TESTING CAUSE AND EFFECT HYPOTHESES
A. INTRODUCTION
The last
lecture introduced correlation as a means for inferring
cause and effect relationships. The purpose
of today's lecture
is to show how correlation can also be used as a
means to test
cause and effect hypotheses.
B. BACKGROUND
The
geologic example used to illustrate how correlation can be
used as a means to test cause and effect
hypotheses is known
as
land
subsidence, the
slow, downward movement of the land
surface. There are
various causes of this phenomenon,
but we
are considering only one specific type (called
deep subsidence
by the authors of your text), which can occur
when pressurized
fluids (water, gas, and/or oil) are removed from
the subsurface:
1. CONFINED GROUNDWATER
Sometimes
the geologic conditions allow groundwater to occur
within highly
permeable materials that are sandwiched between
by lowly permeable
materials. This is called
confined (artesian)
groundwater.
Notice that confined groundwater occurs under
pressure,
which can result in flowing
wells.
2.
OIL AND GAS RESERVOIRS
Oil and natural gas are fluids produced deep within the earth as
buried organic
matter is subjected to heat and pressure. Oil and
gas rise towards
the land surface until they are
trapped by lowly
permeable
formations. These fluids occur under pressure and
are thus analogous
to confined groundwater.
C. OBSERVATIONS
1. WHERE SUBSIDENCE OCCURS
Deep subsidence occurs in
various locations across the
United
States.
What do these locations have in common?
a.
These are locations where fluids are pumped
from layers of
sandy sediment within the subsurface.
b. The
fluids are under pressure: both
confined groundwater
and pressurized oil and gas
reservoirs.
2. EXAMPLES
a. Central
Valley, California:
This is the world's largest area of subsidence
and is an area
where groundwater is pumped from confined
aquifers.
b. Long Beach, California:
This is an area where as much as 30 feet of
subsidence has
occurred over
an oil and gas field.
D. HYPOTHESIS
When fluids under pressure
within the subsurface are removed, the
fluid pressure
drops. If the formation from which fluids are taken is
composed of
compressible sediments, a reduction in fluid pressure
can cause the
overlying
formations to slowly subside.
E.
HYPOTHESIS TESTING
1. CORRELATION
To test
the hypothesis that removal of pressurized fluids causes
land
subsidence, pumping and subsidence data can be plotted
versus
time.
Example:
Santa Clara Valley, California
2. PREDICTION
This hypothesis can also be tested by applying the model in an
effort
to solve subsidence problems.
Example: subsidence in the Long Beach, California area has
been arrested by using injection wells to
pump sea
water back into the oil and gas reservoir.