72
Detailed Island Risk Assessment in Maldives Volume III: Detailed Island Reports L. Gan – Part 1 DIRAM team Disaster Risk Management Programme UNDP Maldives December 2007

Detailed Island Risk Assessment in Maldives - Department of

  • Upload
    others

  • View
    3

  • Download
    0

Embed Size (px)

Citation preview

Microsoft Word - DIRAM_Vol3_DIRP1_LGan_V2Volume III: Detailed Island Reports
L. Gan – Part 1
December 2007
3.1 General environmental conditions
3.3 Environmental vulnerabilities to natural hazards
3.4 Environmental assets to hazard mitigation
3.5 Predicted environmental impacts from natural hazards
3.6 Findings and recommendations for safe island development
3.7 Recommendations for further study
4. Structural vulnerability and impacts
4.1 House vulnerability
4.4 Functioning impacts
3
1. Geographic background 1.1 Location Gan is located on the eastern rim of Laamu Atoll, at approximately 73° 31' 50"E and 1°
52' 56" N, about 250 km from the nations capital Male’ and 3.5 km from the nearest
airport, Kadhdhoo (Figure 1.1). Gan is the largest island in terms of land area and
population amongst 13 inhabited islands of Laamu atoll. It’s nearest inhabited islands
are Kalhaidhoo (7 km), Mundoo (10 km) and Atoll Capital Fonadhoo (10 km). Gan forms
part of a stretch of 4 islands connected through causeways and bridges and is the
second largest group of islands connected in this manner with a combined land area of
9.4km2. The island is exposed to NE monsoon generated winds and waves, and
occasional storm activities originating from the cyclone belt of Indian Ocean. Gan is also
believed to be located in an area where offshore ocean bathymetry could create a
‘funnelling’ effect due to wave refraction during tsunami events originating from
Sumatran Ridge (Shifaz, 2004).
4
1.2 Physical environment
Gan is the largest island in the Maldives with a surface area of 600 Ha (6 km2). It has a
length of 7.2km and a width of 1.5km at its widest point. The island is wider in the north
(1500m) and narrower in the south (400m). There are three settlements on the island,
Thundi (northeast), Mathimaradhoo (east) and Mukurimagu (south). In additional there is
a zone designated as Industrial Development Zone, which has a number of structures,
located within it. All the settlements are located along the coastline but only the Thundi
settlement is located away from the oceanward coastline. Entire settlement of both the
Mathimaradhoo and Mukurimagu are within 300m of oceanward coastline while Thundi
is located approximately 900m away from it.
Gan has been connected to the adjacent island, Maandhoo through land reclamation.
Together the two islands form a land area of 670 Ha (6.7 km2) and covers 21km of
coastline. In addition, the islands of Kadhoo (airport) and Funadhoo (Atoll Capital) are
connected to Maandhoo Island through causeways and bridges. The total length of the
island group is approximately 16km.
The reef of Gan is a large reef system with a surface area of 4500 Ha (45km2), covering
70% of the eastern rim of Hadhunmathi Atoll and stretching to approximately 29km. The
reef also hosts 5 inhabited islands, an Airport island (Kadhoo), 2 industrial islands and 8
uninhabited islands, totalling a 1220ha (12.2 km2) of land. It is the largest concentration
of land in a single reef and Gan comprises half of its land area.
Gan is oriented slightly in a northeast-southwest direction and is located in the middle of
the reef system. The island is located approximately 250m from the oceanward reefline
and 350m from the lagoonward reefline. The reef system is exposed to wind generated
waves during NE monsoons and long distance swell waves from the southeast Indian
Ocean.
In spite, of its size, Gan is a low lying island with an average height of +0.9m MSL. The
oceanward coastline is long and low, exposing the island abnormal rises in sea level.
Vegetation cover on the island is very high but large tracts of land have been cleared for
agriculture and forestry. There are substantial variations in the topography of the island
including a large wetland area, which plays a major role in the drainage system,
especially during rainfall and ocean induced flooding events.
5
The proportion of Gan developed for human settlement is small. However, the
impact of human settlement can be found throughout natural environment of the
island. Parts of the natural environment have been modified to meet the
development requirements of the settlement and the atoll population. Terrestrial
modifications have been undertaken around the entire island for agricultural
development, while coastal modifications have mainly been undertaken in three
main points along the western shoreline which nonetheless have contributed to
change coastal processes around Gan. Low areas within the island have been
settled without proper levelling, leading to flooding in some of these areas.
6
2. Natural hazards
This section provides the assessment of natural hazard exposure in Feydhoo Island. A
severe event history is reconstructed and the main natural hazards are discussed in
detail. The final two sections provide the hazard scenarios and hazard zone maps which
are used by the other components of this study as a major input.
2.1 Historic events The island of Gan has been exposed to multiple hazards in the past although its
exposure has been limited. A natural hazard event history was reconstructed for the
island based on known historical events. As highlighted in methodology section, this was
achieved using field interviews and historical records review. Table 2.1 below lists the
known events and a summary of their impacts on the island.
The historic hazardous events for Gan showed that the island faced the following
multiple hazards: 1) flooding caused by heavy rainfall and 2) swell surges, 3) windstorms
