Physics logo
Lancaster University Homepage
You are here:  SPP / STP Events Database / Search Examples


This page aims to demonstrate how to use the events database to find events which match certain criteria. If you wish to see the results of an example search press the "show me the completed form" button, and then the "submit" button.

Find all substorm onsets identfied in SAMNET data

For this simple search only one dataset should be selected - SAMNET substorm onsets.

It is easy to refine the search, for instance showing only substorm onsets in 1999, by completing the general search options at the top of the page, as demonstrated below. Notice that the start date is inclusive, while the end date is exclusive.

Find riometer absorption events at substorm onset

For this search it is necessary to select the riometer absorption events dataset for output, and correlate it against the times of substorm onsets. The substorm onset should precede any absorption, but as both datasets are only at one minute resolution it is possible that the absorption event was detected before the Pi2 oscillations triggered the SAMNET substorm detector.

This search crudely corresponds to a cause (substorm onset) / effect (riometer absorption) relationship.

Find sudden ionospheric disturbances in riometer data

There are several methods to to acomplish this, depending on the choice of datasets used. One possibility is to search only for SID events in the riometer data, but since these are manually identified (and verified) there are very few in the dataset (see the status page for details).

Instead, the methods chosen take advantage of the power of correlating between datasets. Either way, it is first necessary to select the riometer absorption events dataset for output. As SIDs are only seen during daylight hours the search is restricted by time of day. Although that does not take into account seasonal variations in daylight hours it does provide a simple method to help eliminate events which cannot be caused by solar X-rays. The time of day specified is deliberately skewed away from early evening to miss auroral activity.

Using the BATSE solar flare dataset

The first method demonstrated is to use the BATSE solar flare dataset, whose events are recorded at 1 second resolution. For this reason we don't allow the absorption to precede the start time of the flare. We allow up to 5 minutes from the start of the flare to the start of the absorption event.

Correlating against solar flare data finds the cause/effect relationship.

Using the AAVSO SID dataset

The second method is to use the American Association of Variable Star Observer's SID dataset. Note that in this dataset the times refer to time of maximum disturbance, so it is necessary to allow the absorption event to precede the AAVSO SID time. To prevent too many false results we have chosen the absorption events to start no more than 15 minutes before the time of maximum disturbance. You may wish to modify this value. There seems to be no merit in allowing riometer events to start after the time of maximum disturbance, so the end time of the correlation window is set to the time of maximum disturbance (+/-0d  00:00:00).

This second method illustrates a search for related effects of a common cause.

The fact that the two methods show different search results is not a suprise. The two correlating datasets are generated by very different means. The first used data from an expensive satellite, the second data from a network of radio amateurs using inexpensive (and often homebuilt) equipment. The difference comes not from the price tag of the equipment, but the fact that one is detecting the cause (solar X-rays), and the other an ionospheric effect of that cause. It is also worth noting that most of the AAVSO observers are located in America, which introduces a time of day bias in the distribution of the AAVSO SID events.

A Lancaster University approved page by webmaster.
© Lancaster University.