fluxes of carbon dioxide and water vapor above paddy fields
TRANSCRIPT
Int J Biometeorol (1991) 35:187-194
meteorology
Fluxes of carbon dioxide and water vapor above paddy fields Eiji Ohtaki ~ and Takehisa Oikawa 2
1 College of Liberal Arts and Sciences, Okayama University, Okayama 700, Japan 2 Institute of Biological Sciences, University of Tsukuba, Ibaraki 305, Japan
Received December 15, 1990; revised August 16, 1991; accepted August 19, 1991
Abstract. Atmospheric fluxes of carbon dioxide and water vapor were measured by the eddy correlation tech- nique over a paddy field in 1989. The carbon dioxide was transported downward during daylight hours due to photosynthesis of the paddy crop. The downward flux of carbon dioxide increased with increasing net radi- ation. Maximum values of downward flux varied with the growing stage o f the paddy crop: ca. 0.3 mg m -z s-1 at early vegetative growth stage and ca. 1.3 mg m -2 s-1 at ear formation stage. The daytime totals of downward flux of carbon dioxide also showed seasonal variation reflecting the photosynthetic activity of the paddy crop: ca. 6 g m - 2 at early vegetative growth stage in June and 40 g m - 2 at ear formation stage in September. The sea- sonal variation of daily totals of carbon dioxide flux shows that carbon dioxide of about 28 t h a - 1 is fixed by the paddy crop from transplanting to harvesting. Taking into account the water use efficiency, the paddy crop requires water in amounts at least 100 times that of carbon dioxide fixed by photosynthesis. It is noted that the correlation coefficients between carbon dioxide, water vapor and vertical wind velocity have constant values under near neutral and free convective regimes.
Key words: Carbon dioxide - Water vapor - Eddy corre- lation technique
Introduction
Carbon dioxide and water are the main requirements for crop growth. Accurate determination of the ex- changes contributes to a better understanding of crop productivity and water use efficiency. The eddy correla- tion technique is superior to traditional techniques such as the aerodynamic method and heat balance method. Eddy correlation sensors have been available for measur- ing fluxes of momentum and sensible heat. However,
Offprint requests to : E. Ohtaki
only recently has the eddy correlation sensor been devel- oped for measurement of carbon dioxide flux (e.g., Oh- taki and Matsui 1982; Ohtaki 1984). The sensor can measure atmospheric fluctuations of carbon dioxide and water vapor simultaneously. We therefore used this sen- sor to measure fluxes of carbon dioxide and water vapor over a stand of paddy rice, which is a typical food crop in Japan.
The objective of the present work was to gain a quali- tative understanding of atmospheric fluxes of carbon dioxide and water vapor using data obtained from differ- ent growing stages of the paddy crop and under differing meteorological conditions. The stability dependency of the normalized standard deviation, and correlation coef- ficient of carbon dioxide and water vapor are examined on the basis of Monin-Obukhov similarity theory.
Methods
The measurements of carbon dioxide fluxes were made with the eddy correlation technique over a paddy field at Hachihama Exper- imental Farm of the Faculty of Agriculture, Okayama University. The Hachihama site was reclaimed land and has an area of about 300 x 300 m 2. The surrounding area consisted of a similar paddy surface for at least 500 m in the direction of the prevailing winds. At the Hachihama site, the paddy crop was drilled in rows at the middle of May, and germination occurred about 2 weeks later. The growing phases of the paddy crop comprised an ear formation stage from the middle of August to September and a mature stage from late September to October. The paddy crop was harvested at the end of October.
Observation periods and surface conditions are summarized in Table 1; the daytime totals of carbon dioxide flux (F), daytime averages of wind speed (u), and net radiation (R~) are also shown. Here, the daytime is defined as the period of downward carbon dioxide flux for the crop field. The zero-plane displacement (d) was estimated as 0.7H, where H is crop height. The measuring height is represented by z. In the following discussions, z - d is shown as z for convenience. The soil surface was dry on June 6-7 and muddy on July 19-20. On August 18 and 29 the soil surface was covered with irrigation water (IW) of about 10 cm.
