme were manually evaluated against available evidence to determine validity. In some cases, we determined that automated QC processes had incorrectly flagged the validity of the value, and we submitted the error to the “Datzilla” error correction system at NCDC for review (Shein 2008). Finally, we distributed the revised list to state climate experts (e.g., state climatologists and NWS) for vetting. Any objections to a value resulted in a joint reevaluation to ensure all parties were in agreement with a proposed revision.

In reviewing each existing record or potential challenger, several criteria were used to adjudicate validity. Since many of the values were observed on dates preceding any living memory, or at least occurred several decades ago, interviewing the observer or relevant Weather Bureau/NWS employees would have been impossible or else impractical. Thus, we relied first and foremost on original observation forms (e.g., Fig. 2) and any official meteorological reports. In a nod to the play review rules used by several sports, we assumed a value was accurate unless there was clear and compelling evidence to invalidate it.

Fig. 2.
Fig. 2.

U.S. Weather Bureau 1009 observation form from Perryville, Tennessee, for Aug 1930 showing a maximum temperature of 113°F on 9 Aug. In the remarks area, the reviewing official wrote, “The Perryville maximum record of July [also 113°F] and early August was questioned but the observer replied that the thermometer was always lowered before reading and he was sure readings were correct.”

Citation: Bulletin of the American Meteorological Society 94, 3; 10.1175/BAMS-D-11-00013.1

The observation form was quite often sufficient to declare a potential record invalid. These forms often revealed that the original observation had been incorrectly transcribed when it was digitized. For example, numbers were frequently transposed (e.g., 49 becomes 94), the wrong digit was typed (e.g., 108 becomes 118), or two keys were typed at once (e.g., a 15 becomes a 125). Other errors we documented included the observer using the wrong column on the observing form (e.g., writing snowfall in the precipitation column), omitting a decimal point (e.g., recording 0.41 in. of precipitation as “41”), or recording an observation to the wrong precision (e.g., 3.00 in. snow depth instead of 3 in.). Since the transcribers were restricted to typing the digits as they appeared on the form (omitting any decimal points), these observer errors propagated into the digital data (e.g., 3.00 in. snow depth became 300 in.).

When a digital observation matched the value written on the form, and the observation did not appear out of line for the observation month, we examined secondary information to support the observation. This evidence included observations from surrounding stations and postevent summaries such as those contained in monthly climatological reports (e.g., Fig. 3). Weather maps served to place the event in a synoptic context (e.g., Figs. 4a,b). Tertiary supporting materials such as media articles or photographs (e.g., Figs. 5a,b) were also utilized when they could be located. Because the observations being evaluated are extreme in nature, official reports from the National Weather Service or its predecessors would be expected to mention and corroborate the extreme if it was considered legitimate.

Fig. 3.
Fig. 3.

Narrative from the Louisiana Weather Journal and Agriculturalist, including comparative snow depth observations from various weather stations around southern Louisiana from the storm of 14–15 Feb 1895 supporting an observation of 24-in. snow depth at Rayne, Louisiana (Kerkam 1895).

Citation: Bulletin of the American Meteorological Society 94, 3; 10.1175/BAMS-D-11-00013.1

Fig. 4.
Fig. 4.

(a) U.S. Weather Bureau map from Mon. Wea. Rev. of Feb 1895 showing total snow depth and minimum isotherms for the month. (b) U.S. Weather Bureau surface weather map for 15 Apr 1921 in association with a record-setting 75.8-in. 24-h snowfall at Silver Lake, Colorado.

Citation: Bulletin of the American Meteorological Society 94, 3; 10.1175/BAMS-D-11-00013.1

Fig. 5.
Fig. 5.

(a) Photograph of the McIlroy family standing amidst heavy snow on 21 Dec 1929 in Vaughan, Texas. A storm on 20–21 Dec brought between 20 and 26 in. to the region. Many years later James R. “Mac” McIlroy (the boy in the photo) was still able to describe the storm in great detail to the NWS Fort Worth Climate Focal Point. (Courtesy of McIlroy family and NWS Fort Worth.) (b) Newspaper clipping from the Hillsboro Mirror (Texas) on 21 Dec 1929 describing the strong winter storm that brought over 20 in. of snow to the region. (Courtesy of NWS Fort Worth.)

