in-situ measurements of land-atmosphere exchanges …
TRANSCRIPT
IN-SITU MEASUREMENTS OF LAND-ATMOSPHERE EXCHANGES OF WATER, ENERGY AND CARBON DIOXIDE IN SPACE AND TIME OVER THE
HETEROGENEOUS BARRAX SITE DURING SPARC2004
Z. Su(1), L. Jia(2), A. Gieske(1), W. Timmermans(1), X. Jin(3), J. Elbers(2), H. van der Kwast(4), A. Olioso(5), J.A. Sobrino(6), F. Nerry(7), D. Sabol(8), J. Moreno(6)
(1) International Institute for Geo-Information Science and Earth Observation (ITC), Enschede, The Netherlands,
[email protected] (2) Alterra, Wageningen University and Research Centre, Wageningen, The Netherlands
(3) China University of Geosciences, Beijing, China (4) University of Utrecht, Utrecht, The Netherlands
(5) INRA, Avignon, France (6) University of Valencia, Valencia, Spain
(7) TRIO/ULP, Strasbourg, France (8) University of Washington, Seattle, U.S.A
ABSTRACT In order to advance our understanding of land-atmosphere exchanges of water, energy and CO2 in space and time over heterogeneous land surfaces, an intensive field campaign was carried out at the Barrax agricultural test site in Spain in the period 12-21 July 2004 involving multiple field, satellite and airborne instruments for characterizing the state of the atmosphere, the vegetation and the soil from visible to microwave range of the spectrum. Part of the experimental area is a core site of a 25 km2 area within which numerous crops are grown - on both irrigated and dry land - alongside fields of bare soil. This campaign was carried out in combination with the EU 6FP EAGLE Project. Emphasis of this paper is on the in-situ measurements of land-atmosphere exchanges of water, energy and CO2 as well as the thermal dynamic states of the atmosphere, the soil and the vegetation.
1. INTRODUCTION
Since turbulent fluxes (water vapor, heat and CO2) occur from scales of an air molecule to regional scale characterized by synoptic circulation and are influenced by both internal biophysical characteristics of the soil and vegetation and external forcings (e.g. solar radiation and wind), the measurements of these fluxes are most challenging over heterogeneous terrains. This is because that the terrain heterogeneity causes in addition to the organized patterns and circulations of turbulent fluxes (due to land uses and dominated by radiative forcing) also secondary effects either in terms of surface geometrical conditions (roughness) or thermal dynamic conditions (dryness or wetness) that may cause local circulation of turbulent fluxes which are much more difficult to observe and characterize. A large number of ground based instruments are therefore necessary for a complete
observation and understanding of the turbulent fluxes in space and time. For this purpose several mobile instrument towers, including four eddy correlation devices and two scintillometers, were deployed in the field to monitor the individual components of the energy, water and carbon dioxide flux exchanged over different surfaces at the Barrax site. When these measurements are combined with airborne and satellite data, such as the new Airborne Hyperspectral System (AHS), operated by Spain's Instituto Nacional de Técnica Aeroespacial (INTA), CHRIS data, ENVISAT Medium Resolution Imaging Spectrometer (MERIS) and Advanced Along Track Scanning Radiometer (AATSR) data, data from the Spinning Enhanced Visible and InfraRed Imager (SEVIRI) instrument aboard MSG-1 (Meteosat Second Generation), as well as Landsat data, the Moderate Resolution Imaging Spectroradiometer (MODIS) and Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), detailed quantification of the characteristics of turbulent fluxes over different surfaces at different scales are possible. The availability of such a detailed data set provides unprecedented opportunity for studying of the scaling behavior of the land-atmosphere exchanges of water and energy and carbon dioxide. The in-situ data relevant to land-atmosphere exchanges (fluxes and state variables) included the following measurements:
1) Radiation balance and sensible heat fluxes from the scintillometer systems;
2) Turbulence, H2O, CO2 fluxes and CO2 concentration from the eddy correlation systems;
3) Soil heat flux, soil temperatures, air temperature and humidity and leaf temperatures;
4) Radiometric surface temperature measurements;
5) Photosynthesis, conductance and transpiration measurements at leaf level.
