|
Quantifying
Spatial and Temporal Dimensions of Premonitory
Animal Behavior of the July 16, 1990 Luzon Earthquake
Ricardo M. Zarco*, Jose Tiburcio S. Nicolas*, Mario J. Aguja*
Arturo S. Daag** and Anabelle F. Ringor**
*University of the Philippines
**Philippine Institute of Volcanology and Seismology
METHODOLOGY
1. Nature of the Study
This is a retrospective
field study of premonitory animal behavior which caught the attention of
human observers. A survey- interview of selected respondents is the
procedure of data gathering. The number of recalled events is assumed to
correlate with the actual number of events occurring in a given
location. Events are relatively easy to recall but the exact day or date
and time of the incidents are less easily recalled.
Animals may be our
indicators, they are our precision instruments that pick up chemical and
physical or of geologic stress as an earthquake draws near. Our human
observers selected from the peasantry, housewives, and workers are
assumed to report events objectively, and the role of the behavioral
scientist is to gather and analyze the data for its pragmatic and
science values.
2. Choosing the human
observer
Adults are preferred.
Residents who devote much of their time around the household and
community are selected if they own or have several domestic animals like
dogs, cats, pigs, chickens and if they had stayed home several days
before and during the earthquake. Mental alertness and verbal fluency
are necessary. The principle behind this selection process is to
maximize objectivity, hence validity.
3. The Interview
The interview is structured.
It asks the respondent to recall animal behavior immediately before the
earthquake.
Initially, pretests showed
problems of vagueness since the task of recall was too generalized.
Respondents kept recounting events during and after the quake and
excluded non domestic animals such as snakes, rats, crabs, earthworms,
cockroaches, ants and the like. The interview schedule had to be edited
twice to include a list of specific set of domestic and non domesticates
which the subject is made to specifically respond to. This new procedure
improved data gathering and scoring.
The enthusiasm and
willingness of the observers to recount the earthquake episode is noted.
Not a single refusal was encountered among the 218 respondents chosen.
4. The Research Design
The epicenter of the
earthquake was at 15°42'N and 121°7'E near the town of Rizal (PHIVOLCS
data). Towns surveyed are shown in Figure 1.
The ground rupture line
across Nueva Ecija was surveyed. This task covered six villages in all.
The intention of trying out our measure of intensity of precursory
animal behavior was conducted on those villages to see how the values
would compare against other areas like places away from the ground
rupture line. Previous similar studies in the People's Republic of China
(PROC) had shown that observations of unusual animal behavior follow
ground rupture lines and fracture areas. If this is true for this recent
earthquake then logically these six villages surveyed should reflect
higher intensity coefficients than areas where the earthquake had
lessened.
The second spatial exercise
attempted to test for the effect of epicentrality on animal behavior.
Towns and cities were selected at 20 kilometer intervals from the
epicenter and surveyed. A total of seven (7) towns and cities were
chosen for this exercise. These communities were purposely chosen away
from the affected fault and toward the SW and the southern direction
where the earthquake's effect had lessened. The distant communities from
the fault and the epicenter served as our comparator or control group.
The design, therefore,
involves the selection and classification of areas according to their
location in relation to the epicenter and the affected faultline. The
primary objective of the selection procedure is to identify the possible
effect and interaction of epicentral and faultline factors on the
ability of animals to perceive premonitory signals prior to the July 16
earthquake. If animals perceive tension signals as a result of the
pressure exerted below the faultline, then, one may expect that reports
of anomalous animal behavior would not vary significantly in areas
located along the affected faultline. However, if such geologic signals
are mainly a result of the tension produced below the epicenter one
could expect a gradual decrease in the reportage of anomalous animal
behavior as the distance from the epicenter is increased. On the other
hand, if these geologic stress signals are a product of the interaction
of both the tension produced near the epicenter and along the affected
faultline, a nonlinear trend in the reportage of anomalous animal
behavior may be expected. To test these possibilities, areas were
classified into four (4) major groups - those located along the fault
and nearthe epicenter, those located along the faultbut moderately
distant from the epicenter, those located outside the faultline and
moderately distant from the epicenter, and those which are comparatively
both distant to the epicenter and the affected rupture line (see Table
lb).
5. The Concept of the
Coefficients of Intensity
To quantify premonitory
animal behavior, an area, or location is chosen. This is the unit of
analysis, e.g. a village on a ground rupture line, or a town nearest the
epicenter, or a distant city slightly affected by the earthquake.
Residents, our human observers are selected for interview. They were
asked to recall their observation as these occurred a week ( 7 days)
before the earthquake.

Operationalized, the CI is
location specific, covers a 7 day period, it is the mean value of
observations per observer. This statistic can be accompanied by measures
of variance, standard error, measures of peakedness and symmetry.
>> Research Findings
Send comments or suggestions to Webmaster
Copyright© Philippine Institute of Volcanology and Seismology 2001
All rights reserved. |