Error estimations of dry deposition velocities of air pollutants using 4-m depth temperature under common assumptionsYung-Yao Lan(1), Ben-Jei Tsuang(1), Noel Keenlyside (2), and Sheng Fong Lin (3)

(1) Department of Environmental Engineering, National Chung Hsing University, Taichung 40227, Taiwan;(2) Leibniz-Institut fuer Meereswissenschaften Duesternbrooker Weg 20, D-24105 Kiel, Germany;(3) Green Energy & Environment Laboratories, Industrial Technology Research Institute, Hsinchu 31040 Taiwan

EGU2011-2830

Abstract It is well known that skin sea surface temperature (SSST) is different from bulk sea surface tempe

rature (BSST) by a few tenths of a degree Celsius. However, the extent of the error associated with dry deposition (or uptake) estimation by using BSST is not well known. This study tries to conduct such an evaluation using the on-board observation data over the South China Sea in the summers of 2004 and 2006. It was found that when a warm layer occurred, the deposition velocities using BSST were underestimated within the range of 0.8 – 4.3 %, and the absorbed sea surface heat flux was overestimated by 21 W m-2. In contrast, under cool-skin only conditions, the deposition velocities using BSST were overestimated within the range of 0.5 – 2.0 %, varying with pollutants and the absorbed sea surface heat flux was underestimated also by 21 W m-2. Scale analysis shows that for a slightly soluble gas (e.g., NO2, NO and CO), the error in the solubility estimation using BSST is the major source of the error in dry deposition estimation. For highly soluble gas (e.g., SO2), the error in the estimation of turbulent heat fluxes and, consequently, aerodynamic resistance and gas-phase film resistance using BSST is the major source of the total error. In contrast, for medium soluble gas (e.g., O3 and CO2) both the errors from the estimations of the solubility and aerodynamic resistance are important. In addition, deposition estimations by using various assumptions are discussed. The largest uncertainty is from the parameterizations for chemical enhancement factors, followed by various parameterizations for gas-transfer velocity, followed by neutral-atmosphere assumption, followed by using BSST as SST, and then followed by constant pH value assumption.

Keywords: South China Sea; deposition resistance; Henry constant; chemical enhancement factor; dry deposition; warm layer; cool skin

Introduction: Eddy Covariance System (ECS) was installed onboard during OR1-728, OR1-802, OR1-837, and

OR1-861 cruises. It was installed at 12.12 m above the sea surface. Sensible heat flux, latent heat flux and CO2 flux were measured. The ECS consists of 2 fast-response instruments: a 3-D ultrasonic anemometer (Young 81000 or CSAT3) and an open path infrared hygrometer/CO2 sensor (LICOR 7500). The anemometer measures three orthogonal wind components (u, v, w) and the speed of sound. The open path infrared hygrometer/CO2 sensor measures the concentrations of CO2 and water vapor. In addition, Gyro enhanced Orientation sensor (3DM-G) was located close to the 3-D anemometer, used to measure the ship movements (roll, pitch, yaw), to covert anemometer to the earth reference frame. And all the data were sampled via a datalogger (CR5000), and then used a transmission line to connect with a computer for storage. Gaseous deposition of air pollutants over the sea surface is related to aerodynamic resistance, friction velocity, solubility, and the chemical enhancement factor (Wanninkhof and Knox, 1996; Chang et al., 2004; Kuss and Schenider, 2004; Seinfeld and Pandis, 2006). From the dry deposition theory, two mechanisms are involved with the SST variable for estimating the deposition rate (Seinfeld and Pandis, 2006). First, the solubility of a gas pollutant is a function of SST (denoted as route 1). Second, turbulent heat fluxes, friction velocity and aerodynamic resistance are also a function of SST (denoted as route 2). This study will quantify the errors from the two routes in estimating dry deposition velocities by using BSST for various gases.

