HR: 14:40h
AN: T53C-05 [Abstracts]
TI: Geosynchronous Weather Satellite Thermal IR Measurements Prior to Earthquakes
AU: * Logan, T L
EM: tll@mipl.jpl.nasa.gov
AF: NASA / JPL, 4800 Oak Grove Dr., Pasadena, CA 91109
United States
AU: Bryant, N A
EM: nab@mipl.jpl.nasa.gov
AF: NASA / JPL, 4800 Oak Grove Dr., Pasadena, CA 91109
United States
AU: Zobrist, A L
EM: zobrista@yahoo.com
AF: NASA / JPL, 4800 Oak Grove Dr., Pasadena, CA 91109
United States
AU: Bunch, W L
EM: Walter.Bunch@jpl.nasa.gov
AF: NASA / JPL, 4800 Oak Grove Dr., Pasadena, CA 91109
United States
AU: Clark, J
EM: jxc@mipl.jpl.nasa.gov
AF: NASA / JPL, 4800 Oak Grove Dr., Pasadena, CA 91109
United States
AB:
Thermal anomalies prior to earthquakes have been a controversial topic for some time. Previous investigations of three
earthquake events by the authors showed thermal changes, and as a result a more systematic analysis of fourteen events,
representing a range of magnitudes and epicenter depths was undertaken. Earthquakes associated with plate movement
(strike-slip and thrust faulting), rather than volcanism, were to be considered. It was the purpose of this study to
determine if thermal anomalies could be found in association with known earthquakes by systematically co-registering
geosynchronous weather satellite images at the sub-pixel level and then determining if statistically significant responses
had occurred prior to an event. The automatic co-registration procedures used for this task to accommodate all properties
particular to weather satellite observations taken at night. Spacecraft and sensor ephemeris and the horizontal displacement
due to elevation were all factored in, and final adjustment for minor satellite deviations (related to roll, pitch, and yaw)
were made by using image-to-image tiepoint correlations. Reliance upon visual clues in an image (frequently the subject of
debate in the past) is not required. The technique relies on the general condition where ground cools after sunset. The
technique applies best to the use of the geosynchronous weather satellites (GOES,and Meteosat), where images are taken every
thirty minutes. Use of the geosynchronous satellites also reduces the potential for miscalculation of trends due to weather
front movement or local cloud/fog formation. The data analyzed for each earthquake includes 10 days prior to each event and 5
days after the event as well as the day of the event. The data are for every half hour from sunset to dawn, thermal IR
bands. We also obtained the same Julian dates of the data for the three previous years to use as a baseline. The data sets
were used to systematically measure the observance of thermal anomalies in two key contexts. First is the degree any thermal
anomaly just prior to an event can be demonstrated to deviate from a baseline thermal profile derived from the previous
three years and recorded weather history. Second is the possibility to compare the earthquake epicenter with nearby regions
having similar or different geology using the same imagery but no history of seismic activity. This experimental design will
help address questions regarding the "uniqueness" of thermal anomalies observed by us and others previously, and help
characterize the observed thermal responses under a variety of magnitude, depth, and geologic conditions.
DE: 8194 Instruments and techniques
DE: 1694 Instruments and techniques
DE: 1640 Remote sensing
SC: Tectonophysics [T]
MN: 2004 AGU Fall Meeting