304 W. Luo, J. Goudriaan Agricultural and Forest Meteorology 104 2000 303–313
of heat from the leaf surface, followed by condensa- tion of water vapour. Thus, nocturnal net radiative loss
plays an important role in the dew formation process. Nocturnal net radiative loss is a direct or indirect in-
put of many dew formation simulation models that use the energy balance approach e.g. Pedro and Gillespie,
1982a, b; Jacobs et al., 1990; Wittich, 1995; Wilson et al., 1999. Its effect on dew formation was theoreti-
cally studied by comparing it with model outputs or by sensitivity analyses Scherm and van Bruggen, 1993.
The amount of dew formed on a surface depends on how much nocturnal net radiation is balanced by
latent heat. Simulated dew formation practically al- ways deviates from measured results. This deviation
can be attributed to measurement errors, both of dew formation and of relevant weather data, but also to
deviations in the model structure from reality. The objective of this study was to experimentally inves-
tigate how the dew formation depends on nocturnal net radiative loss. The understanding of these factors
is essential for accurate estimation of dew formation.
2. Materials and methods
2.1. Site, treatment and crop The experiment was carried out at the Interna-
tional Rice Research Institute IRRI, Los Baños 14
◦
11
′
N, 121
◦
15
′
E, 20.0 m amsl, Philippines dur- ing 16 rain-free nights 22–24, 28 February, 1, 3–4,
28, 30–31 March, and 1–2, 6–9 April in 1994 the dry season in the wet tropics. Over rice canopies in
four 4×5 m plots, global radiation at 2.5 m above the paddy water surface, net at 2.0 m above the
paddy water surface and paddy water temperatures at 0.05 m below the water surface were automati-
cally monitored. All aerial sensors were mounted on tripods, each of which was set up at the centre of
one of the four plots. To avoid disturbance of the rice
Table 1 Crop height H and LAI at different development stage variety: IR72
DayMonth 22 February
3 March 28 March
6 April Development stage
Tillering Tillering
Grain filling Dough ripe
H m
0.5 0.6
0.85 0.85
LAI 2.5
3.0 4.5
3.5
canopies, a walk board was installed between the tri- pod and the edge of the field. The four plots were near
the centre of a 25×50 m paddy rice field. To mea- sure the total amount of dew during the whole night,
two plots were kept continuously open without cover control treatment. The other two plots were cov-
ered with a sheet of 4×5 m black plastic from sunset 18:00 till 03:00 or 04:00 or 05:00 of the next day to
create different levels of nocturnal net radiative loss. The plastic cover was supported at a height of 2.5 m
above the ground by a wooden frame. The removal of the cover was achieved by rolling up the plastic.
In this way, two exposure durations to nocturnal net radiative loss were observed each night.
The rice variety used in the experiment was IR72. The experiment was done in the period of crop devel-
opment from tillering to dough ripe stage. Crop height, H
, and leaf area index, LAI, during the experimental periods are given in Table 1.
2.2. Instrumentation An ES230 LI-Cor pyranometer, REBS net radio-
meter model Q-6, and RM Young Wind Sentry anemometer were used to measure global radia-
tion, net radiation, and wind speed, respectively. The manufacturer’s calibrations were used. Air and
paddy water temperature and air humidity were measured with copper–constantan thermocouples
and copper–constantan thermocouple psychrometers which were made and calibrated in the Meteorology
Department of Wageningen Agricultural University. All sensors were connected to a CR10T data logger
and an AM416 multiplexer Campbell Scientific. The sampling interval was 2 s for all the elements
mentioned above except wind speed for which the sampling interval was 10 s. All outputs were averaged
hourly.
