Transcript Slide 1

T21C-1981 Estimating Earth's Heat Flux
Will Gosnold, Department of Geology and Geological Engineering
University of North Dakota
[email protected]
(1) The fundamental difference between recent estimates of Earth’s surface heat flux is whether heat
conduction models of spreading ridges or averages of oceanic heat flow observations accurately represent
oceanic heat flow. Models that relate heat flow to age of oceanic lithosphere (Stein and Stein, 1992) yield
an estimate of 44 TW (Pollack, Hurter, and Johnson, 1993), but averaged observations of oceanic heat flow
yield estimates of 31 TW (Hofmeister and Criss; 2005) and 29 to 34 TW (Hamza, Cardoso, and Ponte Neto,
2008).
HC question the applicability of 1-D cooling models to a 3-D lithosphere, and both HC and HCPN question
effect of hydrothermal circulation and the agreement between the models and the data.
I address the question of hydrothermal circulation by discussion of observed heat flow and the implications
of conductive heat flow above a magma chamber beneath the ridge crest in [2].
[2] Hydrothermal Circulation
In a conductive thermal regime, heat flow is predictable.
Diagram 2a shows temperature and heat flow curves for conductive
continental and oceanic thermal regimes. Continental heat flow
decreases with depth as radiogenic heat decreases, but with
radiogenic heat two orders of magnitude less than that of the
continents, oceanic heat flow remains nearly constant. The change
in slope of the oceanic T-z curve is due to the change in thermal
conductivity at the base of the crust.
Extreme scatter in heat flow in the ocean basin indicates
extreme hydrothermal circulation. Diagram 2b shows 4457 heat
flow observations vs. age of oceanic crust. Four-hundred of the
observed heat flows lie between 0.4 and 27 mW m-2, and 1500 are
less than 50 mW m-2. This is not possible without hydrothermal
input. Diagram 2c shows the temperature profile and heat flow that
would result from a magma chamber 2 km below the ridge crest
with a half-spreading rate of 2.5 cm y-1. It is obvious that heat flow
less than 1000 mW m-2 would require a hydrothermal sink.
Oceanic Temp.
Continental HF
Oceanic HF
Continental Temp.
2a
I address the question of agreement between models and averaged values of oceanic heat flow data by a
simple comparison in [3].
2c
I address the question of heat flow vs. age and cooling models using a 2-D finite difference conductive
heat flow model of lithosphere temperature from the ridge crest to old sea floor (t>160 ma) in [4].
2b
0
1400
-500
Averaged heat flow for 4457 heat flow sites
agrees with all models at ages > 55 ma.
Separation of model predictions from
observations is due to hydrothermal
circulation in the oceanic crust. To
emphasize this point, I quote Bullard’s Law.
"Never take a second heat flow
measurement within 20 km of the original
for fear that it differ from the first by two
orders of magnitude."
MORB
Intraplate
Geotherm
Solidus
Hawaiian Ol.
Tholeiite
ALK OL basalt
Ol Basanite
Ol Nephelinite
Ol Melilitite
Ol Leucite
Mantle Temperatures after Fig. 2,
D.H. Green (2008) Primary
magmas at mid-ocean ridges, ‘hotspots’ and other intraplate settings:
constraints on mantle potential
temperatures, www.
MantlePlumes.org
Ridge Geotherm
Mid-Ocean
Ridge Picrites
-1000
q = 510 t
1000
-.5
Depth (m)
[3] Average Observations and
Models
1200
q = 480 t -.5
q = 473 t -.5
-1500
800
-2000
600
-2500
-3000
400
-3500
200
-4000
0
0
1000
2000
3000
4000
5000
Meters from Ridge Crest
[4a] The heat flow vs. age plot compares heat flow
calculated from the 2-D model (red line) with
GHD1(green line), and averaged observations. The
temperature vs. depth and age shows the thermal
structure of the lithosphere given by the 2-D model.
[4] 2-D Model of Heat Conduction and Advection
The model parameters include thermal conductivity variation with
temperature, a fixed T-z profile at the ridge, constant spreading rate,
and constant heat flow into the base of the lithosphere. The output of
the model is a 2-D temperature vs. depth map based on heat
conduction and mass transport of heat in the spreading lithosphere. I
tested the reliability of the computations using different half-spreading
rates and different node spacings and verified that the models yield
equivalent results at equivalent times and depths. The model result
agrees closely with the GDH1, HSC, and PSM models for ages less
than 5 ma and shows slightly higher heat flow for ages greater than 5
ma. Assuming that continental heat flows are not in question, the
conclusion of this research is that the models of heat flow vs. age
are valid and that global heat flux is approximately 44 TW.
Deg C
0
20
40
60
80
km
100
120
Modern adiabatic upwelling
140
160