Clim Dyn (2009) 33:33–44 DOI 10.1007/s00382-008-0411-9
Influence of the boundary layer height on the global air–sea surface fluxes Erik Sahle´e Æ Ann-Sofi Smedman Æ Ulf Ho¨gstro¨m
Received: 24 January 2008 / Accepted: 15 April 2008 / Published online: 15 May 2008 Ó Springer-Verlag 2008
Abstract Results from large-eddy simulations and field measurements have previously shown that the velocity field is influenced by the boundary layer height, zi, during close to neutral, slightly unstable, atmospheric stratification. During such conditions the non-dimensional wind profile, um, has been found to be a function of both z/L and zi/L. At constant z/L, um decreases with decreasing boundary layer height. Since um is directly related to the parameterizations of the air–sea surface fluxes, these results will have an influence when calculating the surface fluxes in weather and climate models. The global impact of this was estimated using re-analysis data from 1979 to 2001 and bulk parameterizations. The results show that the sum of the global latent and sensible mean heat fluxes increase by 0.77 W m-2 or about 1% and the mean surface stress increase by 1.4 mN m-2 or 1.8% when including the effects of the boundary layer height in the parameterizations. However, some regions show a larger response. The greatest impact is found over the tropical oceans between 30°S and 30°N. In this region the boundary layer height influences the non-dimensional wind profile during extended periods of time. In the mid Indian Ocean this results in an increase of the mean annual heat fluxes by 2.0 W m-2 and an increase of the mean annual surface stress by 2.6 mN m-2.
E. Sahle´e (&) Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149-1098, USA e-mail:
[email protected] A.-S. Smedman U. Ho¨gstro¨m Department of Earth Sciences, Meteorology, Uppsala University, Villava¨gen 16, 75236 Uppsala, Sweden
Keywords Boundary layer height ERA-40 Latent heat flux Momentum flux Sensible heat flux
1 Introduction The global oceans act as huge reservoirs of heat and moisture. They also extract energy from the wind field resulting in waves and currents. The air–sea transport of heat and momentum is to a large extent controlled by the turbulent eddies in the boundary layer. Thus models used to predict the future weather and climate need to describe the turbulence in order to predict the air–sea surface fluxes. The bulk method is often used in models to calculate the air–sea fluxes. This method relates the flux to the mean gradient and the wind speed using an exchange coefficient. The simplicity of this approach is appealing. However, the behavior and value of the exchange coefficient during different environmental conditions needs to be accurately known. Several field experiments have been devoted to the determination of the exchange coefficients. For heat and humidity they have previously been believed to be constant or slightly increasing with wind speed (see, e.g., the reviews in Smedman et al. 2007b and Sahle´e et al. 2008a, b). However, the responses of the exchange coefficients to changes in environmental conditions are not fully known. Recent experimental findings show that due to re-organization of the turbulence structure (Smedman et al. 2007a, b; Sahle´e et al. 2008a) the exchange coefficient for sensible and latent heat increase dramatically during unstable very close to neutral (UVCN) atmospheric conditions. The increase of the air–sea heat fluxes due to the UVCN effects was found to be in parity with the increase of the radiative
123
E. Sahle´e et al.: Influence of the boundary layer height on the global air–sea surface fluxes
34
forcing (at present) due to anthropogenic emission of greenhouse gases (Sahle´e et al. 2008b). The exchange coefficient used for calculation of the momentum flux, the drag coefficient, has been found to increase with increasing wind speed (see, e.g., Smith et al. 1980; Persson et al. 2005). The exact behavior of the drag coefficient at extreme wind speeds is under debate, whether it continues to increase, level off or even decrease (Powell et al. 2003; Donelan et al. 2004). However, experimental and model studies indicate that the momentum transport may differ significantly to what traditionally is assumed also during low to moderate wind speeds when swell, waves moving faster than the wind, is present. During such conditions the ocean and atmosphere show a reversed coupling, that is, momentum is transported from the oceans to the atmosphere (Smedman et al. 1999; Rutgersson et al. 2001; Smedman et al. 2003; Sullivan et al. 2008). The height of the boundary-layer may also influence the exchange coefficients. Model studies by Khanna and Brasseur (1997) (hereafter referred to as KB97) later supported by experimental data (Johansson 2001; 2003; Ho¨gstro¨m et al. 2008) indicate that the non-dimensional wind gradient during unstable conditions is influenced by the height of the boundary-layer. As will be shown in the next section this will have an indirect influence on the exchange coefficients. Here we investigate the importance of this effect when calculating the global mean air–sea surface fluxes using bulk variables from the European Centre for Medium-Range Forecasts (ECMWF) Re-Analyses (ERA-40) database. The bulk formulations and the exchange coefficients are described in Sect. 2, the experimental setup is treated in Sect. 3, results in Sect. 4 and discussion and conclusions in Sect. 5.
2 Theoretical background
The turbulent fluxes can be calculated using the following bulk formulas: E ¼ kqw0 q0 ¼ kqCE U10 ðqs q10 Þ
ð1Þ
H ¼ cp qw0 h0 ¼ cp qCH U10 ðhs h10 Þ
ð2Þ
2 s ¼ qu0 w0 ¼ qCD U10
ð3Þ
where: the latent heat flux (W m-2) heat of evaporation (J kg-1) kinematic flux of specific humidity (m s-1 kg kg-1) air density (kg m-3)
123
U q H cp w0 h0 CH h s u0 w 0 CD
the exchange coefficient for latent heat, the Dalton number wind speed (m s-1) specific humidity (kg kg-1) sensible heat flux (W m-2) specific heat capacity for air at constant pressure (J kg-1 K-1) kinematic flux of potential temperature (m s-1 K) = the exchange coefficient for sensible heat, the Stanton number potential temperature (K) surface stress (N m-2) kinematic flux of momentum (m2 s-2) the exchange coefficient for momentum, the drag coefficient Subscript s refers to the surface and subscript 10 refers to 10 m height.