The locations of the measurements are indicated in Fig. 1, with instruments, period of measurements and surface characteristics given in Tabs. 1-3. In the following sections, the measurements taken in the vineyard are described in detail according to relevant scales. Other relevant issues are described in [1, 2].
Forest nurseryDutch site
garlicvineyard
corn
Bare soil
Wheat
~3 km
3
12
4
5
INRA sitesForest nursery
Dutch site
garlicvineyard
corn
Bare soil
Wheat
~3 km
3
12
4
5
Dutch site
garlicvineyard
corn
Bare soil
Wheat
~3 km
3
12
4
5
INRA sites
Fig. 1. Locations of flux measurements (the base station of the scintillometer and the eddy correlation station were located in the vineyard during the complete campaign period. Other stations were installed for a limited period at different locations).
Tab. 1. Location turbulent fluxes and energy balance measurements (Code – refers to location on Fig. 1) Field (field code)
Code Device height (m)
Period (DOY)
GPS position
Vineyard (V)
Dutch 3.4 195-203
39°03'37"N, 02°06'09"W
Corn – vineyard (C2)
1 4.4 197-202
30S 0577843UTM 4323710
Corn (C2)
2 2.94 197-202
30S 0577804UTM 4323457
Stubble (CW1)
3 - 196-202
30S 0577968 UTM 4323029
Forest nursery (F)
4 1.65 199-201
30S 0578826UTM 4323757alt. 701 m
Bare soil (BS1)
5 1.10 201-202
30S 0579263UTM 4323247
2. IN-SITU OBSERVATION OF LAND-ATMOSPHERE EXCHANGES
2.1 Radiation balance and sensible heat fluxes from the scintillometer systems
The radiation balance was measured at the scintillometer receiver at the base station in the
vineyard from DOY 196 – 202 (July 14 – July 20 2004).
Tab. 2. Locations of scintillometers and energy balance measurements (Code – refers to land use map; T- Transmitter, R-Receiver) Field (Field code)
Device height (m)
Period (DOY)
GPS position
Vineyard (V, Base R)
5.06 196-202
39°03'38"N, 02°06'10"W
Stubble (CW1, Base T)
4.64 39°03'14"N, 02°05'60"W
Corn (C2, R)
2.965 198-199
39°03'25"N, 2°06'04"W
Corn (C2, T)
2.965 39°03'38"N, 2°05'54"W
Forest nursery (F, R)
2.265 199-200
39°03'30"N, 2°05'20"W
Forest nursery (F, T)
2.265 39°03'44"N, 2°05'16"W
Sunflower (SF1, R)
2.265 200-201
39°04'35"N, 2°07'06"W
Sunflower (SF1, T)
2.265 39°04'54"N, 2°07'13"W
Bare soil (BS1, R)
2.215 201-202
39°03'21"N, 2°05'06"W
Bare Soil (BS1, T)
2.215 39°03'10"N, 2°05'00"W
Tab. 3. Surface and vegetation characteristics Field (Field code)
Surface Other characteristics
Vineyard (V)
Row distance=3.35m, Plant distance=1.5m, Plant height=1.5m, LAI=0.52, Vine coverage=33%
Corn (C2)
Plant height=2.0m (16/07/2004), 2.2m (19/07/2004), LAI=1.22, Corn coverage ~ 69%
Stubble (CW1)
-
Forest nursery (F)
Dry plants 20-40 cm height, Plastic nets 35 cm height around young trees, spaced 3 m in row and 4m between rows, Dry plants ~50%
Bare soil (BS1)
-
Sunflower (SF1)
-
The measurements were made at 1 minute interval including shortwave incoming and outgoing radiation, longwave incoming and outgoing radiation (all at height of 4.80 m), soil heat flux (0.5 cm below surface, at 3 locations in the middle of the vine row and below the vine), soil temperature at 4 depths (10.5, 16.5, 21.0, 28.0 cm below surface), relative humidity and air temperature at 2 levels (2.50 and 4.78 m), wind speed at 2 levels (2.60 and 4.88 m), wind direction at 4.88 m, and the large aperture scintillometer (H-LAS) at 5.06 m. Fig. 2 shows the radiation components which were similar for all days of the observation period. Fig. 3 shows the energy balance terms as estimated at the H-LAS receiver site. The sensible heat fluxes measured with the mobile H-LAS are shown in Fig. 4 together with that measured at the base station. The forest nursery field shows the highest value of H and also the highest variability in H.
