HR: 1330h
AN: S12A-0373 [PDF]
TI: Intensity Distribution and Isoseismal Maps for the Denali, Alaska, Earthquake of 2002 November
3
AU: Dewey, J W
EM: dewey@usgs.gov
AF: U.S. Geological Survey, MS 966 Box 25046
Denver Federal Center, Denver, CO 80225 United States
AU: Wald, D J
EM: wald@usgs.gov
AF: U.S. Geological Survey, MS 966 Box 25046
Denver Federal Center, Denver, CO 80225 United States
AB:
Macroseismic data for the 2002 Nov 3 Denali, Alaska, earthquake were were obtained from three sources: (1) the Did You Feel
It? Internet intensity questionnaire, which yielded automatically computed Community Internet Intensities (CII) from over
3400 respondents in 182 zip codes; (2) a postal canvass yielding traditional Modified Mercalli intensities (MMI) assigned by
an analyst to returned canvass cards for 86 zip codes; (3) additional data from the Geological Survey of Canada, the West
Coast and Alaska Tsunami Warning Center, press and engineering reports. The purposes of this study were to present a
carefully evaluated distribution of intensities for the Denali earthquake and to compare intensities derived by methods (1)
and (2) above in order to calibrate the new Internet method against the results of the older postal canvass method. Thus
each of the two data sets ((1) plus (3); (2) plus (3)) was evaluated independently of the other and the results compared
before a final isoseismal map, based on all three, was created. To facilitate investigating these types of intensity-related
phenomena, we have developed a relational database for displaying all the data for each zip code, searching for anomalous
responses, locating sites, finding and marking all reports of specific effects such as seiching, reading and searching
observers' comments, assigning MMIs and comparing CII, RCII (reviewed CII), MMI, etc. With the aid of this database, RCII
were found for each community by including consideration of observers' written comments plus data from sources listed in (3).
One problem so far unique to the Denali earthquake arose from observers on seiche-rocked boats who often responded as if
experiencing the earthquake on land, resulting in high CIIs requiring revision. Other problems resulted from Alaska's low
population density: huge zip code areas that had to be subdivided and too few towns to clearly define the near-fault
intensities and the periphery of the felt area. Although both CII and MMI intensities are based on intensity questionnaires,
the CII and MMI data are derived quite differently. A traditional MMI is often assigned to a single observation from a
community based on a detailed questionnaire. The CII for a zip code is automatically calculated based on a statistical
formula using data from multiple observers in that zip code responding via a short Internet questionnaire. This new CII
method has many obvious advantages including having an initial map quickly available for disaster responders, obtaining far
more data than was previously possible, obtaining better intensities from places with multiple respondents than from a single
observer, and automating some tedious tasks. The three resulting Denali isoseismal maps are in very good agreement within
the damage-level intensities (MMI 6 and above). The outer isoseismals, mostly extending across regions of low population,
varied depending on locations of available information; most notably, an outlying low-intensity felt area in Washington
State appears on the final and RCII-only maps but not on the MMI-only map because it lies beyond the limits of the postal
canvass. This study demonstrates the need for analyst review of the CII for important earthquakes and the advantage of the
new CII method over the old postal canvass in acquiring data from sparsely inhabited places and in locating unanticipated
felt regions.
DE: 7299 General or miscellaneous
SC: Seismology [S]
MN: 2003 Fall Meeting