I. INTRODUCTION
1.1. Background
Interception is a process where a part of rain water retained on a vegetated canopy
surface and later evaporated back to the atmosphere. Interception plays an important
role on rainfall redistribution both spatially and temporally. During rainfall event, the
canopy acts like a sponge for the rain water, after rain water fills the storage capacity
then additional rain water could drip to the ground surface. This storage capacity is the
function of canopy and rainfall characteristics. For a similar depth of gross
rainfall that hit a canopy, a low intensity of rainfall would be higher intercepted than a
high intensity of rainfall. More rain water would also be intercepted in an intermittent
rainfall than in a continuous rainfall. This is because the rain water reaches to the canopy
in a longer period, providing more time for the evaporation process.
Interception is calculated for a vegetated area and expressed as the proportion of gross
rainfall. Other studies show a variable result of interception; Jackson 2000 found 10
of rainfall interception at Agroforestry system in Kenya while Schellekens et al.
1999 found 50 of rainfall interception in a lowland tropical rainforest at northeastern
Puerto Rico.
In a watershed, existence of vegetated area is of importance to extent water
availability for the ecosystem and to prevent flood. Knowledge of interception would
leads to a better understanding of hydrology cycle in this area where interception is one
of the processes that reduce the water input. Neglecting interception loss would result in
an overestimation of water input in a catchments area. By taking into account the
interception loss would result in more precise information when assessing a
catchment water balance and later could lead to a better watershed management.
Previous study to assess the biophysical indicators of Cicatih watershed has been
initiated by Pawitan et al. 2006, while this study focuses on assessing the hydrological
function of the watershed which is about forest canopy ability to intercept rainfall.
1.2. Objectives
The objective of this study is to identify the potency of natural montane forest in
rainfall interception.
II. LITERATURE REVIEW
2.1. Spatial variability of gross rainfall
Gross rainfall distribution varies spatially, but it could be assumed similar for
a relatively small area. To anticipate the variability, researchers installed several rain
gauges in their study site such as Jackson 2000 and Chappell et al. 2001 while
Schellekens 1999 used only one rain gauge. Many rain gauges allow researchers
to get a representative data of gross rainfall but required extra effort for the maintenance.
Some researchers put the rain gauge above the forest canopy such as Asdak et al.
1998a and Schellekens 1999 to avoid forest edge effect.
Asdak et al. 1998a put rain gauge 15 m above ground surface when they
investigated the intercepton on tropical rainforest in Central Kalimantan so that the
angle between the rain gauges and the top of nearest trees was greater than 45
o
to ensure no surrounding environment affect the
measurement. Jackson 2000 used several rain gauges to measure gross rainfall during
his interception study on agroforestry system in Kenya. The rain gauges were placed
above the forest canopy and in an open space near the sampling area, and mean
value from the rain gauges then used as the area’s gross rainfall. Schellekens 1999
used a single rain gauge placed above the canopy at 26 m on a scaffolding tower to
obtain the gross rainfall, while Chappell et al. 2001 were using a network of 34 rain
gauges located within 10 km
2
to investigate the spatial structure of above-canopy
rainfall. Ford and Dean 1978 found that in a
calm weather there was no difference between gross rainfall measured above tree
within 20 m of the sampling area and those 300 m away at 3 m lower and 5 m higher
than surface level of the sampling area. But during windy weather, rain gauge place in
the sampling area above tree measured less rainfall than those at 300 m away.
2.2. The role of canopy characteristics in
controlling the net rainfall
Asdak et al. 1998a mentioned that in a relatively small area, throughfall was a
function of the canopy covers and canopy structure. Leyton and Carlisle 1959 found
larger throughfall as distance increase from the tree stem case study of Lawson cypress
stand. It was also mentioned that because of
1
canopy drip, often throughfall was higher than rainfall in the open space. In contras,
Ford and Deans 1978 found largest amount of throughfall in small area close to the tree
stem case study of young Sitka Spruce Plantation. These show that throughfall
variability was large caused by different canopy characteristics. This canopy
variability should be taken into consideration when determining the troughs
andor gauges placement in the sampling plot so that the value could represent as
much variability as existed in the forest stand.
From other interception studies Table 1, throughfall could vary from 50 – 90 of
gross rainfall, which was obviously caused by the canopy variability. A conical canopy
would have less throughfall than the broadleaved one because the thickness of a
conical canopy would make rainfall that hit the top leaves to be dripped to the lower
leaves. More water then required to saturate the whole canopy before the drops could
reach ground surface. In contras, a broadleaved canopy was thin that would
require less water to saturate its canopy and also makes water hit the top canopy surface
could reach the ground surface in a shorter time as drip.