Barrax 2004 Radiation of Vineyard
-200
0
200
400
600
800
1000
1200
196 197 198 199 200 201 202 203
Day of Year
W m
-2
SW_inc SW_out LW_inc LW_out
Fig. 2. Radiation components (SW_inc: incoming shortwave radiation; SW_out: reflected shortwave radiation; LW_int: incoming longwave radiation; LW_out: outgoing longwave radiation)
Barrax 2004 Fluxes of Vineyard, Scintillometer Measurements
-300
-200
-100
0
100
200
300
400
500
600
700
800
196 197 198 199 200 201 202 203
Day of Year
W m
-2
Rnet G_avg H-LAS LE_Rest.
Fig. 3. Energy balance components (Rnet: net radiation; G_avg: average soil heat flux; H_LAS: sensible heat flux; LE_Rest: latent heat flux estimated as residual of the energy balance)
Barrax 2004 Sensible heat fluxes, Scintilometer Measurements
-200
-100
0
100
200
300
400
500
600
196 197 198 199 200 201 202 203
Day of Year
W m
-2
Vineyard Corn-C2 Reforestation-F1 Sunflower-SF1 Bare-B1
Fig. 4. Energy balance components (Rnet: net radiation; G_avg: average soil heat flux; H_LAS: sensible heat flux; LE_Rest: latent heat flux estimated as residual of the energy balance)
2.2 Turbulent Sensible heat flux, water vapor flux and CO2 flux in the vineyard from the eddy correlation systems
The turbulent H, H2O, CO2 fluxes and CO2 concentrations were measured from DOY 195 – 203 (July 13 – July 21 2004) with an eddy correlation system (Gill 3D sonic + closed path Licor gas analyser: CO2 and H2O + nitrogen reference gas + pneumatic mast + data loggers). The measurement frequency was 20 Hz and the signal was averaged to 10 min, 30 min and 60 min for further analysis (Fig. 5). The most striking feature when comparing the results of the three averages is that temporal averaging smoothes significantly the peak fluxes (e.g. the latent heat fluxes was reduced from 500 W/m2 to approximately 400 W/m2), which poses a serious challenge when fluxes from different measurement and estimation methods are compared to each other [1]. While the sensible heat flux, H, remained approximately constant over the whole observation period, the latent heat flux, LE, decreased gradually with the changing of the local meteorological conditions (changes in wind direction, see Fig. 6). The influence of the drying processes in the vineyard was not likely the cause because the vineyard was drip irrigated and the humidity of the surface air increased from the beginning to the end of the measurement period (see later on discussions on humidity profile).
Fig. 6 shows the observed H from DOY 198-202. On DOY 202, the wind direction has changed from predominantly east and south-east (coming from the corn field) to west and north-west (from the vineyard). While H in previous day were from negative to slightly above zero, H from the last day were at or above 200 W/m2.