As stated by Monin-Obukhov similarity theory (MOST) gradients can be made dimensionless using appropriate scaling variables. For heat, humidity, and momentum these non-dimensional profile functions (u-functions) are: uq ðz=LÞ ¼
oq kz oz q
ð4Þ
uh ðz=LÞ ¼
oh kz oz h
ð5Þ
um ðz=LÞ ¼
ou kz oz u
ð6Þ
where subscripts q, h, and m refer to humidity, sensible heat, and momentum respectively, u* is the friction 0 0 0 0 velocity (m s-1), q ¼ wuq and h ¼ wuh are the scaling parameters for humidity and temperature. z/L is a stability parameter where z is the measurement height (10 m) and L is the Obukhov length (m): L¼
2.1 Bulk formulas
E k w0 q0 q
CE
u3 h0 gkw0 h0v
ð7Þ
where h0 is the potential temperature of the surface layer (K), g is the acceleration of gravity (m s-2), and w0 h0v is the kinematic flux of virtual potential temperature (m s-1 K). Using (4)–(6) the exchange coefficients in (1)–(3) can be expressed as: # " k k CE ¼ ð8Þ lnðz=z0 Þ wm lnðz=z0q Þ wq k k ð9Þ CH ¼ lnðz=z0 Þ wm lnðz=z0t Þ wh 2 k CD ¼ ð10Þ lnðz=z0 Þ wm
E. Sahle´e et al.: Influence of the boundary layer height on the global air–sea surface fluxes
wx ¼
Zz=L
ð1 ux ðfÞÞ=f dðfÞ
ð11Þ
0
Where x = q, h, m, and f = z/L. In most cases the neutral exchange coefficients are of interest, which are obtained by removing the stability influence from (7) to (9). The following forms are then obtained: k k CEN ¼ ð12Þ lnðz=z0 Þ lnðz=z0q Þ k k CHN ¼ ð13Þ lnðz=z0 Þ lnðz=z0t Þ 2 k ð14Þ CDN ¼ lnðz=z0 Þ 2.2 The influence of a limited boundary layer height According to MOST the non-dimensional gradients are function of z/L only. Several field experiments have been devoted to the determination of these functions, see for example, the review by Ho¨gstro¨m (1996). However, comparisons between individual experiments show considerable scatter. KB97 presented a plausible explanation to the observed scatter; using high resolution LES simulations they found the non-dimensional wind gradient to be a function of both z/L and zi/L, where zi is the boundary layer height. During conditions with low boundary layer heights the corresponding um values were shown to be considerably lower than um -values during situations with a deeper boundary layer in the same z/L range. KB97 hypothesize this to be an effect of indirect influence from outer-scale motions. This would indeed show up as scatter when comparing individual averaged experimental results. Variations in other surface layer parameters have also previously been explained by limitation in MOST. The variation in the low frequency part of the horizontal velocity spectrum (Kaimal et al. 1972) and the variations observed in the normalized standard deviation of the horizontal velocity component (Panofsky 1977; Banta 1985) could not be explained solely by variations in z/L. A correct scaling required the inclusion of the boundary layer height. However, KB97 were the first to observe an influence of zi on the non-dimensional wind gradient. Johansson et al. (2001) presented measurements which support the conclusions drawn by KB97. The plotted data in Fig. 1 are from follow up studies where more
measurements, primarily in the marine boundary layer, have been included (Johansson 2003; Ho¨gstro¨m et al. 2008). One can clearly see the division of um according to zi/L values. The solid line (numbered 1) shows the ‘‘standard’’ formulation presented in Ho¨gstro¨m 1996 derived as a mean from several land based experiments, which can be expressed as: um ¼ ð1 19z=LÞ1=4
ð15Þ
The straight lines are tentative fit to the data obtained during conditions with low boundary layer height. The dash dotted line (numbered 2) represents conditions with zi/L in the interval -20 \ zi/L \ -10 and the dashed line (numbered 3) conditions with zi/L in the interval -10 \ zi/L \ 0. When fitting the dash dotted line it has been assumed that it has to passum = 1 at z/L = 0, it is expressed by the following equations: um ¼ 1 þ 3z=L for 0:27\z=L\0
ð16Þ
um ¼ 0:2 for z=L\ 0:27
ð17Þ
The dashed line can be expressed as: um ¼ 1 þ 7:5z=L for 0:12\z=L\0
ð18Þ
um ¼ 0:1 for z=L\ 0:12
ð19Þ
The integrated functions when inserted in (7)–(9) will yield larger exchange coefficients compared to using standard um formulations, which in turn will lead to larger fluxes. 1.2
1
0.8
φm
where k is the von Karma´n constant (0.4), z0, z0q, and z0t are the roughness lengths for momentum, humidity, and temperature, respectively (m). Wm, Wq, and Wh are the integrated non-dimensional profile functions:
35
1
0.6
0.4
2
0.2
3 0 −1
−0.8
−0.6
−0.4
−0.2
0
z/L
Fig. 1 The non-dimensional wind gradient, um, as a function of the stability parameter z/L. Different symbols represent measurements during different zi/L values and were originally presented in Johansson (2003) and Ho¨gstro¨m et al. (2008). Triangles represent conditions with zi/L \ -20, closed circles -20 \ zi/L \ -10 and crosses -10 \ zi/L \ 0. Solid line (numbered 1) is the um equation from Ho¨gstro¨m (1996). Dash dotted line (numbered 2) is a tentative fit to the closed circles (described by (16)–(17)) and the dashed line (numbered 3) is a tentative fit to the crosses (described by (18)–(19))
123
36
E. Sahle´e et al.: Influence of the boundary layer height on the global air–sea surface fluxes
Here we investigate the effect on the long term global air sea surface fluxes.