The fluctuations of carbon dioxide and water vapor were mea- sured by a fast-response carbon dioxide-humidity sensor (Ad- vanced System Co., E-009). This instrument is a revised version
188
Table 1. Observation period and crop con- dition Date Time (h) z H d u R, F IW
(cm) (cm) (era) (m s- 1) (W m - z) (g m-Z) (cm)
June 7 ((7700-2 700) ~ 50 4 3 2_3 583 - 6.0 ail July 20 (0600-1800) 165 45 32 1.4 500 -22 .9 nil Aug. 18 (0600-1900) 160 75 53 1.0 371 -29.1 1t Aug. 29 (0600-1800) 150 90 63 1 ~0 433 - 27.1 8
Daytime downward flux of carbon dioxide (F), measuring height (z), crop height (H), zero-plane displacement (d), daily mean wind speed (u), daily mean net radiation (R,), and irrigation water depth (IW) Soil surface was dry on June 7 and muddy on July 20
Aug. I7, 1989
I m
E
r
i
o I--- ~
Le)
?E
0 E : ~-i)
OD
O"
I I I I I I I I I I I 1 I I I I I I I I
1230 1235 1240 1245 IZSO IZ55
T i m e Fig. 1. Monitor records of turbulent fluctuations of vertical wind velocity (w), temperature (T), carbon dioxide (c), and water vapor (q) over the paddy field on August 17, 1989
of the prototype model reported by Ohtaki and Matsui (1982). The noise level of the instrument was about 0.3 ppm for carbon dioxide measurements and about 0.1 g m -3 for humidity measure- ments. The fluctuations of the three wind velocity components and temperature were measured by a three-dimensional sonic ane- mometer-thermometer w~th a sensing path-length of 20 cm (Kaijo Denki Co. Ltd., DAT-390). The net radiation was measured at the same time (Eiko Co. Ltd., CN-11).
The output signals from each sensor were digitized and stored on a floppy disk at a rate of 10 Hz. The sampling duration of a single run was 15 min. Each run was used to calculate the turbu- lence statistics and spectra.
Results and discussion
Time traces o f carbon dioxide
M o n i t o r r eco rd ings o f ver t ica l w i n d ve loc i ty (w), t emper - a tu re (7) , c a r b o n d iox ide (c), a n d wa t e r v a p o r (q) were m a d e d u r i n g o b s e r v a t i o n s o n a 4 - c h a n n e l osc i l lograph . F i g u r e 1 shows typ ica l example s m e a s u r e d over the ac- t ively g r o w i n g p a d d y field f r o m 1230 to 1255 h o u r s o n A u g u s t ?7, ?989. These examptes were o b t a i n e d u n d e r
o \
b ~
I0
I
0 . 5 k
�9 d u n . 6 | du I . 1 9 / 2 0 o Aug. 1 7 / 1 8
+ Aug. 2 8 / 2 9 + . | ,:m-- +
0 0• - ~ - ~ _ +++ +.
. . . . , , I , , , , , , , , I , , , , , , , , t _ _ _ _
0.01 0.1 I
- z / L
Fig. 2. Variation of normalized standard deviation of carbon dioxide with -z/L. Regression lines are ac/c,=2.6 in near neu- tral conditions, and ao/c, = 0.9 (-z/L)-1/3 in free convective conditions
189
\
b ~"
I 0
0.5
�9 dun. 6 ,~ d u l . 1 9 / 2 0
o Aug. 17 / I 8
'--1-- Aug. 2 8 / 2 9 o
+ +
r T l _ ~ . . ~ . ~ . o o
0 . 0 1 0 . I I
-zlL
Fig. 3. Variation of normalized stan- dard deviation of water vapor with -z/L. Regression lines a r e IOq /q , ] =
2.6 in near neutral conditions, and ]~q/q,I = 0.9(--z/L)- 1/3 in free con- vective conditions
unstable conditions. The effects of plumes in fluctuations of T and q can be seen, as reported by Kaimal and Businger et al. (1971); that is, gradual increase and sharp decrease in T and q signals. It is noted that the fluctua- tions of c and T or q correlate closely, though the fluc- tuating direction is opposite. It is also noted that the distinct negative c fluctuations occur when w is positive, and vice versa.
