ATMOSPHERIC ELECTRIC FIELD OBSERVATIONS, ANIMAL BEHAVIOR, AND EARTHQUAKES

by

C. BufeI'. 5. Geological Survey

iiun~!o Park, California

and

J. NaneviczStanford Research Institute

Menlo Park, California

Open-file report

76-876

This report is preliminary and has not been edited or revievMed for conformity with Geologcazl Survey standards

95.

Observations

Late in. 1975, a modest field program was initiated to observe atmospheric

electric field along the San Andreas and Calaveras faults in central California.

An objective of the program was to detect changes in atmospheric electric field

associated with earthquakes and to compare these changes with those observed

during storms. Unusual animal behavior reported preceding earthquakes resembles,

in many ways, that observed before r-tonn activity. Typical weather - related

field-changes observed elsewhere are shown in Figure 1, Maximum changes are of

the order of -[- 1C,000 V/m.

The fault-zone program was carried out by Joel Sharp of the USGS using

field mills (instruments that measure atmospheric electrostatic field) provided

by the Stanford Research Institute. Observations were begun in the winter of

1976 and have continued to the present time. The data are recorded on Rustrak

charts at a speed of 1/4" per hour. Full scale range is about 2500 V/m. The

field mill locations and periods of operation are tabulated below (Table 1).

These periods contain some minor data gaps occasioned by bearing failure of

the field mill. ...

Preliminary analysis of the data has shown the following characteristics:

1. Normal "quiet" field is positive (ground negative) and ranges from

0 to 300 V/m.

2. Rainstorm activity is accompanied by large field fluctuations

(> 1250 volts + and -) and reversals, with several reversals often

occurring in the course of an hour or two (see Figure 2).

3. A site-dependent diurnal variation ranges from about 200 V/rn at

St. Francis Retreat to more than 1000 V/m at San Felipe Valley.

The diurnal variation is accompanied by high frequency, spikey

noise during the daylight hours.

96

TABLE 1

FIELD MILL SITES IN CENTRAL CALIFORNIA

Sj_te_Name Code

Stone Canyon STC

Melendy Ranch MEL

St. Francis Retreat SFR

Cienega -Valley ADS

San Felipe Valley SFV

Stanford Obs. Primate SOP Lab

. N.

36-38.10

36-35.7

36-49.2

36-43.2

37-18.0

37-24.5

Long. ^

121-14.00

121-11.2

121-29.9

121-21.0

121-37.5

122-13 0

Period of Operation

12-17-75 to 4-7-76

1-21-76 to 8-17-76

2-5-76 to present

4-7-76 to 6-23-76 8-17-76 to 11-5-76

6-25-76 to present

11-9-76 to present

97

Figure 1. Weather-Related Atmospheric Electric Field Fluctuations

-15

1200 11-20

to ;~

oj>

10 j~

'-i

_^^X±^aj

77-:.;-, ?

LQC/L Mil AN Tl^ 1^

'. /\ I ! V\ ; \ \ 1.

I '/

t

, ;

98

F

'?ure

2*

Atmo

sphe

ric

E-lectHc Fi

eld

Vari

atio

ns During Ra

inst

orm

Acti

vity

, Al

ong

the

San

Andreas

Faul

t.

.._, _ . __ .

_. ..._

_......

.rrrL_

I^rr".

.....

. ._._

_._. .

.....

_ _

: :

:^-;'

:. :.- .:,

" .

.'"

;"""

; - T:

' "' -

. #

* "

'

- ""

( ;; $

!!.' i

'-' ;:!

\>* '-

-\^

. .-..

-. : ;

: -::

->.: ! i

:' .

i .

"

-.; i-.

-:-"

; . :

.;: -

H:-

; ;r

H-"-

:- '

"~ ft" .

':7;

"!".

.:t"

'~"T

' :;

---n

-vj-

-;-;

- /

--,

- ;

; .-

-r's

V;?

;£.

(

- :'

] V;" '|.

TV

.

__.

...

_ ..r

._.j

......_....

. _.

. ,_

., :

.

.-

-"~

'""

--

. .

. <

'

. _.....

...._._.

.,...v

....r.

i

' .

' .

-J ,

.-. ,:

U V

M&f-

*f&'i

-te.

x: \

\.:

-:.-

liV

.F'

: ! Jl

.r:

.'.

t '

!'^^^^^^^A .

; ;

-, - .

- .

. .

.. ;-

>-:

" '

'

I .-..?} ."

' -

- :;(

-. --lr-t)

-;.'

4. Several small step-like fitId decreases at St. Francis Retreat

appear to be correlated with abrupt increases in humidity as

measured at a nearby weather station. The source of the abrupt

change in humidity is not known.