3 Experimental setup Two calculations of the global air sea fluxes were performed, one reference using the standard um function, that is, (15) and one test run using also (16)–(19) for small zi/L values. The non-dimensional gradients for heat and humidity were set to be equal, a reasonable assumption according to Edson et al. (2004), and also unmodified between the reference and test run. The equation presented in Ho¨gstro¨m (1988) was used: uq ¼ uh ¼ 0:95ð1 11:6z=LÞ1=2
ð20Þ
The flux calculations were made using data from the ERA-40 database. The ERA-40 is a global reanalysis dataset created by ECMWF, which cover the period September 1957 to August 2002. However, 1979 was chosen as a start year due to the large improvements in the observational system at this time, such as: improved instrumentation on observational satellites, increased amount of aircraft data, deployment of drifting ocean buoys and improved quality of wind information from geostationary satellites (Uppala et al. 2005). The model used in ERA-40, the Integrated Forecast System (IFS) has a T159 spectral resolution, or a horizontal resolution of approximately 125 9 125 km with 60 vertical levels. The vertical model levels uses a hybrid coordinate, (Simmons and Burridge 1981; Simmons and Stru¨fing 1983) that is, the upper levels uses a purely pressure coordinate, at mid levels a hybrid between pressure and terrain following r-coordinate is used whereas at the lowest few levels a pure r-coordinate is utilized. The lowest model level is very close to 10 m (depending weakly on temperature and surface pressure). See Ka˚llberg et al. (2004) for a complete list of model levels. ERA-40 has utilized a 3-D variational data assimilation system (3D-var). A wave model, WAM (Komen et al. 1994) is coupled to the atmospheric model which allows the oceans to interact with the atmosphere. A more detailed description of the IFS can be found at: http://www.ecmwf. int/research/ifsdocs/index.html. The atmospheric boundary layer height is diagnosed in ERA-40 using the method presented by Troen and Mahrt (1986). Basically the boundary layer height is defined at the model level where the Richardson number reaches the critical value of 0.25. If the critical value is reached between two model levels a linear interpolation is done to get the exact height. The exchange coefficients were calculated from (8) to (10) using the roughness lengths from ERA-40. However,
123
it turned out that the roughness lengths for heat, z0t (=z0q), in the ERA-40 database yielded unrealistically low CH and CE. Thus, instead these had to be solved from (12) to (13) by assuming constant neutral exchange coefficients, CEN ¼ CHN ¼ 1:1 103 : In ERA-40 the sea surface roughness is calculated using a form of the Charnock relation (Beljaars 1995): z 0 ¼ aM
m u2 þ aCh u g
ð21Þ
Where aM is a constant (0.11), m is the kinematic viscosity (m2 s-1), and aCh is the Charnock coefficient which is returned from the ocean model. The first term on the right hand side dominates during low wind speeds, smooth flow when z0 scale with the kinematic viscosity. The second term is recognized as the original Charnock relation (Charnock 1955) and dominates during rough flow.
4 Results All results are based on calculations using analysis and forecast fields from ERA-40 every sixth hour. Figure 2 shows the mean boundary layer height over the oceans averaged over the entire year for the period 1979–2001. Over large areas the mean boundary layer height is between 600 and 1,000 m. Some regional features can be distinguished. Along the west coast of North- and South America as well as along the west coast of Africa the boundary layer is fairly shallow. This is connected to the cold ocean currents that prevail in these regions, which effectively suppresses convection. A minimum in zi is seen over the tropical regions, which is due to low mean wind speeds. The seasonal variation of the boundary layer height is shown in Fig. 3. In the mid-latitudes the mean boundary layer height varies about 600 m between seasons. Maximum in zi is reached in the hemispheric winter due to the high wind speed and large temperature gradients, minimum is reached in the hemispheric summers. The highest mean zi, 1,400–1,600 m is found over the Pacific Ocean east of Japan during December to January, due to high mean wind speeds and a large air–sea temperature difference. Here we also observe the largest seasonal variation, during June to August the mean boundary layer height is only about 400–600 m, that is, a variation of about 1,000 m. The tropics do not experience much seasonal variation in zi due to small seasonal variation of wind speed and temperature. Figure 4 shows percentage of time of a year when zi/L is in the range -20 to 0. From this figure it is clear that zi/L in this range is quite common, especially in the tropics between 30°S to 30°N, where large areas experience these conditions between 80 and 100% of the time. There is some seasonal variation (not shown) however, this is not
E. Sahle´e et al.: Influence of the boundary layer height on the global air–sea surface fluxes Fig. 2 Mean boundary layer height over the global oceans, averaged over the year for the years 1979–2001
37 m 1200
400
800 8400 00
60 400 0
60
0
408 000
0
800
400
400 600
0
800 400600
800 40 0
6000 80080
800 600 400
80
0 40
600
80
600
200
600
600 400
600
1000
800
400
0 80
180oW
600
0 80
800
0 0 40 60 0
60
800
1000
80
800
400 0 40
0
80
800
600 60 0
00 86400
400
0 80
600
o
60 S
0
800
800
600 0 80
1000
0 80
400
40 0
400
800
0 00 4060 100
800 400 6
400
0 80
30oS
800
606 00 0
400
60800 0
0
80
1000
0 0 401000 60
400
64000 0 60
600
80 0
400
60 040 0 600
80
0
40
400
0 40
0o
400
400
400 800
0
40
600
000 806
0 40 0 60
30 N
0
60
600 40 0
0 600 60
400
800 o
80 0
40 0
600
0 80
40 0
0 40
0
40
4000 40
0
60
1000 400
60 N
8040 00
0
400
400
o
80 0
400 600
400
0 40060
0
40
600 400
0 00 60 4
00 40 4 0 0 40
600
40 0
400
600
200
o
120 W
DJF
200
o
0o
60 W
60oE
200
200
120oE
o
180 E
200
MAM
m 1600 1400 1200 1400 1200 1000 800
1000 600
1000 800 800 600
800
800
800
1000
400 200 0
SON
JJA
600 600 800
800 800 1000
800 1000
1200 800
Fig. 3 Mean seasonal boundary layer height over the oceans
large and the global pattern does not change, that is, maximum at low latitudes gradually declining towards the poles.