Characteristics of turbulent statistics involved in vertical transport
Vertical turbulent t ransport o f carbon dioxide and water vapor can be expressed as the correlation coefficient R and standard deviation a as follows:
w' c'= Rwo aw ao for carbon dioxide, w'q'=Rwq aw aq for water vapor.
In the following sections are examined the characteristics of standard deviation and correlation coefficient on the basis o f Monin-Obukhov similarity theory.
Figure 2 shows the standard deviation of carbon dioxide normalized by the scaling parameter (c, = - c'w'/u,) as a function of stability parameter ( -z /L) of the air layer, u , is the friction velocity and L is the Monin-Obukhov stability parameter. In order to exam- ine the stability dependency of normalized standard de- viation, the stability parameter was divided into three classes: near neutral (0.004 < - z/L < 0.03), transition ( 0 . 0 3 < - z / L < 0 . 1 ) , and free convection (0.1 < - z / L < 2). In the present study, 47 data f rom near neutral, 32 data f rom transition and 39 data f rom free convective conditions were obtained.
The ratio of ac/c, seems to have a constant value under near neutral conditions. The average value and standard deviation of ac/c, for the 47 data is 2.59__ 0.40; the ratio of ao/c, decreases with increasing instability with the - 1 / 3 power law in free convective conditions. Ohtaki (1985) and Maitani and Ohtaki (1989) have pre- viously reported a similar stability dependency of a~/c,. The functional relationship of ac/c, in the free convec- tive regime can be represented as ac/C, =Ac (--z/L)-1/3. The coefficient A~ is specified by the value of normalized
190
-n
\
5
�9 dun. 6 �9 du I . 19120 o Aug. 17/I O
Aug. 2 8 / 2 9 ++/
(D
. . . . , I , , , , , , , , I , , , , , , , , I
0.01 O. I I
Fig. 4. Variation of normalized stan- dard deviation of vertical wind veloc- ity with -z/L. Regression lines are aw/U, = 1.3 in near neutral condi- tions, and aw/U,=3.2(--z/L) 1/3 in free convective conditions
- z / L
- i
b:
20
I0
�9 dun. 6 du I. 19 /20
o Aug. 17/I 8 o + Aug. 2 8 / 2 9
�9 e ~ ~o ~ ~ 1 1 ~
0 0 d O go J
o
I I I I i [ I I I I I I I I [ I I I I I I I I I
0.01 0.1 I
I I
Fig. 5. Variation of normalized standard deviation of horizontal wind velocity with --z/L. Regres- sion lines are aJu, = 2.6 in near neutral conditions, and au/u, = 6.2(--z/L) 1/3 in free convective conditions
- Z / L
standard deviation at - z / L = 1. Ar = 0.9 is obtained in the present case. A similar value of A~= 1.1 has been obtained for data over wheat fields by Ohtaki (1985).
The stability dependency of normalized standard de- viation of water vapor, laq/q,[, is very similar to that of ao/c, (Fig. 3). The ratio of IO-q/q,I is 2.56___0.51 under near neutral conditions and decreases with increasing instability following the - 1/3 power law in the free con- vective regime. The coefficient of Aq is 0.9. Similar re- sults have been demonstrated by other researchers (e.g., H6gstr6m and Smedman-H6gstr6m 1974; Ohtaki 1985; Takeuchi et al. 1980).
It is apparent that the ratios of aw/U, and a~/u, (Figs. 4, 5) are respectively 1.34___0.51 and 2.63+0.27 under near neutral conditions. Though there are some scatters in the plotted data, these values are consistent with consensus values reported by other researchers (e.g., Lumley and Panofsky 1964). The 1/3 power law of aw/U, and au/u, for the free convective regime is also shown in Figs. 4 and 5. The coefficients of Aw and Au are estimated to be 3.2 and 6.2 in the present case.
