29 March 2005

Sumatra Earthquake: Quick look at divergent Tsunami forecasts

-->
It's interesting to contrast the Indian and US Geological forecasts. Yet, regardless the epicenter, the models forecasted a sizeable Tsunami, worrying some.

Yet, it was when the Indian government looked at the ridge lines that we came up with some different results.

This hopes to support a study looked at whether the divergence is related to modeling errors, or a fact of life related to the uncertainty of errors.

Initial reports

Models predicted a high Tsunami. These were later downgraded. Appears as though the initial reports were downgraded because of the location. After carefully reviewing the ridge line, India forecasted that there would be no likely Tsunami threat.

US Geological survey stated the epicenter was approximately +/- 2.9 NM, approximagely 0.05 degress at the equator. However, we found that regardless the actual location, models still forecasted a high Tsunami.

Actual Tsunami was 3M high, suggesting the affected areas were not well connected to the Internet to report actual wave heights.

It appears as though the Earthquake occurred right on a ridge line, causing confusion about the location of the Tsunami:



Warning mechanisms

Most warnings were non-electronic. Most physically felt the Earthquake. Those working on their computer or watching television did not have flash message. Some SMS alerts were more succesfful. Monestaries and mosques issued alerts via speaker.

Recommendations

Use divergent data from non-US sources for cross check.

Use real-time event simulation to target notifications.

Update subscription optionsn to allow users to pick their location, and the XML-platforms will automatically report based on location, not magnitude.

Review audio-warnings via home-phone, even when power is out and phone is still on the hook.

It's interesting to contrast the Indian and US Geological forecasts. Yet, regardless the epicenter, the models forecasted a sizeable Tsunami, worrying some.

Yet, it was when the Indian government looked at the ridge lines that we came up with some different results.

This hopes to support a study looked at whether the divergence is related to modeling errors, or a fact of life related to the uncertainty of errors.

Initial reports

Models predicted a high Tsunami. These were later downgraded. Appears as though the initial reports were downgraded because of the location. After carefully reviewing the ridge line, India forecasted that there would be no likely Tsunami threat.

US Geological survey stated the epicenter was approximately +/- 2.9 NM, approximagely 0.05 degress at the equator. However, we found that regardless the actual location, models still forecasted a high Tsunami.

Actual Tsunami was 3M high, suggesting the affected areas were not well connected to the Internet to report actual wave heights.

It appears as though the Earthquake occurred right on a ridge line, causing confusion about the location of the Tsunami:



Warning mechanisms

Most warnings were non-electronic. Most physically felt the Earthquake. Those working on their computer or watching television did not have flash message. Some SMS alerts were more succesfful. Monestaries and mosques issued alerts via speaker.

Recommendations

Use divergent data from non-US sources for cross check.

Use real-time event simulation to target notifications.

Update subscription optionsn to allow users to pick their location, and the XML-platforms will automatically report based on location, not magnitude.

Review audio-warnings via home-phone, even when power is out and phone is still on the hook.

" />