The difference of the yearly mean surface stress between the reference run and the test run is shown in Fig. 5a. By using the zi/L dependent um and Wm functions, as is done in
123
E. Sahle´e et al.: Influence of the boundary layer height on the global air–sea surface fluxes
38 Fig. 4 Mean percentage of time of a year when zi/L is in the range -20 \ zi/L \ 0
% 100
60oN
80
o
30 N 60 0o 40
30oS
o
20
60 S
0 180oW
120oW
the test run, the surface stress is increased. Maxima of 4–6 mN m-2 are found in regions between the latitude bands 30°S and 30°N. Notice that the outline of these regions follow the outline of maximum occurrence of zi/L in the range -20 to 0, as shown in the previous figure. In Fig. 5b the corresponding relative increase of the surface stress is shown. Again the maximum increase is found between 30°S and 30°N. The largest increase, 6–8%, is found over the oceans west of the Angolan and Namibian coastline. The seasonal variation of the reference and test run difference in surface stress is shown in Fig. 6. Mid to high latitudes display maxima during the hemispheric winters. In the North Atlantic the maxima reach 6–8 mN m-2 during December to February. The tropical regions show much smaller seasonal variation except over the Indian Ocean, Arabian Sea, and Bay of Bengal, which display a distinct maximum during June to August. This is most likely linked to the monsoon circulation. Figure 7a shows the average increase of the yearly air–sea latent heat flux when using the modified zi/L dependent um and Wm functions. Only the areas between 30°S and 30°N display an increase above 1 W m-2. Again, these areas coincide with the areas which most frequently experience zi/L values between -20 and 0. In no region is the mean annual increase larger than 4 W m-2 and only a small region in the Indian Ocean between Australia and Madagascar shows an increase greater than 3 W m-2. The corresponding relative increase is shown in Fig. 7b. The relative increase is in general between 1 and 2% which is less, about half of the increase shown for the surface stress in Fig. 5b. This can be expected by examining (8)–(10): Wm has a larger impact on CD compared to its impact on CE and CH. In Fig. 8 is shown the seasonal variation in the increase of the latent heat flux. The pattern is similar to that shown
123
60oW
0o
60oE
120oE
180oE
for the surface stress in Fig. 6, that is, small seasonal variation except in the Indian Ocean, Arabian Sea, and Bay of Bengal which have a maximum of 3–4 W m-2 during June to August. The increase of the sensible heat flux in absolute terms, shown in Fig. 9, is much smaller compared to that of the latent heat flux. This is due to the fact that the sensible heat flux is generally smaller than the latent heat flux over the oceans by a factor of about 10 (except high latitudes in the northern hemisphere over the warm ocean currents, where the sensible heat flux is in parity with the latent heat flux), see for example, the climatology by Ka˚llberg et al. (2005). However, the relative increase (not shown) is about the same as the relative increase of the latent heat flux. Table 1 shows the difference between the reference and test calculation using modified Wm functions of the surface fluxes. Results are shown for the global ice free oceans and for three regions: Indian Ocean between 20°S and the equator, south east Atlantic Ocean, that is, the ocean area bounded by the Angolan and Namibian coastline and 20°W between 30°S and 15°S. The third region is part of the east Pacific Ocean, limited by the Californian peninsula and 130°W between the latitude bands 20°N to 40°N. The table is divided into three parts: part (a) shows mean values for the entire year, part (b) mean values for December to February, and part (c) mean values for June to August. In the mean the heat fluxes increase by 1% and the surface stress increase by 1.8% when using the zi/L dependent Wm function. This amounts to an increase of the heat fluxes by 1.14 W m-2 and an increase of s by 2.1 mN m-2. Corresponding increases when averaging over the entire globe (not only the oceans) are 0.09 W m-2 for the sensible heat flux, 0.68 W m-2 for the latent heat flux and 1.4 mN m-2 for the surface stress.