The correlation coefficient between c and w is plotted in Fig. 6 as a function of -z /L. From the definition
of correlation coefficient mentioned above, Rw~ can be
rewritten as R,w= w'c'/(awac)= _u , .c , . Taking into ac- O- w O- c
count the stability dependency of ajc , and aw/U,, the value of R w o = - 0 . 2 9 is predicted under near neutral regimes. The average of 47 near neutral data gives Rwc = - 0.30 + 0.05, representing good agreement with the pre- dicted value of -0 .29 . In the free convective regime, we also have the constant value of R w o = - 0 . 3 4 + 0 . 0 9 . This constant value comes from the fact that ao/c, de- creases with (-z/L)-1/3, but aw/U, increases with ( - z / L ) 1/3 in free convective regimes. Thus, the stability ef- fects of ao/c, and aw/U, cancel out each other. It is noted that the value Rwc=--0.35, which is obtained from 39 data in free convective conditions, is very close to the value predicted by the relation of I/(A~.Aw). The value of Rwr varies from - 0 . 3 0 to - 0 . 3 4 in transition regimes of the stability parameter, - z / L . Thus, Rw~ can be ex- pressed as a function of the stability parameter from near neutral to free convective regimes. Therefore, the carbon dioxide flux can also be expressed by a function of - z / L , ao, and aw.
0 i
m
-0.2 -
(.}
n.- - 0 . 4
-0.6
-z/L
0.01 O. I
' ' ' " 1 ' ' ' ' ' ' " 1 ' ' '
o~ | �9 ++ O
O
�9 dun. 6 �9 J u l . 1 9 / 2 0 o Aug. 1 7 / 1 8
+ Aug. 2 B / 2 9
I I I I I I I
o 4- �9
+ ,,f-
O Fig. 6. Correlation coefficient of carbon dioxide and vertical wind velocity as a function of --elL. Regression lines are Rwr in near neutral conditions, and Rw~= -0.34 in free convective conditions
191
�9 J u n . 6 �9 J u l . 1 9 / 2 0
o Aug. 1 7 / 1 8
4- Aug. 2 8 / 2 9
0 "
n~
0 . 6 - -
0 . 4 -
0 . 2 -
0
" ~ ' r 1 6 2 r162 r L ~ q-+4 ~ o
t- o +
0.01 0.1 I
Fig, 7, Correlation coefficent of water vapor and vertical wind veloci- ty as a fanction of --z/L. Regression lines are R,,q=0,30 in near neutral conditions, and Rq~=0.35 in free convective conditions
-z/L
:3
pc-
0 i I
-0.1 -
- 0 .2
-0.:5
- 0 . 4
- 0 , 5
- z / L
0.01 0.1 I . . . . . , . . . . . . . . , ' ' / ;
_ ~
o ~ du I. 1 9 / 2 0
o Aug. 1 7 / I 8 + Aug. 2 8 / 2 9
Fig. 8. Correlation coefficent of vertical and horizontal wind velo- cities as a function of --z/L. Re- gression lines are R,v, =--0.29 in near neutral conditions, and Rw, = -O.5(- z/L)- 2/~ in free convective conditions
The stability dependency o f Rw, is very similar to that o f R,vc, The value o f Rwq is 0 . 3 0 i 0 . 0 6 under near neutral condi t ions and 0.35 _+0.07 under free convective condi t ions (Fig. 7). Rw, the correlat ion coefficient be- tween vertical velocity (w) and hor izonta l wind speed (u), has the cons tan t value o f -0 ,29_+0 ,04 under near
ne~tral condit ions, and decreases with increasing insta- bility fol lowing the - 2 / 3 power law o f - z / L (Fig, 8). This is based on the fact that aw/u, and ~ , j u . bo th increase with the 1/3 power law o f - e l l in free convec- tive regimes. The da ta o f normal ized s tandard deviat ion and correlat ion coefficient are summarized in Table 2.