Fig. 5. Time-averaged sensible heat flux, H2O and CO2 Fluxes measured with the eddy correlation station
Sensible Heat Flux (H) and Wind Direction at the edge of Vineyard and Corn Field
-100
-50
0
50
100
150
200
250
300
350
400
12:0
012
:30
13:0
013
:30
16:0
08:
209:
5010
:20
10:5
012
:20
15:5
016
:20
7:40
8:10
8:40
10:3
011
:00
11:3
014
:20
14:5
011
:00
11:3
012
:00
12:3
0
Hour/Day
H (W
.M^2
)
0
60
120
180
240
300
360
Win
d D
irect
ion
HWind Direction
Fig. 6. Time-averaged sensible heat flux measured with a 3D sonic anemometer at the corn-vineyard edge (code 1 in Fig. 1 and Tab. 1; North is set at 0 degree)
2.3 Soil heat flux, soil temperatures, air temperature and humidity and leaf temperatures at the eddy correlation location
The soil heat fluxes at the eddy correlation location were measured with 4 soil heat fluxes buried at a depth
of 1cm below surface, with two placed in the middle of the row and two under the vine and shaded. Fig. 7 shows the dynamics and variability of the measured fluxes. The soil temperatures were measured with thermal couples at two levels (Fig. 8), air temperature and humidity measured at 3 levels (Fig. 9) and leaf temperatures (thermal couples) (Fig. 10). All these measurements were recorded at 10 min interval from DOY 195 – 203 (July 13 – July 21 2004).
30 Minutes Averages, Eddy Correlation Measurements
-500
-300
-100
100
300
500
time
flux
(H, L
E)
-20-10
010
2030
40
CO2
Flux
H W m-2 LE W m-2 Fc µmol m-2 s-1
60 Minutes Averages, Eddy Correlation Measurements
-500
-300
-100
100
300
500
time
flux
(H, L
E)
-20
-10
0
10
20
30
40
CO
2 Fl
ux
H W m-2 LE W m-2 Fc µmol m-2 s-1
10 Minutes Averages, Eddy Correlation Measurements
-500
-300
-100
100
300
500
time
flux
(H, L
E)
-20-10
010
2030
40
CO2
Flux
H W m-2 LE W m-2 Fc µmol m-2 s-1
Soil Heat Flux
-100
-50
0
50
100
150
200
250
1830 90
0
1530
2020 11
0
600
1050
1540
2030 12
0
610
1100
1550
2040 13
0
620
1110
1600
2050 14
0
630
1120
1610
2100 15
0
640
1130
1620
2110 20
0
650
1140
1630
2120 21
0
700
Time
W/m
^2
SH1SH2SH3SH4
Fig. 7 Soil heat fluxes (Soil heat flux plates SH1 and SH4 were in the shade; SH2 and SH3 were in the sunlit areas)
Soil Temperature at -5 cm and -30 cm below Surface
0
5
10
15
20
25
30
35
40
45
50
1830 90
0
1530
2020 11
0
600
1050
1540
2030 12
0
610
1100
1550
2040 13
0
620
1110
1600
2050 14
0
630
1120
1610
2100 15
0
640
1130
1620
2110 20
0
650
1140
1630
2120 21
0
700
Time (in 10-minute measurement points)
0 C
Ts1_30Ts2_5Ts3_30Ts4_5Ta_ref
Fig. 8. Soil temperatures measured with thermal couples at two levels (Ts1_30 – 30 cm, Ts2_5 - 5cm, Ts3_30 - 30, Ts4_5 – 5 cm below surface; Ts1_30 and Ts2_5 are under the vine, while Ts3_30 and Ts4_5 are in the middle of the row; Ta_ref – the reference air temperature measured at the eddy correlation station as shown in Fig.5) From Fig. 7, it can be observed that the variability of soil heat flux is very large (e.g. from 50 to 200 W/m2) even within a short distance due primarily to the radiation and soil moisture states. The same physical factors also influence the soil temperature as shown in Fig. 8 in additional to the variations caused by the depth of the sensor’s locations. The humidity information shown by the wetbulb and drybulb temperatures (Fig. 9) indicated that from the height of 0.65 m (approximately the vertical center of the crown layer which may also be thought as the averaged source of the transpiration) to just above the
top of the canopy at 1.65 m and to about twice the canopy height at 3 m, the humidity reduces gradually, indicating the primarily vertical water vapor diffusion into the dry air above the field.