Citation: Bulletin of the American Meteorological Society 94, 3; 10.1175/BAMS-D-11-00013.1

Critical to this effort was establishing clear provenance to any value considered an official statewide extreme. We wanted all listed values to be fully defensible, and our evaluation process and all evidentiary documentation to be open to scrutiny by the scientific community. Unfortunately, some values on the existing extremes tables could not be supported in this regard. In a few instances, we discovered that the listed record value was simply in error. For example, the 118°F (47.8°C) all-time maximum temperature for Colorado (Bennett on 11 July 1888) was well out of line with most statewide maximum temperatures in the 90° to 105°F range (32° to 40°C range) on that date. Although a U.S. Signal Service voluntary observer was located in Bennett, U.S. National Archives records show no observation form exists for July 1888 (Finneran 1965). Bennett was first included in the Colorado monthly climatological summary in January 1893. However, tracking changes to the value through time sheds some light on the possible discrepancies. The statewide maximum high temperature of 105°F (40.6°C; Glenwood Springs, same date) for July 1888 was altered to 115°F (46.1°C) in the July 1900 climatological summary. In July 1948 (coincidentally the year these data were first transcribed to magnetic media), the value was further changed to 118°F. We are uncertain how Bennett became associated with this value. However, even if a July 1888 observation form from Bennett was located that showed a 118°F observation, the value would remain questionable, as it is substantially higher than all surrounding observations. The Colorado state climatologist's office agreed with this assessment (N. Doesken 2010, personal communication).

More commonly, the existing extreme values came from sources that were not recognized observation stations, and that a number of the records— principally precipitation—were estimated values rather than actual observations from an operational, purpose-built instrument (e.g., graduated rain gauge). These values commonly did not have any direct record of observation, but often had ample secondary documentation because of their extreme nature. In many of these cases, the observation was obtained from a postevent “bucket” survey of in situ containers (e.g., buckets, oil drums, and milk cans). However, despite the appearance of these values in official documents, all must be considered estimates having varying degrees of accuracy (and not official extreme values) because the disposition of the collecting vessel and any liquid it may have contained prior to the storm was not documented.

Two examples stand out in this regard. The first is 34.50 in. (876 mm) of rain that fell at Smethport, Pennsylvania, on 17 July 1942. This value is widely acknowledged as legitimate (Jennings 1950), and was estimated from the volume of water [37 quarts (35 L)] present in a 40-quart (37.9 L) milk can (remarks indicate “Can filled to a point near top”; U.S. Weather Bureau 1942, p. 14). But, the dimension of the can mouth and precise quantity of water added during the storm, both of which would be required to accurately derive volumetric depth, can only be approximated. Similarly, 38.70 in. (983 mm) of rain that fell in Yankeetown, Florida, during the passage of Hurricane Easy on 5 September 1950 is considered legitimate and is officially documented (Harns 1952). But substantial uncertainties and assumptions surround the accumulation of that precipitation [it was mathematically estimated from filled 12-oz (0.35 L) soda bottles that had been found in a crate on the sidewalk after the storm], and as an extra-station estimate, observational data were not preserved. These and other similar examples necessitated the inclusion of an addendum to the tables. In this addendum, we acknowledge the likely validity of the values in an unofficial context. Overall, the review of existing record extreme values resulted in seven values being invalidated and another six becoming unofficial.

The inclusion of all recognized record extreme values (both official and unofficial) is important to understanding the extreme nature of climate. Record extreme meteorological conditions do not necessarily coincide with officially sanctioned observation stations. Thus, the ability to observe a record climate extreme is affected by the number of available observing stations, and this number has varied widely, from a few hundred throughout most of the 1800s and early 1900s to over 10,000 today (Finneran 1965; www.nws.noaa.gov/om/coop/what-is-coop.html). When the density of a station network is insufficient, some extremes (especially precipitation) may be registered by other means (e.g., rain collecting in a bucket). Additionally, in recent years, a number of non-NOAA weather observation networks have been established (e.g., state mesonets and voluntary cooperatives) that greatly enhance the density of weather observations throughout the United States. Although the revision of the NCDC statewide climate extremes tables was undertaken using only federally collected weather observations archived at NCDC, we decided that any observation could be considered official if it was obtained from a station situated according to minimum federal siting standards, the station was operated with the intent of collecting a long-term series of observations, it was not located for the purposes of intentionally capturing extreme microclimatological observations, it had comprehensive and unambiguous metadata, and that it was openly and indefinitely archived for public review. A few such values have been suggested by state climatologists, and at least one temperature observation collected by the U.S. Geological Survey was evaluated by the SCEC and determined to be a new all-time minimum temperature record for Maine (SCEC 2009). It is likely that the substantial expansion of high-quality weather networks will contribute to an increase in the frequency of record-setting observations. Thus, the role of the SCEC to investigate these observations as they occur becomes more important to maintaining a current listing of statewide climate extremes.

RESULTS.