Field Changes and Animal Behavior

Reports of unusual animal behavior preceding earthquakes are sometimes

accompanied by reports of mysterious luminosity or lightning (Derr, 1973).

Tributsch (1976) has suggested that the piezoelectric effect may be responsible

for unusual animal behavior preceding the May 6, 1976 earthquake (M = 6.7) in north­

eastern Italy. Finkelstein," et.al. (1973) have concluded that, the piezoelectric

effect is likely to generate lightining only in circumstances where the rock

resistivity is unusually high. A more plausible explanation may be electro-

kinetic phenomena induced by ground water flow as discussed by Mizutani, et.al. s

(1976).

Unusual animal behavior preceding earthquakes may not necessarily by the

result of electrostatic field changes, but rather in response to related

ionization phenomena and the release of gases associated with creack opening. If

gas emission is a factor, the development of sense of smell in many animals is

sufficiently advanced over that of humans as to offer an explanation of the

apparent superiority of animals in anticipating earthquakes. It is also possible

that the human s with constant exposure to pollution and distracting thought processes

and activities, is less likely to notice subtle changes in his environment.

Human anxiety or depression are nearly always attributed to psychological stress;

physic?! illness to organic disease. K.rueger and Reed (1976) have studied the

biological impact of small air ion imbalance on mice, rats, and humans. They

conclude that the production of serotonin, a powerful neuro hormones is enhanced

by high concentrations of positive ions. The "serotonin irritation syndrome"

(migraine, nausea, vomiting, amblyopia, irritability, hyperistalsis, edema,

loo

conjunctivitis, congestion of th? respiratory tract, etc.) is associated with an'

increase in positive ions preceding, by 24 to 48 hours, the onset of hot, dry wind

conditions. This human condition can be relieved by inhalation of air containing

large numbers of negative small air ions. Negative ions have also been found to

have a tranquilizing effect on rats and mice.

Air ionization can result from the decay of radioactive material. Pierce

(1976) has calculated atmospheric electric field changes which would result from

variations in radon release. The changes amount to about 30% of the fair weather

field and thus would be on the order of 100 V/m, a value probably too low to

alone acount for unusual animal behavior, since field changes of this, size are common,

Field Changes And Earthquakes

The only observations of atmospheric electric field during an earthquake

sequence of which we are aware were those of Kondo (1968) during the 1966

Matsushiro sequence. He found that significantly more earthquakes occurred at

times of decreased electric field than at other times.

During the period of our observations, earthquake activity along the San

Andreas and Calaveras faults has been unusually light. Locations of the largest

earthquakes to occur near a field mill are shown in Figure 3. One shallow

(h - 2.5 km., M = 3.5) earthquake which occurred very near the ADS mill was

preceded by a field reversal (Figure 4) at ADS about 24 hours earlier and field

reversals just prior to the earthquake at SFR. The ADS signal change is sus­

pected to result from local cultural activity, and an uncertainty in timing

makes it impossible to ascertain whether the changes at SFR actually accompanied

the earthquake.

Two other earthquakes (M = 3.5, M = 3.8) occurred within a 7-hour period

in late October near the SFR mill. These events occurred on two different

faults at depths of about 8 km. The electrostatic field at SFR and ADS during

101

rrfc 10

20

15

25

20 STAlUTC

KILOMET^fiS

CONTOUR IN'TFTVAL 200 r-LCT- *JDOTTED UK:.-. F«i.'RCSLNT 100-fOOT CONTOURS

L-'.ToM IS MF.AN SEA LEVEL DEPTH CURVES IN FEET DATUM IS MEAN LOWER LOW WATERSHORtLINt SHOWN WE c«ESENrS IM£ APPROXIMATE LINE OF MEAN HIGH WATER

P iX *30'_____~*^\i GbtZY/" 77- jpsres^y/ . j -^ . .__,^^7771-7 r -\ s \v^ i "iiv-'"'^ >! " *^^ '/ £^ \ i /:s"»*.* « '' >

--,- c/pro \ \ 'V ' \ -V "" v ( -ft- \'J> ^-^; ;-. rA.-- ,i ^ t - ^-..' i ,s -.,-J rj ,,;^' ;. c ,,, -> v1 '--x^-- «\'- >^»^: y^'v. - ( V 0- -. ' \ " M -~- !> - '-.- '/-n 'v ^- iC 'vi <. $'- . >' ^>i;>Ly^y ^XVT^-t-^^v \ - v- \^.^r^-' -\ \ >v^;^v / 'M^*-- '/..^ SVrN^v- V ^\carn.e^p4r^^--^\-- ; v^.'.p.'-;'nl-\xs\ /> ;--^ .- ^/^Tcv'^vrv 1^-K^f 3 Oo^/" v-co X. x\i--- '/v-^n FelipJ'--;( -:.- 1 l-- «*«* ~ ' '* v -^-^^^-^M'-^U^;'V