10
15
20
25
30
35
1830 83
0
1510
1950 30 51
0
950
1430
1910
2350 43
0
910
1350
1830
2310 35
0
830
1310
1750
2230 31
0
750
1230
1710
2150 23
0
710
1150
1630
2110 15
0
630
1110
1550
2030 11
0
550
Time
0 C
Tw1Td1Tw2Td2Tw3Td3
Fig. 9. Air temperature and humidity measured at 3 levels (Tw – wetbulb and Td – drybulb temperature; Tw1 and Td1 were set at the height of 0.65 m, Tw2 and Td2 at 1.65m, and Tw3 and Td3 at 3.0m)
Shaded (d) and Sunlit (s) Leaf Temperature
10
15
20
25
30
35
40
45
1830 90
0
1530
2020 11
0
600
1050
1540
2030 12
0
610
1100
1550
2040 13
0
620
1110
1600
2050 14
0
630
1120
1610
2100 15
0
640
1130
1620
2110 20
0
650
1140
1630
2120 21
0
700
Time (in 10-minute measurement points)
0 C
Tl1dTl2sTl3sTl4dTa_ref
Fig. 10. Leaf temperatures measured with thermal couples attached to the back of leaves (Tl1d and Tl4d were for shaded leaves, Tl2s and Tl3s were for sunlit leaves, Ta_ref is also shown as a reference) The measured leaf temperatures indicated that the differences in leaf temperatures in the sunlit leaves and shaded leaves reached 8 0C, while it has been often suggested in the literature that leaf temperatures could be treated as equal. Similarly, the difference between the air temperature and the leaf temperatures could vary between 1 0C in the earlier morning when the air temperature is the lowest up to 11 0C in the afternoon (6 0C for sunlit leaves).
2.4 Radiometric surface temperature measurements
In addition to the leaf temperatures measured with thermal couples, component radiometric temperatures were measured with a Rytek portable radiometer (Fig. 11). The differences among the shaded leaves, shaded soil, sunlit leaves and sunlit soil (in general order of increasing radiometric temperature) can be clearly
distinguished (a range of 30 0C on DOY 200). This indicates that in any modeling or validation studies, care must be taken in interpretation of the variability in temperature measurements when only limited small scale measurements are available. It is all too easy to reach false conclusions if the thermal dynamic state of the surface is not well quantified. Fig. 12 shows explicitly the differences between sunlit and shaded radiometric temperatures measured with a fixed setup and recorded at an interval of 10 min. The maximum difference between sunlit and shaded leaf temperatures reached 10 0C in the later afternoon (a rather interesting finding), while the difference between sunlit and shaded soil temperature reached 26 0C at noon.
10
20
30
40
50
60
70
9:28
10:0
310
:26
11:0
411
:25
15:3
216
:06
17:0
417
:38
18:3
219
:08
9:32
11:0
011
:35
12:3
013
:06
14:3
015
:05
18:3
319
:30
20:0
521
:00
9:03
10:0
010
:35
11:3
212
:04
13:0
113
:34
14:3
115
:30
16:0
417
:00
17:3
318
:30
19:3
2
Time
Fig. 11. Component radiometric temperature measurements for leaves and soil (Tl_sun, Tl_shade, Ts_sun, Ts_shade indicates respectively temperature of the sunlit leaves, shaded leaves, sunlit soil and shaded soil) Fig. 12 also shows a clear phase shift of approximately 1 hour in the surface radiometric temperature of the soil, while such is not observed in leaves. The most plausible cause of such a phase shift is that the shaded soil (as opposed to sunlit soil) is mainly heated up by multiply reflected solar radiation, thermal radiation and sensible heat flux exchanged with the in canopy air, the combination of these effects is less fast than direct solar heating of the sunlit soil. Another factor of importance is the solar radiation transmitted through the canopy layer reaching the shaded soil. Fig. 13 further indicates that while the shaded leave temperatures agree with each other to within 2 0C when measured with a radiometer and a thermal couple, the sunlit leave temperatures can reach a difference of 10 0C in the afternoon due to differential heatings when the solar position changes. Fig. 14 shows a thermal imagery taken with a FLIR thermal camera with a observed radiometric temperature range of 35 0C from 21.67 0C to 56.67 0C. Clearly the component temperatures observed in previous figures can also be separated easily (observation high resolution and not in the hot-spot
Com
pone
nt T
empe
ratu
res
( C )
o
TL_sunTL_shadeTs_sunTs_shade
198 199 200
position). The lowest temperature observed is on the wet soil under the vine due to drip irrigation.