192
Table 2. Normalized standard deviation and correlation coefficient as a function of stability parameter
Near neutral Free convective condition condition (0.004 < -- z/L < 0.03) (0.1 < -- z/L < 2)
~c/c, 2.59 + 0.40 0.9 ( - z /L) - 113
I%/q,[ 2.56 + 0.51 0.9 ( - z /L) - 1/3
aw/u, 1.34 _+ 0.51 3.2 (-- z/L) 1/3 au/u, 2.63 _+ 0.27 6.2 (--z/L) 1/3 Rw~ -- 0.30_+ 0.05 -- 0.34 + 0.09 Rwq 0.30_+ 0.06 0.35_+ 0.07 Rwu - 0.29 _+ 0.04 - - 0.05 ( - z/L) - 2/3
Time variation of vertical f luxes
Figure 9 shows time variations of vertical fluxes of car- bon dioxide (F), net radiation (R,), sensible heat (H) and latent heat (2E) measured on August 17-18 at ear formation stage of the paddy crop. F is positive with a steady value below 0.3 mg m -2 s -1 in the night-time, representing an upward t ransport of carbon dioxide. This upward transport of carbon dioxide is associated with the crop respiration. The carbon dioxide flux due to soil respiration was negligible in the present case, be- cause the soil surface was covered by irrigation water of about 1 ] cm depth. After sunrise, F values changed from positive to negative. The negative value indicates downward transport o f carbon dioxide due to photosyn- thesis of the paddy crop. It is noted that the downward flux of carbon dioxide increases with increasing net radi- ation. The max imum value of downward flux of carbon dioxide was about 1.3 mg m -2 s - 1 at forenoon, and var- ied with the growing stage of the paddy crop.
Figure 9 also shows that H was small compared with 2E during daylight hours over the paddy field. The paddy field transferred heat to the overlying air almost
% t } % E
0 .5
- 2 0 0 o ~
- 0 . 5
- R n W n~ =
2 0 0 4 0 0 6 0 0 8 0 0 I000 I l l U UI I I I l I 1 I
" . , , O ' ~ o
= = � 9 �9 , J u n e 6 / 7 , 1 9 8 9
~P o 6 : 0 �9 . . . s t .
- I . 0 � 9 1 4 9 �9 If'�9 h �9 �9 �9 �9
- I . 5 August 17/18, 1989
Fig. I 0 . Relation between daytime downward flux of carbon diox- ide and net radiation. White circles indicate data measured at early vegetative growth stage (June 6-7), and solid circles indicate data measured at ear formation stage of the paddy crop (August 17 18)
exclusively through evapotranspiration. Evapotranspira- tion of about 370 W m - a was measured around midday, corresponding to an evapotranspirat ion rate of about 0.5 m m h -1. The variation of 2E was primarily associat- ed with variation in R.; 2E was about 50% of R. re- ceived by the paddy field. The upward flux of )oR throughout the day was due to the surface conditions covered by irrigation water. Similar results o f evapo- transpiration over the paddy fields have been reported by Takeuchi et al. (1980). In order to check the energy balance, the soil heat flux should be taken into account. However, the main interest of the present work is to show the turbulence characteristics of carbon dioxide and water vapor; thus, further discussion on soil heat flux is not presented here.