As a result of the nearly 3-yr effort, the aforementioned statewide climate extremes tables were brought up to date. Of the 250 reviewed record values (the five elements for 50 states), 104 were revised. The fewest updates occurred with maximum and minimum temperature (seven each). Seventeen updates were made to 24-h precipitation, while snow depth and 24-h snowfall had the most changes—36 and 37, respectively. Tracking was also newly established for all-time maximum temperature, all-time minimum temperature, and all-time greatest 24-h precipitation in Puerto Rico and the U.S. Virgin Islands. The revised tables are available on the SCEC website (www.ncdc.noaa.gov/extremes/scec/).

Review of the revised tables by the members of the American Association of State Climatologists and the regional and local climate focal points of the NWS led to 12 values being questioned. Each of these values was investigated again in the context of additional information in the possession of the local NWS or state climate office. Most of the questions concerned the relegation of previously accepted values to “unofficial” status, but all agreed that the distinction was necessary in order to ensure all official values were supported by accurate and defensible observations. In a few instances, valid observations that were archived outside of NCDC were identified and subsequently were included as official record extremes (e.g., snow depth for Washington). In others, information held by a state or local office revealed questionable siting or observing practices that were sufficient to invalidate what had been thought to be a legitimate observation.

As this investigation has shown, the evaluation and verification of climate extremes is an ongoing process in which a listed extreme value may be updated unexpectedly as new information or new observations become available. Even while searching for additional information to support this summary, a climatological review from the Harvard College Observatory (New England Meteorological Society 1885, p. 3) revealed the existence of an undigitized minimum temperature observation from Mount Washington on 22 January 1885, indicating that it “was the lowest ever observed at that station.” The original observation form and monthly climate summary for New England for January 1885 were found, and the observation of −50°F (−45.6°C) was determined to be legitimate, subsequently replacing the existing New Hampshire record extreme (also at Mount Washington) of −46°F (−43.3°C).

Despite our best efforts at objectivity, some subjectivity is involved in adjudicating historical records. Forms may be missing or illegible, and an absence of contrary evidence does not always mean a value thought to be legitimate is actually valid. In addition, every day thousands of new observations are added to the climate record, and the digitization of historical records extends that record farther back in time. Our nation's climate observation legacy dates back to the mid-1700s, and systematic nationwide observations began in earnest in 1814 (Finneran 1965), but our current digital record contains relatively few observations predating 1900. This means that each day brings the possibility of a new climate record being observed or discovered. To that end, the SCEC now evaluates any potential challenges to the existing records and updates the records tables as necessary (cf. National Weather Service 2011). Such attention ensures that the tables will no longer be allowed to become outdated and inaccurate, and that each potentially record-breaking observation will be thoroughly examined by atmospheric scientists and data experts to ensure its validity.

The work of the SCEC in reviewing extreme observations also helps to reduce the amount of effort that is currently placed on the periodic review of the entire suite of climate extremes tables. Instead, an individual table entry will be reviewed each time the SCEC investigates a particular record value. This new review process will allow NCDC to direct its efforts to updating additional statewide climate extremes tables, such as greatest annual precipitation or highest temperature by month. Also, the SCEC has the leeway to investigate and approve records for meteorological variables not routinely tracked by NCDC (such as hail size). The next step in this work, however, is for the addition of tracking for record-setting observations from the U.S.-affiliated Pacific islands (i.e., Guam, American Samoa, Republic of Palau, Federated States of Micronesia, Marshall Islands, and U.S. Minor Outlying Islands) in the five revised tables. Observations from these islands have traditionally been compiled under a general “Pacific Islands” category, and will need to be associated with each individual island before they can be evaluated.

ACKNOWLEDGMENTS.

The authors wish to acknowledge the extensive work of Michael Changery, Grant Goodge, and Tom Ross of NCDC and Robert Leffler and Andrew Horvitz of NWS in creating the previous NCDC records tables, along with the efforts of NCDC's Datzilla administrator, Bryant Korzeniewski, to ensure erroneous data found during the course of this investigation were corrected. Also, we acknowledge the invaluable work of the employees of the National Climatic Data Center (and its predecessors), as well as the thousands of volunteer weather observers, whose dedication has made this effort possible. We thank the anonymous reviewers, whose suggestions were extremely helpful.

REFERENCES

  • American Meteorological Society, 1910: Climatological districts of the United States. Mon. Wea. Rev., 38 (12), d.

  • Arndt, D. S., , M. O. Baringer & , and M. R. Johnson, Eds, 2010: State of the climate in 2009. Bull. Amer. Meteor. Soc., 91 (7), S1S222.

  • Bell, G. D., , E. Blake, , K. C. Mo, , C. W. Landsea, , R. Pasch, , M. Chelliah, , S. B. Goldenberg & , and H. J. Diamond, 2006: The record breaking 2005 Atlantic hurricane season [in “State of the Climate in 2005”]. Bull. Amer. Meteor. Soc., 87 (6), S44S45.