E. Sahle´e et al.: Influence of the boundary layer height on the global air–sea surface fluxes
mN m
a 2 2
2
2 2
4
2
−2 8
2
2
2
2
2 2
2
2
2
2
o
2
2
6
2
2
2
60 N 2
2
2
4 4
2
2
2
2
2 2
4
2
2
2
4
2
2
2
2 2
2
4
2
4
2
2
2
2 2
0
180oW
2
2
2 2
2
2
2
2
2
2
2
2
2
2
2
2
60 S
2
2
2
2
o
2
2
2
2
30oS
4
2
4
2
2
2
2
2
2
2
0o
4
4
2
2
2
4
2
2
2
o
30 N
4
Fig. 5 a Mean annual difference in surface stress calculated using bulk formulations between the test run and reference b As a but showing the relative difference
39
0
120oW
0
60oW
0o
0
0
60oE
120oE
180oE
% 2
2
2
2
2 2
2
2
2
4
2 2
2
2
2
2
2
2
60oN
2 2
2
2
2
24
2
2
2
4
8
2
2
2
2
2
2
b
2
4
2 2
2 2
46 2
4 4
2
2
2
2
2
2
4 4
2
4
2
4
4
2
2
o
30 S
2 2
2
4
4
4
4
2 2
2
2
2
2 2
2
22
4
4
2
2
4
42
4
2
2 2
4
4
2
4
0o
2 2
2
4
24
2
2
2
22 2
2 2
2
2
o
30 N
2 2
2
6 2 2
2
2 o
60 S
2
2
2
2
42
2
2
2
2
2 2
2
2
2
2
2 0 o
180 W
0 o
120 W
Certain regions experience a larger increase than the mean of the surface fluxes as seen in the table. Maxima are found in the Indian Ocean and the ocean outside the Californian peninsula where s increases by 3.6 and 3.9%, respectively, during June to August. Also notice the relatively large seasonal variation between DJF and JJA over the Indian Ocean. This variation is most likely connected to the monsoon circulation.
5 Discussion and conclusions The results presented here is an estimation of how the surface fluxes would be expected to respond when a zi/L dependence is introduced in the bulk formulas. No feedback effects have been considered, which would require a
0 o 60 W
o
0
0
60oE
0
120oE
180oE
0
fully coupled global circulation model. The results indicate that the surface fluxes increase during conditions with low boundary layer height compared to what traditional theory predicts. A possible negative feedback is that the increased fluxes would lead to an increased depth of the boundary layer, which then would reduce the enhanced fluxes. The increased latent heat fluxes may also influence cloudiness and precipitation and in combination with the increased sensible heat flux lower the SST. A lowered SST would have an impact on the mixed layer in the ocean as well as influence the atmospheric stratification. Increased s may have an effect on the ocean’s mixed layer and also affect ocean currents and the wind field. The relatively large scatter in Fig. 1 makes it hard to accurately determine the dependence of um on –z/L during situations with a shallow boundary layer. The trend is clear
123
E. Sahle´e et al.: Influence of the boundary layer height on the global air–sea surface fluxes
40
DJF
mN m
−2
MAM
8
6
4
2
0
JJA
SON
Fig. 6 Mean seasonal difference of the surface stress between the test run and reference
but the scatter adds to the uncertainty of the results. More measurements of this kind are needed. Uncertainty may also be introduced by the marine boundary-heights diagnosed by ERA-40. To the authors knowledge no thorough evaluation of this parameter has been published. However, some experimental evidence support that these values are reasonable. Boundary layer heights over the oceans were measured during the lidar in space technology experiment (LITE). A lidar mounted on the space shuttle Discovery in September 1994 provided global measurements of marine boundary layer heights. Although the experiment only lasted for 9 days these data are in good agreement with the results presented in Figs. 2 and 3. During this period the zonally averaged marine boundary layer heights varied between 600 and 1,000 m (Randall et al. 1998; Beljaars and Kohler 2000). Zeng et al. (2004) presents boundary layer heights over the eastern Pacific from 973 radio soundings compiled during 11 cruises from 1995 to 2001. These data also support Figs. 2 and 3, showing a minimum of 350–750 m close to the equator gradually increasing further to the north and south. The boundary layer heights measured over
123
the tropical Indian Ocean and central Arabian Sea during January–March 1999 (Subrahamanyam et al. 2003) vary between 450 and 1,070 m which is close to the mean values in Fig 3. Airborne measurements of the boundary layer height over the Japan Sea, February 3, 2000, between Vladivostok and Japan are somewhat lower than depicted by the climatology, however, a similar structure is seen where the boundary layer height increases from 400 m to about 1,000 m in the NW to SE direction (Dorman et al. 2004). Over the Baltic Sea the boundary layer is generally shallow (a few hundred meters), which also is seen in the September 2003 measurements presented in Ho¨gstro¨m et al. (2008). We are aware of that these comparisons doesnot validate the ERA-40 boundary layer heights however, they lend confidence that they are reasonable. The total increase of the heat fluxes (latent + sensible) averaged over the entire globe is 0.77 W m-2. For comparison, the intergovernmental panel on climate change (IPCC) estimate in the fourth assessment report (AR4) that the anthropogenic emissions of greenhouse gases up to present day have increased the global radiative forcing by