The downward fluxes (F) of carbon dioxide measured on June 6-7 and August 17-18 are plotted against the mean values of net radiation in Fig. 10. It is seen f rom
"t E
I 0 0 0
8 0 0
6 0 0
4 0 0
2 0 0
0
- 2 0 0
I I I I I I
i i I I I I
Aug.I7 2 2 0 2 4
I I I I I I I I i I I
Rn
X E
I I I I I I I I I I I
Aug. 18 6 8 IO 12 14 1 6 h
E
E
0.5
0
-0.5
- 1.0
-1.5
- 2.0
i t i i i i I i i i i i i i i i i
I I [ I I I I I I I I I i I I I I
Fig. 9. Time variation of vertical fluxes of carbon dioxide (F), la- tent heat (,~E), sensible heat (/4), and net radiation (R.) over the paddy field with irrigation water of 11 cm
E o
"l- L)
!00
5 0
0
193
N 'F:
LL
50
3 0
201
I 0
0
I I0 2 0 I Io 2 0 I io 2 0 i
Jun dul Aug
Ill J l,, I0 20 I I0 20
Se pt 0 c t
rl
3 0
Fig. 11. Seasonal variat ion of dayt ime flux o f ca rbon dioxide over paddy fields, and paddy height. Solid bars indicate data measured by the eddy correlat ion technique in 1989; open bars indi- cate data measured by aerody- namic and heat balance tech- niques in 1969
the figure that the downward flux increased with increas- ing net radiation. Maximum values on June 6-7 and August 17-18 were respectively about 0.3 and 1.3 mg m -2 s -1. The downward fluxes of carbon diox- ide were integrated as a measure of the photosynthetic activity of the paddy crop and are shown in Fig. 11 to- gether with results reported in previous papers (Seo and Ohtaki 1978), in which fluxes were measured by the aero- dynamic and heat balance techniques. The daily totals of carbon dioxide flux show a seasonal variation of ca. 5 g m-2 at early growth stage on June 6-7 and reaching 45 g m-2 at the ear formation stage on September 3-4. The daily fluxes fell rapidly in late September and de- creased to about 2 g m - 2 just before harvest. Taking into account the seasonal variation of daily totals of carbon dioxide flux illustrated in Fig. 11, it can be esti- mated that about 28 t ha- 1 of carbon dioxide is fixed by the paddy crop from transplanting to harvesting.
Water use efficiency
Water use efficiency is defined as the total water required to fix the total dry matter of the paddy crop. The water use efficiency of the paddy crop is considered to be 690 to 710 (e.g., Tani 1963). In the present study, w'q'/w'c' was calculated and is shown as a function of time in Fig. 12. For convenience, the sign difference between w'q' and w'c' is ignored. Except for the early growing stage of the paddy crop, the values of w'q'/w'c; increased monotonically from a minimum value of 70 at 0730 hours to a maximum of 300 at 1800 hours. This implies that the paddy crop is better hydrated in the morning when the temperature is less likely to exceed the opti- mum value. Anderson et al. (1984) reported similar diur- nal variations of w'q'/w'c' measured over a soybean crop. Roughly speaking, the ratio of w'q'/w'c' over the paddy crop ranges from 70 to 300 during the daylight hours
and becomes large during the night-time. It is, however, noted that the value of w'q'/w'c' is smaller than the con- sensus value of water use efficiency mentioned above. One reason for this discrepancy arises from the fact that w'q' and w'c' values include evaporation and respiration from the soil surface respectively. When the water use efficiency is examined, long-term variations of w'q' and w'c' should be taken into account; however, the ratio of w'q'/w'c' presented here is the value for selected peri- ods of time, which thus provides another reason for dis- crepancy. Further work on w'q' and w'c' is required to obtain a precise understanding of water use efficiency. It is, however, noted that the paddy crop requires an amount of water at least 100 times that of carbon dioxide fixed by the photosynthesis. Here, we would like to no-
500
400
�9 Jun. 6, 1989 Jul. 20, ,,
o Aug. 18, .
+ Aug. 29, .
0
" " 3 0 0 - �9
�9 CD �9 �9 0 0 o
�9 oo o o~,oOooo o CD ,+
I "~ 0 0 - o _ , ~ ' ~ oo
e e e e + p~-~o
,oo o o
O f I I I I I [ i i I I I
6 8 I0 12 14 m6
T i m e Fig. 12. Time variat ion of water use efficiency, w'q'/w'c'
t
18 h
194
tice the water use efficiency of other crops: 460-550 for wheat, 510 530 for barley and 230-380 for corn (Tani 1963).
Conclusions
The vertical transports of carbon dioxide and water va- por were measured by the eddy correlation technique over a paddy crop. Atmospheric fluctuations of carbon dioxide and water vapor were investigated on the basis of Monin-Obukhov similarity theory using data ob- tained over paddy fields for a wide range of unstable conditions, 0.004 < - z / L < 2.