    • Search Google Scholar
    • Export Citation
  • Cerveny, R. S., , J. Lawrimore, , R. Edwards & , and C. Landsea, 2007: Extreme weather records: Compilation, adjudication, and publication. Bull. Amer. Meteor. Soc., 88, 853860.

    • Search Google Scholar
    • Export Citation
  • Easterling, D. R., , G. A. Meehl, , C. Parmesan, , S. A. Changnon, , T. R. Karl & , and L. O. Mearns, 2000: Climate extremes: Observations, modeling, and impacts. Science, 289, 20682074.

    • Search Google Scholar
    • Export Citation
  • Finneran, H. T., 1965: Preliminary inventory of operational and miscellaneous meteorological records of the Weather Bureau (record group 27). General Services Administration, National Archives and Records Service, National Archives, 160 pp.

    • Search Google Scholar
    • Export Citation
  • Harns, J. E., 1952: Storm study SA 5–8, 3–8 September 1950. U.S. Army Corps of Engineers, 201 pp.

  • Jennings, A. H., 1950: World's greatest observed point rainfalls. Mon. Wea. Rev., 78, 45.

  • Kerkam, R. E., 1895: Snowfall in February, 1895. La. Wea. J. Agriculturalist, 8, 4.

  • Lawson, T., 1851: Meteorological Register for Twelve Years, from 1831 to 1842 Inclusive: Compiled from Observations Made by the Officers of the Medical Department of the Army at the Military Posts of the United States. C. Alexander, 324 pp.

    • Search Google Scholar
    • Export Citation
  • National Climatic Data Center, cited 2006: US snow climatology background. [Available online at www.ncdc.noaa.gov/ussc/USSCAppController?action=scproject.]

  • National Weather Service, 1997: Evaluation of the reported January 11–12, 1997, Montague, New York, 77-inch, 24-hour lake-effect snowfall. U.S. Department of Commerce, 41 pp.

    • Search Google Scholar
    • Export Citation
  • National Weather Service, 2011: National Weather Service instruction 10- 1004. NWS Publ. NWSPD 10-10, 43 pp. [Available online at www.nws.noaa.gov/directives/sym/pd01010004curr.pdf.]

    • Search Google Scholar
    • Export Citation
  • New England Meteorological Society, 1885: Summary for January, 1885. Bull. New England Meteor. Soc., 3, 38.

  • Quetelard, H., , P. Bessemoulin, , R. S. Cerveny, , T. C. Peterson, , A. Burton & , and Y. Boodhoo, 2009: World record rainfalls during tropical cyclone Gamede. Bull. Amer. Meteor. Soc., 90, 603608.

    • Search Google Scholar
    • Export Citation
  • Ross, T. F. & , and R. T. Truesdell, 2010: NOAA's Climate Database Modernization Program—A decade of data rescue and modernization activities. Eighth Presidential History Symp., Atlanta, GA, Amer. Meteor. Soc., J7.2. [Available online at https://ams.confex.com/ams/90annual/techprogram/paper_159403.htm.]

    • Search Google Scholar
    • Export Citation
  • SCEC, 2009: Record minimum temperature in Maine. State Climate Extremes Committee Memo. to NCDC Director, Asheville, NC, 8 pp. [Available online at http://www1.ncdc.noaa.gov/pub/data/cmb/scec/reports/SCEC-ME-TMIN-20090116.pdf.]

    • Search Google Scholar
    • Export Citation
  • Shein, K. A., 2008: Interactive quality assurance practices. Preprints, 24th Conf. on Interactive Information Processing Systems (IIPS), New Orleans, LA, Amer. Meteor. Soc., 6A.9. [Available online at https://ams.confex.com/ams/88Annual/techprogram/paper_131217.htm.]

    • Search Google Scholar
    • Export Citation
  • Shein, K. A., , D. P. Todey, , F. A. Akyuz, , J. R. Angel, , T. M. Kearns & , and J. L. Zdrojewski, 2012: Evaluating statewide climate extremes for the United States. J. Appl. Meteor. Climatol., 51, 20472059.

    • Search Google Scholar
    • Export Citation
  • U.S. Weather Bureau, 1911: Climatological Service, District No. 9: Colorado Valley. U.S. Department of Agriculture, 11 pp.

  • U.S. Weather Bureau, 1942: Daily and hourly precipitation supplement storm of July 17–18, 1942 New York–Pennsylvania. U.S. Department of Commerce, 40 pp.

    • Search Google Scholar
    • Export Citation
  • Wilson, W. M., 1898: The snow storm of February 19–20, 1898. U.S. Weather Bureau Rep. for February 1898, Vol. III, No. 2 (Wisconsin section), 8 pp.

    • Search Google Scholar
    • Export Citation
Save