E. Sahle´e et al.: Influence of the boundary layer height on the global air–sea surface fluxes
−2
Wm
a
5
60oN
1
2
1
2
1 2
2
1
1
1
1
2
2
2
1
1
1 1
2
1
2
1
2
1
30 S
3 1
1 3
2
1
1 2
1
1
2
2 o
1 2
2
1
1
2
1
1
1 1
2
1
1
1
1
2
2
1 12
3
1 1
1
0o
1
1
1
2
1
30oN
1
1
4
1
Fig. 7 a Mean difference of the latent heat flux between the test run and reference. Averaged over the year for the period 1979–2001 b As a but showing the relative increase of the latent heat flux when using the zi/L dependent Wm functions as is done in the test run
41
o
1 1
60 S
0
0
o 180 W
120oW
0
0
0
o
o 60 W
o
o
0
60 E
o
120 E
180 E
%
1 1
1
1
1
1 2
1
1
1
1
1
1
1
1
1
1
1
1
2
2
1 1
1
21
2
2
2 1
1
23 1
1
1
1
30oS
1
1
1 1
1
1
1
3
1
2 1
1
1
2
1 1
2
1
2
1
1
0o
1
1
3 1
1
1
1
1 1 1
1
12
2
1
1
1
1
1 1
2
1
1
11
1 1
30oN
1
1
1
1
1
1
2
1
1
3
1
1
1 1
1 2
3 22 1
3
o
60 N
2
1 2
2
4
1
1
1
2
1
b
1
1
1 o
2.63 ± 0.26 W m-2 (Forster et al. 2007). However, this forcing can be expected to be more or less uniformly distributed whereas the increase of the fluxes presented in this study display some regional variation, as shown in Table 1. Other parameters besides the boundary layer height may have an influence on the exchange coefficients. Esau (2004) reviews some studies indicating that it is necessary to incorporate effects of stability in the free atmosphere above the boundary layer in the CD parameterization. This effect is also seen in large eddy simulations and can be quantified by the Zilitinkevich number (Zilitinkevich and Calanca 2000), that is, the ratio N/|f|, where N is the BruntVa¨isa¨la¨ frequency and f is the Coriolis parameter. Larse´n et al. (2004) present data showing that presence of a low level wind maximum above the stable boundary
1
1
1
1
120oW
1
1
1 1
1
1
1
0
180oW
1 1
1
1
1
1
1
1
1
1
60 S 1
0
60oW
1
1 1
0
0o
o
60 E
0 o
120 E
180oE
0
layer likely affects the air sea exchange. The mechanism explaining this phenomenon is called shear sheltering and is thoroughly described in Hunt and Durbin (1999) and Smedman et al. (2004). In short, given certain criteria are fulfilled, the strong shear layer will act as a vortex sheet which will produce an upward force preventing penetration of downward moving large eddies, thus effectively decreasing the efficiency of the turbulent transfer near the surface. Sea spray droplets produced by breaking waves possibly influences the air–sea exchange of sensible and latent heat during high wind conditions. The spray is likely to influence the air–sea exchange mainly due to the following processes: energy exchange through the droplet interface, evaporation from the droplet which in
123
E. Sahle´e et al.: Influence of the boundary layer height on the global air–sea surface fluxes
42 DJF
Wm
−2
MAM
5
4
3
2
1
0
SON
JJA
Fig. 8 Difference in latent heat flux between the test run and reference for the different seasons Fig. 9 As Fig. 7a but for sensible heat flux
−2
Wm 0.1
00.1 .2
0.1 0.1
0. 1 0.1
turn also will cool the surrounding air thus also enhancing the sensible heat transfer from the sea surface directly. Microphysical models such as the one presented
0.1
0.1
0.2
0.1
0.1
0.1
0.
0.2
1
0.2
0
0 o
120 W
0.2 0.2
0.1 0.1
0.3
0.1
0.1
0.1
00.1.2
0.1
0.1
0.2
0.1
0.1
0
180oW
0.2
0.02.1
0.2
0.1
0.1
0.1
1 0.1
0.1
0.
0.3
0.1 0.2
0.1
0.1
1
0.1
0.1 0.2 0.3
2
0.
1
0.1
0.1
0.3
0.1
0.1
0.
o
123
0.2 0.4
0.1
2
0.4
0.
0.2
0.1
0.1
60 S
1
1
0.
60oW
0.5
0.
0.
0.2
0.1
0.1
30oS
0.2 0.1
0.1
0.2
0
0.1
2
0.1
o
0.1
0.1
0.
0.2
0.1
30.2 .1 0
0.1
0.1
30oN
0.
2 0. 0.1
0.1
0.1
0.1 0.2. 0 3
0.1
0.1
0.3
0.2 0.1
0.6
0.1
0.2
0.2
1
0.
0.2
60 N
0.1
o
3 02.1 0. 0.1
0.2 0.1
0.2
0.
0.2
0.1
0o
60oE
120oE
180oE
0
in Andreas (2004) are being developed for quantification of the spray influence however, no consensus on this issue has yet been reached.
E. Sahle´e et al.: Influence of the boundary layer height on the global air–sea surface fluxes
43
Table 1 Regional influence of using a zi/L dependent Wm-function when calculating the bulk fluxes s Diff. (mod.-ref.) mN m-2
H Diff. (mod.-ref.) W m-2
E Diff. (mod.-ref.) W m-2
s Diff. (mod.-ref.) %
H Diff. (mod.-ref.) %
E Diff. (mod.-ref.) %
2.1
0.14
1.0
1.8
0.9
1.0 1.3
(a) Year Ice free oc. Indian Ocean
2.6
0.17
1.8
2.9
1.3
S.E. Atl. oc.
2.4
0.16
1.2
3.0
1.1
1.2
Pac. Cal.
1.9
0.17
1.0
3.0
1.1
1.1
Ice free oc.
2.1
0.14
1.0
1.8
0.9
1.0
Indian Ocean
1.7
0.11
1.3
3.1
1.0
1.0
(b) DJF
S.E. Atl. oc.
2.2
0.13
1.3
3.2
1.2
1.2
Pac. Cal. (c) JJA
1.5
0.13
0.8
2.1
0.8
0.9
Ice free oc.
2.1
0.13
1.0
1.8
1.0
1.0
Indian Ocean
3.8
0.22
2.3
3.6
1.5
1.5
S.E. Atl. oc.
2.2
0.15
1.0
1.5
0.9
0.8
Pac. Cal.