Shaded (d) and Sunlit (s) Leaf and Soil Radiometric Tempertures
0
10
20
30
40
50
60
70
20:4
0
21:5
0
23:0
0
0:10
1:20
2:30
3:40
4:50
6:00
7:10
8:20
9:30
10:4
0
11:5
0
13:0
0
14:1
0
15:2
0
16:3
0
17:4
0
18:5
0
20:0
0
21:1
0
22:2
0
23:3
0
0:40
1:50
3:00
4:10
5:20
6:30
7:40
8:50
10:0
0
Time
0 C
T3_leaf_dT1_leaf_sT4_soil_sT2_soil_d
Diff T_leaf, T_soil
-10.00
-5.00
0.00
5.00
10.00
15.00
20.00
25.00
30.00
Tim
e
22:3
0
0:30
2:30
4:30
6:30
8:30
10:3
0
12:3
0
14:3
0
16:3
0
18:3
0
20:3
0
22:3
0
0:30
2:30
4:30
6:30
8:30
Time
0 C diff T_leafdiff T_soil
Fig. 12. Component radiometric temperature (upper panel) and difference (lower panel) for leaves and soil (diff T_leaf indicates the radiometric temperature difference in sunlit and shaded leaves, diff T_soil for that of soil) Fig. 13. Component radiometric temperature versus component leaf temperature measured with thermal couples (the symbols are similar to previous figures)
2.5 Leaf Level Photosynthesis, conductance and transpiration measurements
Leaf level photosynthesis, conductance and transpiration measurements were taken from 17-19 July 2004 using a handheld CIRAS instrument from PP
systems. Attempts were made to cover the complete 3 days from down to dusk with a measurement interval of 1 hour, however reliable data from first day were only available from 10 AM due to battery failures. Measurements were made from two vine plants by selecting healthy young and old leaves at different height position of the canopy and including both sunlit and shaded leaves. The total usable measurements are 725 (Fig. 15).
Fig. 14. Radiometric temperature of the vineyard measured with a FLIR thermal camera
0
100
200
300
400
500
600
700
800
10:0
010
:24
12:1
013
:49
14:1
115
:30
15:5
017
:32
17:5
019
:42
20:0
76:
569:
019:
1910
:41
11:1
713
:00
13:3
014
:49
15:1
816
:44
17:1
019
:01
19:3
420
:55
21:2
56:
267:
319:
049:
3211
:04
11:3
112
:54
14:1
614
:37
16:2
816
:53
19:4
020
:10
20:3
321
:00
-10.00
-5.00
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
Ci(ppm)EVAP(millimole.M^-2.S^-1)Tleaf(0C)An(millimole.M^-2.S^-1)
1015202530354045505560
20:4
022
:00
23:2
00:
402:
003:
204:
406:
007:
208:
4010
:00
11:2
012
:40
14:0
015
:20
16:4
018
:00
19:2
020
:40
22:0
023
:20
0:40
2:00
3:20
4:40
6:00
7:20
8:40
10:0
0
Time
0 C
Trad_L_shd
Trad_L_sun
Trad_s_sun
Trad_s_shd
TLshd_cont
TL_sun_cont
15-17 July 2004Vineyard, Barrax
Fig. 15. Time series of Internal leaf CO2 concentration (Ci, refer to lhs axis), Transpiration (EVAP, rhs axis), Leaf Temperature (Tleaf, rhs axis) and assimilation (An, rhs axis)
Fig. 16 indicates the leaf internal CO2 concentration with an average range from 200 to 500 ppm, which is quite different as those often assumed in photosynthesis models especially those for dynamic vegetation simulation. Though the transpiration, leaf temperature, and the assimilation show positive correlation, detailed analysis is necessary in order to quantify the details and is presented in following sections.