Theory One-column ocean model

A one-column ocean model (Tu and Tsuang, 2005) was used to simulate the SSST and deposit

ion-related parameters. The model was previously tested at a TOGA site. The sea water temperatur

e (T), current ( u ) and salinity (S) are determined (Martin, 1985; Gaspar et al., 1990) as:

z

F

c

R

z

Tk

t

T

ww

snhh

02

2

2

2ˆ

z

ukukf

t

umm

2

2

z

Sk

t

Shh

On the ocean surface, the friction velocity (u*) (m s-1), Monin-Obukhov length (L) (m); sensible heat flux (H) (W m-2) and latent heat flux (LE) (W m-2) are determined according to the similarity theory (Businger et al., 1971; Brutsaert, 1982) as:

L

z

L

dz

z

dz

kuu

mm00

0

0

*

ln v

aa

kgH

cuL

3*

a

asaa

r

TTcH

a

asva

r

qTqLLE

* 4slu TR

where k is von Karman’s constant (0.4); z is the measurement height (m); z0 is surface roughness (m); d0 is zero displacement (m) (=0 over ocean); is dimensionless momentum profile function; g is the acceleration of the gravity (m s-2); is the density (kg m-3) of air; ca is the specific heat of air at constant pressure (J kg-1 K-1). u and are wind speed (m s-1) and potential temperature (K) measured at height z; Ts is SSST (K); Ta is surface air temperature (K); q* is saturated specific humidity (kg kg-1); qa is surface air specific humidity (kg kg-1); and ra is aerodynamic resistance (s m-1).

m

Setting up an Eddy Covariance System (ECS) onboard

Figure 1 (left) Eddy covariance system installed in OR1 during the OR1-802 cruise at the Kaoshiung Harbor on 18 July 2006. The flux was measured at 12.12 m above the ocean surface. The two boxes in the lower mast were the control box of infrared hygrometer and Datealogger (CR5000); (right) Three-axis sonic anemometer (Young 81000) and infrared hygrometer (LICOR 7500) used to measure the fluxes.

Figure 2 Short-wave and long-wave radiation measurements in OR1 during

Figure 3 Datalogger (gray box ) and GPS (black circle box) installed on R/V OR1.

Deposition Theory:The transport property of each layer is different and the resistance for the deposition at each layer can be described as aerodynamic resistance (ra), gas-phase film resistance (rb), and aqueous-phase film resistance (rc) (Wesely, 1989; Duce et al., 1991; Seinfeld and Pandis, 2006). The total resistance (rT ) is the sum of all three resistances that can be determined (e.g., after Chang et al., 2004).

cbaTd rrrr

v

11

*

001

0

01ln

ku

L

z

L

dz

z

dz

rhh

a

3/2

*

5cab S

ur

qHkr

wc

*

1

Table 2 Molecular diffusivity coefficient in air (Dg), Henry constant (H), effective Henry constant (He), dimensionless

Henry constant (H*), dimensionless effective Henry constant (He*), chemical enhancement factor (a) and molecular gas-

transfer velocity under clam wind conditions (q) of various gaseous pollutants.

where Rsn is net solar radiation at the surface (W m-2); F(z) is the fraction (dimensionless) of Rsn that penetrates to the depth z; kh and km are eddy diffusion coefficients for heat and momentum (m2 s-1), respectively. Both the values of the eddy diffusion coefficient for heat within the cool skin and for momentum within the viscous layer are set at zero.

Chemical species Dg (m2 s-

1) at 298K

H (M(aq) atm-1-gas) at 298K

He (M(aq) atm-1-

gas) at 298K, pH 8.1

H* (M(aq) M-1-gas) at 298K

He* (M(aq) M-1-g

as) at 298K for pH 8.1

q

SO2 1.4×10-5 1.23 1.9×107 30.1 5.5×108 (Seinfeld and Pandis, 2006)

0

CO2 1.5×10-5 3.4×10-2 1.9 0.8 (a) 0.94(b) 0.94-1.3(c) 53(d) 1.06

(Seinfeld and Pandis, 2006) (d) 1.6×10-8e1.98 pH+1 (Kuss and Schneider, 2004)