2.4
0.20
1.2
3.9
1.9
1.7
Ice free oc. = Ice free oceans, S.E. Atl. Oc. = South East Atlantic Ocean, Pac. Cal. = Pacific Ocean, west of the Californian peninsula. Table 1a shows the yearly averages, whereas Table 1b and Table 1c show the averages for December, January, February and June, July, August, respectively
During conditions with swell, that is, long wave moving faster than the wind, measurements (e.g., Smedman et al. 1999; Rutgersson et al. 2001) and model studies (Sullivan et al. 2008) have shown that wave field strongly influences the atmospheric surface layer. The momentum flux during these conditions has an upward directed component induced by the long waves influencing the wind and turbulence profiles. Larse´n et al. (2004) found that the non-dimensional wind gradient is modified during these situations, which thus will modify the exchange coefficients through eqs. (7)–(9). Recent studies of the exchange coefficient have shown that the air–sea exchange of sensible and latent heat are enhanced during unstable very close to neutral conditions due to a very special reorganization of the turbulence structure in the atmospheric surface layer (Smedman et al. 2007a, b; Sahle´e et al. 2008a). During such conditions the exchange coefficients for sensible and latent heat were observed to be much larger than predicted by traditional theory. In Sahle´e et al. (2008b) the global impact of this enhancement was found to be of equal magnitude as the increased radiative forcing due to anthropogenic emissions of greenhouse gases. The results presented in this study in combination with the other effects such as those mentioned above may have a significant impact on the global energy cycle. Thus, we speculate that the uncertainty in the calculations of the future weather and climate could possibly be reduced by inclusion of these processes.
Acknowledgments We would like to thank Per Ka˚llberg for the help with ERA-40 and also ECMWF for making the ERA-40 data available.
References Andreas EL (2004) A bulk air–sea flux algorithm for high-wind spray conditions, version 2.0. Preprints 13th conference on Interactions of the sea and atmosphere, Portland ME, Amer Meteor Soc P1.5 Banta RM (1985) Late-morning jump in TKE in the mixed layer over a mountain basin. J Atmos Sci 42:407–411 Beljaars ACM (1995) The parameterization of surface fluxes in largescale models under free convection. Quart J Roy Meteor Soc 121:255–270 Beljaars ACM, Kohler M (2000) LITE/ECMWF boundary layer height intercomparison. www.ecmwf.int/research/physics/lite Charnock DV (1955) Wind stress on a water surface. Quart J Roy Meteor Soc 81:639–640 Donelan M, Haus BK, Reul N, Plant WJ, Stiassnie M, Graber HC, Brown OB, Saltzman ES (2004) On the limiting aerodynamic roughness of the ocean in very strong winds. Geophys Res Lett 31. doi:10.1029/2004GL019460 Dorman CE, Beardsley RC, Dashko NA, Friehe CA, Khelif D, Cho K (2004) Winter maritime atmospheric conditions over the Japan Sea. J Geophys Res 109:C12011. doi:10.1029/2001JC00197 Edson JBC, Zappa J, Ware JA, McGillis WR (2004) Scalar flux profile relationship over the open ocean. J Geophys Res 109. doi: 10.1029/2003JC001960. Esau IN (2004) Parameterization of a surface drag coefficient in conventionally neutral planetary boundary layer. Ann Geophysicae 22:3353–3362 Forster P, Ramaswamy V, Artaxo P, Berntsen T, Betts R, Fahey RW, Haywood J, Lean J, Lowe DC, Myhre G, Nganga J, Prinn R, Raga G, Schultz M, Van Dorland R (2007) Changes in atmospheric constituents and radiative forcing. In: Solomon S,
123
44
E. Sahle´e et al.: Influence of the boundary layer height on the global air–sea surface fluxes
Win D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, Miller HL (eds) Climate Change 2007: the physical science basis. Contributing of working group 1 to the fourth assessment report of the intergovernmental panel on climate change. Cambridge University Press, Cambridge Hunt JCR, Durbin PA (1999) Perturbed vortical layers and shear sheltering. Fluid Dyn Res 24:375–404 Ho¨gstro¨m U (1988) Non-dimensional wind and temperature profiles in the atmospheric surface layer. Bound-Layer Meteorol 42:263– 270 Ho¨gstro¨m U (1996) Review of some basic characteristics of the atmospheric surface layer. Bound-Layer Meteorol 78:215–246 Ho¨gstro¨m U, Sahle´e E, Drennan WM, Kahma KK, Smedman A, Johansson C, Pettersson H, Rutgersson A, Tuomi L, Zhang F, Johansson M (2008) Momentum fluxes and wind gradients in the marine boundary layer—a multi platform study. Boreal Env Res (in press) Johansson C, Smedman A, Ho¨gstro¨m U, Brasseur JG, Khanna S (2001) Critical test of the validity of Monin-Obukhov similarity during convective conditions. J Atmos Sci 58:1549–1566 Johansson C (2003) Influence of external factors on the turbulence structure in the boundary layer. PhD-thesis, Department of Earth Science, Uppsala University Kaimal JC, Wyngaard JC, Izumi Y, Cote´ OR (1972) Spectral characteristics of