T * " *-v Q- *J ^ vx ""^^"s ***" V » \ *. x^ ^^» !%-' i- *' \ * 7x,«w t -, ^c ^-*s i"TA /*M ro/v /*^^*v*rN A^/R/POSA -^ I \", < -^ Nj *-s\ -^-/ »v \C\ >^S^ ~\ V ^s.; * ^1 ^S. t-r I » ' I ^^^*''^»-»*JI/H V,-L/>r\M C-wot-? - - _ f _ J ^

^ Mill Locations Along the San Andreas Fault and Epicenters of j^-SA-^

tflquakes (M 3.5) During Period of Operation. The epicenters .<-.!

/L of the San Andreas Fault are believed to be mis located and

occurred nearer the mapped fault trace. i^-» --'

-1000 V

/m

1, , V- t

.

i^!V:>

'

SO

OO

V

/m

SE

PT

2

O

u>

A AD

B-M

3.0

M 3

.5

,(

;' '>

'"-*»

>V

'

0000

GM

Tt

f '

'>C

CS

EP

T

3

^jL..j

.-iLy

- v

..

i

" i

' 7

~"

^'

-"-1

1 - ip

m

, »

,

m

,-y /

"N

^^

;

A

-**"

*- -»

v^

^

- ^

H./*V

^^W

V-

^*

>?'

* \

» i

. v

: i

v, -

-

' r

C^*

^ P*

^"**

t^^*

"^^

*%*'»

ri*^

fr

-«*f»

f&

tMK

iy'r

iL--

' '

- v'

'^

'>

->\;

V 0

'>

V

C

>;^

"";r

7^

;r-<f

<L/

~

i 0

C

C G

^ ^

>n

A

***\

, '

- *

' ""

:"~*

"""~

t=rx

. ^y-

>**v

v-*<

t

.j**s

.; '

? ,,

. Y

.

v< ^

-

- i;

(- -

V A

'' /

i> 4

\ /

Vfe<

r/ U

*r

'"«'

-,. *»

*-.

4

»\

'

(.",'* )

. X

/

A-**-

-

. -

^ -

-'

-J

>

/

£.

> _,

{, -

'v

-,

> ->

-,

Figure 4

. Atmospheric

Electric-Field V

aria

tion

s Near t

he T

ime

of a

Magnitude

3.5

Eart

hqua

ke N

ear

the

ADS

Fiel

d Mi

ll.

SFR

is about

18 k

m northv/est

.'&:>

''.

the

eart

hqua

ke.

10

the period preceding the earthquakes is shown in Figure 5. Data from a more

distant mill at SFV, 50 km north of SFR, is also shown. The storm-like

pattern of electric field changes occurring at SFR and ADS the day before

the earthquakes does not appear to correlate with any large scale change in

weather. However, similar field changes are not unusual enough to allow

unequivocal attributions of these changes to some process associated with the

earthquake. Additional near observations of larger earthquakes are needed to

establish such an association. As soon as any other system (tiltmeters, self-

potential, etc.) reports a developing anomaly, the USGS will place the field

mills in the expected epicentral region.

104

* "

v.,;,

is Re

trea

t (SFQ\

(SFR)

the

sam

p sp

ni

*u three

instruments

are

oper

atin

g at

£slp$

12

REFERENCES

Derr, 0. S., Earthquake lights: A review of observations and present theories,

Bull, Seism. Soc. Amer., 63, 2177, 1973

Finkelstein, D.-, R. D. Hill, and J. R. Powell, The piezoelectric theory of

earthquake lightning, J. Geophys. Res., 78, 992-993, 1973,

Kondo, G., The variation of the atmospheric field at the time of earthquake,

Mem. Kakioka Magn. Observ., 13, 11-23, 1968.

Krueger, A. P., and E. J. Reed, Biological impact of small air ions, Science,

193, 1209, 1976. See Appendix 3.

Mizutani, H. T. Ishido T. Yokokura, and S. Ohnishi, Electrokinetic phenomena

associated with earthquakes, Geophys. Res. Lett., 3, 365, 1976.

Pierce, E. T., Atmospheric electricity and earthquake prediction, Geophys.

Res. Lett., 3, 185, 1976

Tributsch, H., Pre-earthquake animal behavior: A consistent pattern indicates

unrecognized electrostatical geophysical phenomenon, unpublished manuscript,

1976.

106