The relationship between assimilation and PAR (Photosynthetic Active Radiation) is presented in Fig. 16, from which it can be concluded that a positive relation exists between the two. Although the coefficient of determination is 0.76 indicating a high correlation, the scatter is nevertheless quite large especially at large PAR values (e.g. above 1000 millimole.M-2.S-1, the scatter covers practically the whole range of observed assimilation). This reveals the limitations in photosynthesis models based on only PAR and empirical relationships.
Assimialtion vs Photosynthetically Active Radiation
y = -7E-06x2 + 0.0189x - 0.0445R2 = 0.76
-20.0
-10.0
0.0
10.0
20.0
30.0
0 500 1000 1500
PAR (micromole.M-2.S-1)
An
(mill
imol
e.M
-2.S
-1)
Fig. 16. Relationship between assimilation and PAR
In Fig. 17 the relationship between assimilation and leaf conductance is examined. While in the state of art photosynthesis models the relation between photosynthesis and leaf conductance is explicitly formulated and interactively solved, our measurements have shown that such relationship is valid most likely only for sunlit leaves, since there is practically no assimilation in the shaded leaves even though the leaf conductance is not zero.
The relationship between assimilation and leaf transpiration is shown in Fig. 18, which is broadly similar to that shown in Fig. 17. If sunlit and shaded leaves are not separated in the analysis, the resulted coefficient of determination is very low, indicating the general failure of regression type of model in estimation of assimilation on the basis of transpiration and vice versa. It is also interesting to note that in the shaded leaves there is transpiration taking place while the assimilation is zero. A possible explanation to this phenomenon is that the leaves need to maintain at some biologically tolerable temperature range (the measured maximum is 37.2 0C on 19/07/2004 with CIRAS and 40.38 0C on 14/07/2004 and 40.34 0C on 18/07/2004 with thermal couples as shown in Fig. 10; The minimum leaf temperature measured was 12.1 0C at 7.28 AM on 19/07/2004 with CIRAS and 10.710C at 7.10 AM on 19/07/2004 with thermal couples). Under
such conditions, although the assimilation is zero because the leaves are in the shade, the stomata are open and transpiring to keep the leaves cool.
Assimilation vs Leaf Conductance
-10.0
-5.0
0.0
5.0
10.0
15.0
20.0
25.0
0 100 200 300 400
gs (micromole.M-2.S-1)
An (m
illim
ole.
M-2
.S-1
)
Shaded Leaves
Sunlit Leaves
Fig. 17. Relationship between assimilation and leaf conductance
Assimilation vs Transpiration
y = 2.4185x - 1.5168R2 = 0.42
-20.0
-10.0
0.0
10.0
20.0
30.0
0.00 2.00 4.00 6.00 8.00 10.00
Tanspiration (millimole.M-2.S-1)
An (m
illim
ole.
M-2
.S-1
)
Shaded Leaves
Sunlit Leaves
Fig. 18. Relationship between assimilation and transpiration
3. OBSERVATION UNCERTAINTIES ANDENERGY CLOSURE
In this section, the fluxes measured with different instruments are compared so that a first order analysis of the uncertainties could be given. Fig. 19 shows H measured by the eddy correlation system (H_ED) and that by the scintillometer (H_LAS), while the scatter plots of these H are shown in Fig. 20. From both figures, it can be observed that H_LAS is higher than H_ED for both 10 and 60 min averages. The uncertainties may be caused by the different heights of the instruments and certainly by the different fetches covered by each instrument. In addition, the LAS covered part of the stubble field with
higher sensible H where its transmitter was located. H_LAS is approximately 13% higher than H_ED.
H_ED & H_LAS, 10 Minutes Average
-150
-100
-50
0
50
100
150
200
250
300
350
400
Tim
e_ed 3.
