W1188 Progress Report Tyson E. Ochsner USDA-ARS St. Paul, MN Jan. 3, 2005.

W1188 Progress Report

Tyson E. Ochsner USDA-ARS St. Paul, MN

Jan. 3, 2005

The surface energy balance

LEHGRn 0

SGG r 0

Gr = heat flux at the reference depth zr

S = rate of change of heat storage in the soil above the reference depth

G0 = heat flux at the soil surface

Methods for measuring soil heat flux

1. Plate method Most common by far Thermopile measures vertical temperature difference

across a plate embedded in the soil Flux through the plate may not equal flux through the

soil

Methods for measuring soil heat flux

1. Plate method Most common by far Thermopile measures vertical temperature difference

across a plate embedded in the soil Flux through the plate may not equal flux through the

soil

2. Gradient methods Rarely used Soil temperature gradient and thermal conductivity

measured In situ thermal conductivity measurements are

challenging

Leads

Epoxy

Thermocouple

Resistanceheater SS tubing

6 mm

6 mm

1.3

40 mm

Methods for measuring soil heat flux

1. Plate method Most common by far Thermopile measures vertical temperature difference

across a plate embedded in the soil Flux through the plate may not equal flux through the

soil

2. Gradient methods Rarely used Soil temperature gradient and thermal conductivity

measured In situ thermal conductivity measurements are

challenging

3. Self-calibrating plate method Relatively new Heater on top of plate permits in situ correction for

heat flux divergence Large plate and unpublished theory

3 needle gradient Gr (W m-2)

-150 -75 0 75 150 225

1 ne

edle

gra

dien

t Gr (

W m

-2)

-150

-75

0

75

150

225

HF

T1.

1 pl

ate

Gr (

W m

-2)

-75

0

75

150

225Bare soil site

a

b

Slope 0.751r ² 0.908

Slope 1.01r ² 0.841

Flux estimates from HFT1.1 plates ~25% smaller than from three-needle gradient method.

Bare soil site – 2001

Two independent gradient methods gave similar flux estimates.

3 needle gradient Gr (W m-2)

-60 -40 -20 0 20 40 60 80

1 ne

edl

e g

radi

ent

Gr (

W m

-2)

-60

-40

-20

0

20

40

60

80

HF

T1

.1 p

late

Gr (

W m

-2)

-40

-20

0

20

40

60

80Corn site

a

b

Slope 0.755r ² 0.982

Slope 0.976r ² 0.988

Corn site – 2002

Flux estimates from HFT1.1 plates ~25% smaller than from three-needle gradient method.

Two independent gradient methods gave similar flux estimates.

3 needle gradient Gr (W m-2)

-40 -20 0 20 40 60

HF

P01

SC

pla

tes

Gr (

W m

-2)

-40

-20

0

20

40

HF

T1.

1 pl

ates

Gr (

W m

-2)

-40

-20

0

20

40

60Soybean site

a

b

Slope 0.755r ² 0.968

Slope 0.991r ² 0.968

Soybean site – 2004

Flux estimates from HFT1.1 plates ~25% smaller than from three-needle gradient method.

Self-calibrating plates and three-needle gradient method gave similar flux estimates.

Bare soil site – 2001

Flux estimates by the HFT1.1 plates were better than those from the other three types we tested.

Regression statistics for plate versus three needle gradient heat flux estimates.

Plate Slope Intercept r2 Mean absolute

difference

W m-2 W m-2

HFT1.1 0.751 2.28 0.908 14.2

CN3 0.640 1.82 0.913 17.4

GHT-1C 0.522 3.49 0.852 22.8

610 0.328 1.19 0.912 29.4

Conclusions

• Heat flux plates typically underestimate the magnitude of soil heat flux.

• The best of the four types of plates we tested underestimated the flux by about 25%.

• The three-needle gradient method and the self-calibrating plate method are promising alternatives.

• Further development of these methods are needed for frozen soil conditions.

Collaborators

• Thomas J. Sauer, USDA-ARS, National Soil Tilth Laboratory, Ames, IA

•Robert Horton, Dep. of Agronomy, Iowa State University, Ames, IA