The normalized standard deviations of carbon diox- ide and water vapor show constant values under near neutral conditions, and decrease with increasing instabil- ity according to the - 1 / 3 power law in free convective regimes. Similar stability dependency is consistent with results obtained over a wheat field (Ohtaki 1985). It is also noted that the normalized standard deviations of vertical and horizontal components of the wind velocity show a constant value in near neutral conditions, and increases with a 1/3 power law of the Monin-Obukhov stability parameter in free convective regimes. These re- sults agree with those observed in the surface layer over a wide range of field conditions (e.g., Maitani and Oh- taki 1987; Lumley and Panofsky 1964). Taking into ac- count the functional relationships between normalized standard deviations of carbon dioxide, water vapor and vertical wind velocity, the correlation coefficients of car- bon dioxide and water vapor can be divided into three classes: near neutral, transition, and free convective. R w c = - 0 . 3 0 and Rwq=0.30 under near neutral condi- tions (0.004 < - z / L < 0.03). Rwc = - 0.34 and Rwq ----= 0.35 in free convective conditions (0.1 < - z / L < 2). The tran- sition of Rw~ and Rwq occurs in stability ranges from - z /L = 0.03 to - z /L = O. 1.
Considering the seasonal variation of photosynthetic activity, the paddy crop can fix carbon dioxide of about 28 t h a - 1 from transplanting to harvesting. Though the ratio of w'q'/w'c' varies with the growing stage of the
crop, it should be noted that the amount of water re- quired is at least 100 times that of the carbon dioxide fixed by photosynthesis.
Acknowledgements. The authors express their thanks to Prof. Y. Miyake of Okayama University for kind permission to use the experimental fields. This work was partly supported by a grant under the Monbusho International Scientific Research Program of Joint Research (No. 02044102).
References
Anderson DE, Verma SB, Rosenberg NJ (1984) Eddy correlation measurements of CO2, latent heat, and sensible heat fluxes over a crop surface. Boundary-Layer Meteorol 29:263-272
H6gstr6m U, Smedman-H6gstr6m AS (1974) Turbulence mecha- nism at an agricultural site. Boundary-Layer Meteorol 7:373- 389
Kaimal JC, Businger JA (1971) Case studies of a convective plume and a dust devil. J Appl Meteorol 9:612-620
Lumley JL, Panofsky HA (1964) The structure of atmospheric tur- bulence. Wiley, New York
Maitani T, Ohtaki E (1987) Turbulent transport processes of mo- mentum and sensible heat in the surface layer over a paddy field. Boundary-Layer Meteorol 40: 283-293
Maitani T, Ohtaki E (1989) Turbulent transport processes of car- bon dioxide and water vapor in the surface layer over a paddy field. J Meteorol Soc Jpn 67:809-815
Ohtaki E (1984) Application of an infrared carbon dioxide and humidity instrument to studies of turbulent transport. Bound- ary-Layer Meteorol 29:85-107
Ohtaki E (1985) On the similarity in atmospheric fluctuations of carbon dioxide, water vapor and temperature over vegetated fields. Boundary-Layer MeteoroI 32:25-37
Ohtaki E, Matsui T (1982) Infrared device for simultaneous mea- surement of fluctuations of atmospheric carbon dioxide and water vapor. Boundary-Layer Meteorol 24:109-119
Seo T, Ohtaki E (1978) Atmospheric flux of carbon dioxide over paddy fields. In: Monsi M, Saeki T (eds) Ecophysiology of photosynthetic productivity. University of Tokyo Press, pp 145-151
Takeuchi K, Ohtaki E, Seo T (1980) Turbulent transfer of water vapor over paddy fields. Ber Ohara Inst Landw Biol Okayama Univ 18:1-30
Tani N (1963) The wind over the cultivated field. Bull Natl Inst Agric Sci [A] 10:1-99 (in Japanese)