surface-layer turbulence. Quart J Roy Meteor Soc 98:563–589 Khanna S, Brasseur JG (1997) Analysis of Monin-Obukhov similarity from large-eddy simulation. J Fluid Mech 345:251–286 Komen GJ, Cavaleri L, Donelan M, Hasselmann K, Hasselmann S, Janssen PAEM (1994) Dynamics and modelling of ocean waves. Cambridge University Press, Cambridge, p 532 Ka˚llberg P, Simmons A, Uppala S, Fuentes M (2004) The ERA-40 archive. ERA-40 project reports series No. 17, European Center for Medium Range Weather Forecasts, Shinfield, Reading, UK. http://www.ecmwf.int/publications Ka˚llberg P, Berrisford P, Hoskins B, Simmons A, Uppala S, LamyThe´paut S, Hine R (2005) ERA-40 Atlas. ERA-40 project series No. 19, European Center for Medium Range Weather Forecasts, Shinfield, Reading, UK. http://www.ecmwf.int/publications Larse´n XG, Smedman A, Ho¨gstro¨m U (2004) Air–sea exchange of sensible heat over the Baltic Sea. Quart J Roy Meteorol Soc 130:519–539 Panofsky HA, Tennekes H, Lenschow DH, Wyngaard JC (1977) The characteristics of turbulent velocity components in the surface layer under convective conditions. Bound-Layer Meteor 11:355– 361 Persson POG, Hare JE, Fairall CW, Otto WD (2005) Air–sea interaction processes in warm and cold sectors of extratropical cyclonic storms observed during FASTEX. Quart J Roy Meteorol Soc 131:877–912 Powell M, Vickery PJ, Reinhold TA (2003) Reduced drag coefficient for high wind speeds in tropical cyclones. Nature 422:279–283 Randall DA, Shao Q, Branson M (1998) Representation of clear and cloudy boundary layer in climate models. In: Holtslag AAM, Duynkerke PG (eds) Clear and cloudy boundary layers. Roy Neth Acad Arts and Sci, Amsterdam, pp 305–322 Rutgersson A, Smedman A, Ho¨gstro¨m U (2001) Use of conventional stability parameters during swell. J Geophys Res 106:27117– 27134 Sahle´e E, Smedman A, Ho¨gstro¨m U, Rutgersson A (2008a) Reevaluation of the bulk exchange coefficient for humidity at sea
123
during unstable and neutral conditions. J Phys Oceanogr 38:257– 272 Sahle´e E, Smedman A, Rutgersson A, Ho¨gstro¨m U (2008b) Influence of a new turbulence regime on the global air–sea heat fluxes. Submitted to J Climate Simmons AJ, Burridge DM (1981) An energy and angular-momentum conserving finite-difference scheme and hybrid vertical coordinates. Mon Weather Rev 109:758–766 Simmons AJ, Stru¨fing R (1983) Numerical forecast of stratospheric warming events using a model with a hybrid vertical coordinate. Quart J Roy Meteorol Soc 109:81–111 Smedman A, Ho¨gstro¨m U, Bergstro¨m H, Rutgersson A, Kahma KK, Pettersson H (1999) A case study of air–sea interaction during swell conditions. J Geophys Res 104:25833–25851 Smedman A, Guo Larse´n X, Ho¨gstro¨m U (2003) Effects of sea state on the momentum flux over the sea during neutral conditions. J Geophys Res 108. doi:10.1029/2002JC001526 Smedman A, Ho¨gstro¨m U, Hunt JCR (2004) Effects of shear sheltering in stable atmospheric boundary layer with strong shear. Quart J Roy Meteorol Soc 126:31–50 Smedman A, Ho¨gstro¨m U, Hunt JCR, Sahle´e E (2007a) Heat/mass transfer in the slightly unstable atmospheric surface layer. Quart J Roy Meteorol Soc 133:37–51 Smedman A, Ho¨gstro¨m U, Sahle´e E, Johansson C (2007b) Critical reevaluation of the bulk transfer coefficient for sensible heat over the ocean during unstable and neutral conditions. Quart J Roy Meteorol Soc 133:227–250 Smith SD (1980) Wind stress and heat flux over the ocean in gale force winds. J Phys Oceanogr 10:709–726 Subrahamanyam DB, Ramachandran R, Guptka KS, Mandal TK (2003) Variability of the mixed layer heights over the Indian Ocean and central Arabian Sea during Indoex, IFP-99. BoundLayer Meteor 107:683–695 Sullivan PP, Edson JB, Hristov T, McWilliams JC (2008) Large-eddy simulations and observations of atmospheric marine boundary layers above nonequilibrium surface waves. J Atmos Sci 65:1225–1245 Troen I, Mahrt L (1986) A simple model of the atmospheric boundary layer; sensitivity to surface evaporation. Bound-Layer Meteor 37:129–148 Uppala SM, Ka˚llberg PW, Simmons AJ, Andrae U, da Costa Bechtold V, Fiorino M, Gibson JK, Haseler J, Hernandez A, Kelly GA, Li X, Onogi K, Saarinen S, Sokka N, Allan RP, Andersson E, Arpe K, Balmaseda MA, Beljaars ACM, van de Berg L, Bidlot J, Bormann N, Caires S, Chevallier F, Dethof A, Dragosavac M, Fisher M, Fuentes M, Hagemann S, Ho’lm E, Hoskins BJ, Isaksen L, Janssen PAEM, Jenne R, McNally AP, Mahfouf J-F, Morcrette J-J, Rayner NA, Saunders RW, Simon P, Sterl A, Trenberth KE, Untch A, Vasiljevic D, Viterbo P, Woollen J (2005) The ERA-40 re-analysis. Q J R Meteorol Soc 131:2961– 3012. doi:10.1256/qj.04.176 Zeng X, Brunke MA, Zhou M, Fairall C, Bond NA, Lenschow DH (2004) Marine atmospheric boundary layer height over the eastern pacific: Data analysis and model evaluation. J Clim 17:4159–4170 Zilitinkevich SS, Calanca P (2000) An extended similarity-theory for the stably stratified atmospheric surface layer. Quart J Roy Meteorol Soc 128:1571–1587