The quantity of water required to saturate a dry canopy referred as canopy capacity.
During storm event, some raindrops hit the canopy, retained temporarily on its surface,
then evaporated back to the atmosphere or flow down as canopy drip. The canopy
would retain variably quantity of water on its surface depends on its structure such as
leaf size, surface character, and the nature of branching. Water would falls from the
canopy as drip after the canopy get saturated, but some wind or turbulence
might interference the retained water causing the canopy to retain less water and
reduce the canopy capacity. Hall 2003 mentioned that for coniferous trees which
have large leaf area index LAI but small canopy size, the wetting of the canopy
determined by drip from upper leaves while for broadleaved trees with small LAI but
large canopy size, raindrop size significantly influence the canopy capacity.
Canopy capacity did not have a certain interval of values because it was depend on
the canopy characteristics. As shown in Table 2, canopy capacity from forest stands
could vary between 1 – 3.5 mm but there was a possibility that in other canopy this
value could be less or more. Some forests which were dominated by broadleaved trees
might have lower canopy capacity than the conical one and a single layer canopy would
also have lower canopy capacity than the multi-storey one. It because in a broadleaved
and single layer canopy, more canopy cover was exposed to the rainfall area so that less
rainfall was required to saturate the whole canopy while in a conical and a multi-storey
canopy, the same quantity of rainfall might only saturate the top leaves of the canopy
but not the lower ones.
Jarvis 2000 found that epiphytes existence did add the total interception on a
forest stand, but he could not mention the proportion of interception contributed by
those epiphytes.
Despite raindrops that hit the canopy surface, other raindrops could reach the
ground surface directly without ever touching the canopy, identified as free
throughfall. The proportion of canopy gaps that allow rain water to reach ground surface
without hitting the canopy first identified as canopy porosity. Its value varies between 0 -
1. High canopy porosity would indicate large gaps existence in the forest canopy that
allows large quantity of rain water falls as free throughfall. Hall 2003 mentioned that
free throughfall contributes to the net rainfall that occurred in a short time lag with
the gross rainfall. This means that when a forest has larger canopy porosity then the
time lag between gross and net rainfall at the beginning of a rainfall event would be
shorter because of the free throughfall contribution.
Another contribution to the net rainfall was from stemflow; part of raindrops that
flow on the tree branches, accumulated on the trunk and finally reach the ground
surface. Ford and Deans 1978 mentioned that stemflow was positively correlated with
the crown projected area. Their study on young Sitka Spruce plantation found largest
stemflow volume on tree with largest crown projected area. Both bark texture and height
of the stem will also influence the quantity of stemflow reaching the ground surface
Jackson 1975. The thicker and coarser the bark, more water was needed to saturate it
before water can reach the ground surface. This will result in a longer time lag between
rainfall and stemflow occurrence at the beginning of a rainfall event. In agreement
with Ford and Dean 1978, Aboal et al. 1999 found that trees with the largest
2
crown area, not dominated by their neighbors, and with the smoothest bark,
collect the largest volume of stemflow case study of laurel forest.
However, stemflow contributes a relatively small portion to the net rainfall
Table 2. Asdak et al 1998a found stemflow 0.3 of gross rainfall in the
logged forest of Central Kalimantan. In this case, throughfall was controlling the net
rainfall. The sum of throughfall and stemflow
measured identified as the net rainfall, the actual rainfall that reach the ground surface,
and mentioned as the proportion to the gross rainfall. Obviously, canopy characteristics
leaf shape, size, surface character, and the nature of branching did have great
influence on the net rainfall, besides the rainfall distribution itself.
Table 1. Comparative result from other interception studies of gross rainfall
Study Site Location
T
f
S
f
I Source
Unlogged forest Indonesia
87.2 1.4
11.4 Asdak et al. 1998a
Logged forest Indonesia
93.5 0.3
6.2 Asdak et al. 1998a
Agroforestry system with no intercropped
Kenya 88.4 0.7
10.9 Jackson
2000 Hardwood stand
Canada 76.4
4.3 19.3
Carlyle-Moses et al. 1999 Natural forest
Indonesia 72.3
1.7 26.1
Anwar 2003 Lowland tropical
rainforest Puerto Rico
- -
50 Schellekens et al. 1999
Tropical upland mixed cropping system
Indonesia 24 188
Van Dijk and Bruijnzel 2001
Laurel forest Canary
Islands -
6.85 -
Aboal et al. 1999 Tropical montane
cloud forest Columbia - -
42 Jarvis 2000
2.3. Influence of canopy characteristics on Interception