0

O3 1.5×10-5 1.0×10-2 5.3×10-2 0.28 1.5 5.3 (Chang et al., 2004) 1.8×10-4 m s-1

NO2 1.5×10-5 1.0×10-2 1.0×10-2 0.24 0.24 1 (Seinfeld and Pandis, 2006) 0

NO 1.7×10-5 1.9×10-3 1.9×10-3 0.05 0.05 1 (This study) 0

CO 1.7×10-5 9.5×10-4 9.5×10-4 0.02 0.02 1 (This study) 0

SF6 1.3×10-5 2.4×10-4 2.4×10-4 0.006 0.006 1 (Liss and Merlivat, 1986; Wanninkhof, 1992; Wanninkhof et al., 1993; Nightingale et al., 2000)

0

** HH e

2211

][][1

H

KK

H

K sss

2211

][][1

H

KK

H

K ccc

(a) 1 (Wanninkhof 1992)(b) 1.0-1.4 (Wanninkhof and Knox, 1996)(c)

Results Table 1 summarizes the meteorological conditions during the occurrence of the warm layer (SSST > BSST) and the occurrence of the cool skin only (BSST>SSST). The cool skin always appeared, whilst the warm layer only occurred under the conditions of low wind speed (~4 m s-1) and high incoming solar radiation (~455 W m-2). They show that most (95%) of the time, BSST-SSST was positive, since the cool skin always appeared, while the warm layer only occurred under low wind and strong isolation conditions.

Table 1. Summary of the meteorological variables under warm layer (SSST > BSST) and cool skin (SSST < BSST) conditions.

Hours BSST-SSST (K)

Wind speed(m s-1)

Rs(W m-2)

Ta(K)

ΔLE (W m-2)

ΔH(W m-2)

ΔRlu(W m-2)

ΔG(W m-2)

Δu*(m s-1)

warm layer

40 (5%) -0.7 4.4 455 304.5 -14.1 -2.8 -4.3 21.2 -4.4×10-3

cool skin

714 (95%)

0.23 9.3 193 302.2 15.7 3.5 1.4 20.6 3.3×10-3

Total 754 0.19 9.0 207 302.4 14.1 3.2 1.1 18.4 2.9×10-3

Error using BSST Table 3 also lists the deposition velocity of that temperature parameter using BSST. It was found that when the warm layer occurred, the errors using BSST were underestimated within the range of 0.8 – 4.3 %, and when only cool skin condition occurred, the errors using BSST were overestimated within the range of 0.5 – 2.0 %, varying with pollutants. The warm layer would cause the bias of latent heat flux, sensible heat flux, and deposition velocities to fall down to the lowest point. Under this stable condition of higher pH and lower wind speed, the bias of latent heat flux and sensible heat flux were decreased by 60 and 10 W m-2, and the bias of deposition velocities was underestimated within the range of 5.4 - 37 %. Most of the time, only a cool skin occurred. An unstable condition of lower pH and higher wind speed have caused the bias of latent heat flux, sensible heat flux, and deposition velocity to increase a local peak. Table 3. Typical dry deposition velocity (vd) for some important atmospheric gases over ocean.

Species vd (m s-1) 

Profile This study (using SSST)

Using BSST

Others Avgbias*

(%)

Maxbias(%)

Error (%)

route 1(solubility)

route 2(ra, rb)

 

SO2 warm layercool skintotal

5.31×10-3

1.19×10-2

1.17×10-2

5.09×10-3

1.21×10-2

1.18×10-2

1×10-3-9×10-3 (Foltescu et al., 1996)1.5×10-2-2.1×10-2 (Raymond et al., 2004)