8
7.8
11.8
15.8
19.8
23.8 3.8
7.8
11.8
15.8
19.8
23.8 3.8
7.8
11.8
15.8
19.8
23.8 3.8
7.8
11.8 16 20 0 4 8
12.2
16.2
20.2 0.2
4.2
8.2
12.2
16.2
Time
W.M
^-2
H_edH_las
H_ED & H_LAS, 60 Minutes Average
-150
-100
-50
0
50
100
150
200
250
300
350
400
Time_
ed 4 8 12 16 20 1 5 9 13 17 21 2 6 10 14 18 22 3 7 11 15 19 23 4 8 12 16 20 1 5 9 13 17
Time
W.M
^-2
H_edH_las
Fig. 19. Time series of sensible heat fluxes measured by the eddy correlation system and the scintillometer system with 10 min average (upper panel) and 60 min average (lower panel)
In Fig. 21, the energy closure is shown for 10 minutes averaged and 60 minutes averaged fluxes, in both cases the energy closure is not reached with the sum of the turbulent fluxes (H_ED + LE_ED) measured by the eddy correlation system being 10% higher than the available energy (Rn – G0). Assuming that all the measurements are of comparable reliability, this is considered a striking finding in its own right, since if this is indeed true the extra energy measured by the eddy correlation system must be advected to the field being measured from neighboring dry fields by local circulation.
H_ED vs H_LAS
y = 1.1373xR2 = 0.866
-200
-100
0
100
200
300
400
-200 -100 0 100 200 300 400
H_L
AS
H_ED
Energy Closure, 60 min Average
y = 0.9068xR2 = 0.9592
-100
0
100
200
300
400
500
600
-100 0 100 200 300 400 500 600
Rn-
G0
H_ed+LE_ed
Fig. 20. Scatter plots of sensible heat fluxes measured by the eddy correlation system and the scintillometer system with 10 min average (left panel) and 60 min average (right panel)
Energy Closure, 10 min Average
y = 0.9023xR2 = 0.9184
-200
-100
0
100
200
300
400
500
600
700
800
-200 -100 0 100 200 300 400 500 600 700 800
H_ed+LE_ed
Rn-
G0
Energy Closure, 60 min Average
y = 0.9068xR2 = 0.9592
-100
0
100
200
300
400
500
600
-100 0 100 200 300 400 500 600
H_ed+LE_ed
Rn-
G0
Fig. 21. Energy closure with 10 min average (left panel) and 60 min average (right panel)
4. CONCLUSIONS
This paper has presented a complete account of the in-situ measurements of the land – atmosphere exchanges of energy, water and CO2 from leaf level to field scale, including both fluxes and thermal dynamics of the surface and the near-surface atmosphere of the heterogeneous vineyard. The preliminary analyses of the observation and relevant findings have been discussed in each section accompanying the observations. The three most important findings are 1) The most striking observation is that temporal averaging smoothes significantly the peak fluxes (e.g. the latent heat fluxes was reduced from 500 W/m2 to approximately 400 W/m2 when the averaging time changed from 10 min to 60 min), which poses a serious challenge when fluxes from different measurement and estimation methods are compared to each other; 2) Models of photosynthesis and leaf conductance must be used with caution and must take into account temperature difference in sunlit and shaded leaves (up to 10 0C observed), there is practically no assimilation in the shaded leaves even though the leaf conductance is not zero; 3) A striking finding is the energy disclosure (the sum of the turbulent fluxes, H_ED + LE_ED, measured by the eddy correlation system was 10% higher than the available energy, Rn – G0). This indicates that the extra energy measured by the eddy correlation system must be advected to the field being measured from neighboring dry fields by local circulation.
5. REFERENCES
1. Timmermans, et al., Intercomparison of Energy Flux Models using ASTER Imagery at the PARC 2004 Site (BARRAX, SPAIN), ESA Proceedings WPP-250, SPARC Final Workshop, ITC Enschede, 4-5 July 2005, The Netherlands, 2005. 2. Gieske, A.S.M., et al., Processing of MSG-1 SEVIRI Data in the Thermal Infrared – Algorithm Development with the Use of the SPARC2004 Data set, ESA Proceedings WPP-250, SPARC Final Workshop, ITC Enschede, 4-5 July 2005, The Netherlands, 2005.