About Severe Storms Archive

Introduction

The Severe Storms Archive contains data relating to recorded Severe Thunderstorm and related events dating back to the 18th Century.
This data is now directly accessible to the public at http://www.bom.gov.au/australia/stormarchive/

But before using the data you should note the following limitations:

Completeness of data

This is a database of all recorded severe local storm events. Many stormsare not recorded because:

  • It impacted an unpopulated area
  • Nobody saw it
  • Nobody reported it
  • It was not detected by any observation systems (e.g. radar, AWS)

This will cause two types of bias to be imposed on the database.

(1) There is an artificial increase in observed events over time.

This is due to the increase in population and spread of population. Also, there has been an increase in the number and range of observation systems, particularly radar. The recording of severe thunderstorm data was also enhanced in the 1980's and 1990's as the Bureau introduced dedicated Severe Weather Sections in its main Regional Offices, and the Bureau's Storm Spotters program was developed.

(2) The observed distribution of severe thunderstorms across the country is influenced by population density

Where there are no people, there is nobody to see the event, hence there will be an artificial maximum in storm frequency observed in populated areas.

Quality of data

Due to the relative scarcity of robust observation systems across the country, as well as the relatively small impact area of an indivdual severe thunderstorm, it is necessary to utilise information that can be gleaned from all possible sources. This often requires the use of subjective information gathered from the public impacted by the event (e.g. "the hail was the size of cricket balls!" ). The quality of the data can be assessed by quality metadata or plain-language comments recorded in the databse with the report.

Timeliness of data

Considerable time may pass between the severe thunderstorm event and the insertion of data relating to the event into the database. This delay may be caused by any number of the following reasons:

  • Time required to assess and gather impact data from the site
  • Time required to locate and interview those who saw the event
  • Delay in the receipt of reports from people observing the event
  • Time required to verify reports received
  • Availability of Bureau staff to insert new data into the databse

Out of Scope

This database does not include the following:

  • Most phenomena directly related to tropical cyclones
  • Severe winds at sea (except waterspouts near the shore)
  • Fire weather events
  • Riverine flooding
  • Non-severe events
  • Extreme temperature events
  • Earthquakes, tsunamis & landslides

Guidance on basic database structure

The database is structured on severe phenomena occuring during severe thunderstorm events.

  • EVENT: This is the development, lifetime and decay of an individual thunderstorm or squall line.
  • SEVERE PHENOMENON: During the lifetime of a storm event, there may be one or more occurrences of severe weather conditions (large hail, damaging wind gusts, etc). These may affect different locations along the storm path. Each of these is a "severe phenomenon" and can be categorised as one of the following:
    • Severe Wind Gust
    • Damaging Hail
    • Tornado
    • Heavy Rain
    • Lightning (causing injury or death)
    • Waterspout
    • Damaging dust-devil

Any thunderstorm event may cause one or many occurrences of severe phenomena. These may occur at different times during an event, or some may occur at the same time and place.

The following diagram illustrates the main database structure. The diagram does not show all the detail (like hail size, source of data, duration of hail, etc) for each datbase entry.

severe thunderstorm database main structure

Note: the data depicted in the diagram is not real