Blotting paper was used to collect dew because it can be easily weighed. Its thermal emissivity is close
W. Luo, J. Goudriaan Agricultural and Forest Meteorology 104 2000 303–313 305
to unity , similar to that of a real leaf, and also its heat capacity is small, so that thermal equilibration occurs
rapidly. The amounts of dew at H, 23H, and 12H were measured by weighing the blotting paper installed be-
fore sunset. In February, a circular blotting paper with diameter of 90 mm was used as a substitute leaf for
dew formation, with two replicas at each height. In March and April, the same kind of paper was used
to collect the dew with five replicas at each height. To avoid the blotting paper being saturated by water,
each replica had five pieces of paper pinned together by two paper clips and was installed horizontally, at-
tached to an erect bamboo stick. At crop height, care was taken to prevent the blotting paper from touch-
ing the rice leaves. However, at 23H and 12H inside the canopy, this could not be prevented. The blotting
paper was kept in the field until the next morning and weighed three times during each observation day, i.e.
before sunset, before removing the cover, and around sunrise 06:00. The dew amount at each height was
calculated from the increase in weight of the blotting paper installed at that height divided by the area of
the blotting paper. Thus, the unit of the dew amount was kg m
− 2
or millimetre, referring to leaf or blotting paper area. The onset and disappearance of dew was
observed visually. After sunset, it was sensed man- ually whether dew had appeared on leaf surfaces or
not. In this way, the rice leaves were checked every 15 min until dew was detected. After sunrise, to de-
tect drying, the rice leaves were visually checked with the same frequency. The dew duration was calculated
as the period between the onset and disappearance of dew, expressed in hours.
2.3. Model The MICROWEATHER model Goudriaan, 1977
was used to simulate dew amount and duration. This model is a multilayer model. In this model, the energy
and mass balances of a canopy, the partitioning of the absorbed radiation R
n
into sensible heat H and latent heat λE is calculated based on the combination
of the following energy balance equations: R
n
− H − λE =
H = T
l
− T
a
ρc
p
r
b,h
λE = e
S
T
l
− e
a
ρc
p
γ r
l,v
+ r
b,v
e
S
T
l
= e
S
T
a
+ sT
l
− T
a
where T
l
and T
a
are leaf and air temperatures, re- spectively, ρc
p
the volumetric heat capacity of the air, e
S
T
l
and e
S
T
a
the saturated vapour pressure at leaf temperature and air temperature, respectively, e
a
the actual vapour pressure, γ the psychrometric con- stant, s the slope of the saturation vapour pressure
curve at air temperature, r
b,h
and, r
b,v
the boundary layer resistances to heat and water vapour, respec-
tively, and r
l,v
the leaf resistance to water vapour. Resistances for these fluxes were computed from the
wind profile and microclimatic conditions, following the stomatal-photosynthesis model of Leuning 1995.
The profiles of temperature and vapour pressure were found by integration of the net fluxes over time for
different canopy and soil layers. The partitioning of the available net radiation at soil surface into sensible
heat, latent heat and soil water in paddy field heat fluxes was also computed. At night, the leaf resistance
is very large, but when the latent heat flux becomes negative, or when dew is present, it is reduced to zero.
Dew formation on different layers was then simulated using the following equation Monteith and Unsworth,
1990:
λE = sR
n
+ ρc
p
Dr
h,v
s + γ 1
where D=e
S
− e
a
is the local vapour pressure deficit. 2.4. Guttation
During the 16 experimental nights, dew never oc- curred in the covered plots before the cover was
removed. But the blotting papers installed inside the canopy at 12H and 23H which were meant to just
collect dew also gained weight due to leaf guttation. Inside the canopy, the blotting paper could not be
prevented from interception of guttation water. There- fore, the ‘dew’ amount measured inside the canopy
was contaminated by guttation, and only the dew data amount and duration measured on top leaves of the
rice crop were used in the quantitative data analysis.
The time of cover removal was different from night to night. In order to compare the guttation amount be-
tween different nights, the mean guttation rate for each
306 W. Luo, J. Goudriaan Agricultural and Forest Meteorology 104 2000 303–313
night was estimated by dividing the weight increment of the blotting paper at the moment of cover removal
by the time elapsed since sunset. The total amount of guttation water intercepted by the blotting paper in-
side the canopy was calculated as this mean guttation rate multiplied by night length 12 h.
Interception of guttation drops is subject to stochas- tic variability which can be estimated from the obser-
vation data themselves. Intercepted water is not the same as the true quantity of guttation, and rather it
tends to be an underestimate for the following rea- sons. In the first place, only drops beyond a certain
size will get detached and fall. Secondly, the vertically projected area of the blotting paper is what counts,
and thirdly only water from the leaf area above the blotting paper will be collected. Only if a leaf physi-
cally touches the blotting paper, may it directly deliver water, thereby raising the collected amount.
3. Results and discussion