-4.31.51.1

-37.011.1

000

100100100 

CO2

(Wanninkhof, 1992)

warm layercool skintotal

2.68×10-5

8.08×10-4

7.80×10-5

2.66×10-5

8.13×10-4

7.83×10-5

5.6×10-5 (Wanninkhof, 2002) -1.80.50.5

-13.52.4

96.998.198.0

3.11.92.0 

CO2 (Kuss and Sch

neider, 2004)

warm layercool skintotal

3.71×10-5

1.11×10-4

8.94×10-5

3.66×10-5

1.12×10-4

8.99×10-5

-1.80.60.5

-13.52.4

95.997.997.8

4.12.12.2 

CO2 (Seinfeld and

Pandis, 2006)

warm layercool skintotal

1.20×10-3

3.26×10-3

3.15×10-3

1.17×10-33.31×10-3

3.20×10-3

-2.31.71.0

-17.73.8

68.273.673.4

31.826.426.6 

O3 warm layercool skintotal

2.10×10-4

2.99×10-4

2.93×10-4

2.08×10-4

3.01×10-4

2.97×10-4

1.6×10-4-7.8×10-4 (Chang et al., 2004)7.0×10-4 (Hauglustaine et al., 1994)7.5×10-4 (Muller, 1992)

-0.80.90.4

-5.43.1

72.989.588.1

27.110.511.9

NO2 warm layercool skintotal

9.21×10-6

2.64×10-5

2.50×10-5

9.04×10-6

2.69×10-5

2.53×10-5

2.0×10-4 (Hauglustaine et al., 1994) -1.92.01.0

-16.04.8

97.797.797.7

2.32.32.3 

NO warm layercool skintotal

1.51×10-6

4.36×10-6

4.14×10-6

1.49×10-6

4.44×10-6

4.28×10-6

3.0×10-5 (Hauglustaine et al., 1994) -1.51.70.8

-12.63.7

99.899.999.9

0.20.10.1 

CO warm layercool skintotal

7.49×10-7

2.17×10-6

2.09×10-6

7.39×10-7

2.20×10-6

2.13×10-6

0 (Hauglustaine et al., 1994) -1.41.70.8

-12.13.5

100100100

000 

Conclusion It was observed that the skin sea surface temperature is different from the bulk sea surface temperature by about a few tenths of a degree Celsius. It was found that when the warm layer occurs, the dry deposition velocity estimated by using BSST is underestimated within the range of 0.8 – 4.3 %; under the other conditions (i..e., only cool skin occurring), the dry deposition velocity estimated by using BSST is overestimated within the range of 0.5 – 2.0 %, varying with pollutants, along the route of the research ship. There are two mechanisms which can change the deposition rate due to the error in using BSST. Firstly, the solubility of a gas pollutant is a function of sea temperature. Secondly, the difference of the estimated turbulent heat fluxes, determined between using BSST and SSST, causes a difference in the estimated stabilities, aerodynamic resistances and, consequently, dry deposition rates. Additionally, it is found that under the warm-layer dominated condition, latent heat flux, sensible heat flux and surface upward longwave radiation were underestimated by 14 W m-2, 3 W m-2 and 4 W m-2, respectively. That is, the net surface ground heat flux absorbed by sea water was overestimated by 21 W m-2. In contrast, under the cool-skin only condition, using BSST as SST, latent heat flux, sensible heat flux and surface upward longwave radiation were overestimated by 16 W m-2, 4 W m-2 and 1 W m-2, respectively. That is, the net surface ground heat flux absorbed by sea water was underestimated by 21 W m-2.

The deposition velocities of CO2 with the enhancement factors proposed by Wanninkhof (1992), Kuss and Schneider (2004) and Seinfeld and Pandis (2006) are determined to be 7.8×10-5, 8.9×10-5 and 3.2×10-3 m s-1, respectively. It can be seen that the deposition of CO2 using the chemical enhancement factor proposed by Seinfeld and Pandis (2006) is higher by about fifty times than those by Wanninkhof (1992) and Kuss and Schneider (2004). Wanninkhof (1992) assumes the enhancement factor to be fixed at 1, and Kuss and Schenider (2004) estimate the enhancement factor (~ 1.15) as an exponential function of pH. It is beyond the scope of this study to judge which enhancement-factor parameterization for estimating CO2 transfer velocity is more accurate. Nonetheless, the direct flux measurements over ocean using the eddy covariance system (Edson et al., 1998; Nilsson and Rannik, 2001; McGillis et al., 2001; 2004) are close to the result estimated using the chemical enhancement factor proposed by Seinfeld and Pandis (2006).