and 4) tsunami. Impacts and frequency of these events vary significantly. Flooding
caused by rainfall is the most commonly occurring hazard events. Windstorms have also
been reported as frequent especially during the southwest monsoon. Swell surges have
been reported as infrequent and as having little impact.
Table 2.1. Known historic hazard events of Gan Metrological hazard
Dates of the recorded events
Impacts
Events commonly occurring during SW monsoon.
There are areas in the 3 settlements (Thundi, Mathimaradhoo and Mukurimagu) which are prone to rainfall flooding. All these settlements have wetland areas in close proximity to the settlement. As settlements expand to the low areas exposure to flooding becomes imminent. Impacts from these events are usually minor with damage to household goods and disruption to daily activities such as businesses and schools.
Flooding caused by swell surges
• 1950’s (exact date unknown)
• 5 July 1966
There was one major flooding event reported for Gan, which is dated back to 1950’s. Exact date is not known, but residents say there were reports of fish near the northern wetland area, which is located 400m inland. No substantial
7
Windstorms • 11 July 19661
• 5 May 1977
• 12 May 1978
• 28 Sept 1984
No major recent events have been reported. Written records show damage to vegetation and crops. Little damage to property was reported.
Droughts No major event have been reported
Earthquake No major event have been reported
Tsunami 26th Dec 2004 At least 70% of the island was flooded during the tsunami of 2004. Flood heights were recorded at 2.0m (maximum). Flood heights and their distances in Mathimaradhoo are as following 2.0m – at a distance of 30m from
shoreline 1.5m at a distance of 100m from
shoreline 1.0m at a distance of 150m from
shoreline less than 0.5m – at a distance between
300m and 600m from shoreline The primary reason for tsunami inundation may be found in the low ridge of the island and the presence of very low areas towards the centre of the island.
2.2 Major hazards Based on the historical records, meteorological records, field assessment and Risk
Assessment Report of Maldives (UNDP, 2006) the following meteorological, oceanic and
geological hazards have been identified for Viligilli.
• Heavy rainfall (flooding)
• Windstorms
• Tsunami
• Earthquakes
• Climate Change
1 All dates in italics are adopted from MANIKU, H. A. (1990) Changes in the Topography of Maldives,
Male', Forum of Writers on Environment of Maldives. And news paper reports.
8
2.2.1 Swell Waves and Wind Waves
Being located on the eastern rim of Laamu atoll, Gan is relatively protected from
the year round swell waves approaching from a west to southerly direction. There
are no specific wave studies undertaken for Gan, but studies undertaken around
the country reports a predominantly southwest to a southerly direction for swell
waves (Kench et. al (2006), Young (1999), DHI(1999) and Binnie Black & Veatch
(2000)). A similar pattern could be expected for swell waves reaching Laamu
Atoll. Laamu atoll is also one of the most closed atolls in Maldives with only 6
major reef passes, 5 of which are narrower than 600m. The widest channel (4km
wide), is located in the southern rim facing a south easterly direction. Hence the
probability of swell waves, approaching from the southwest, propagating through
the atoll is very limited.
The east and west coastlines of Gan are exposed to wind waves, however.
During the NE monsoon between November and March, the eastern
(oceanward) coastline may receive strong waves. Wave studies done in similar
settings in GA. Viligilli (EDC, 2006), and K.Hulhule’ (Binnie Black & Veatch,
2000) reported wave heights less than 2.0m and with wave periods of 2-4
seconds. The west coast is exposed to wind generated waves during SW
monsoon, originating within the atoll due to the 30 km fetch and usually with
wave heights less than or about 0.5m.
Despite, being located away from the predominant swell wave direction, Gan is
still exposed to abnormal swell waves originating from intense storms in the
southern hemisphere between 73°E and 130°E longitude. Waves generated from
such abnormal events could travel against the predominant swell propagation
patterns in the Indian Ocean (Goda, 1998), causing flooding on the eastern rim
island of Maldives. The historical flood events on the eastern coastline are most
likely to be the result of such waves since the probability of storm surge is low
due to the proximity to the equator.
The occurrence of abnormal swell waves on Gan reef flat is dependent on a
number of factors such as the wave height, location of the original storm event
9
within the South Indian Ocean, tide levels and reef geometry. It is often difficult to
predict occurrence of such abnormal events as there is only a small probability,
even within storm events of similar magnitude, to produce waves capable of
flooding islands.
Based on the current data available it is impossible to link the swell incidents to
the known cyclonic events in the Indian Ocean. Detailed assessment using
synoptic charts of the South Indian Ocean corresponding to major flooding
events are required to delineate any specific trends and exposure thresholds for
Gan from southern swells. Unfortunately this study does not have the resources
and time to undertake such an assessment but is strongly recommended for any
future detailed assessments.
Udha
Flooding is also known to be caused in Gan by a gravity wave phenomenon
known as Udha. These events are common throughout Maldives and especially
in the southern atolls of Maldives. No specific research has been published on
the phenomenon and has locally been accepted as resulting from local wind
waves generated during the onset of southwest monsoon season. The
relationship has probably been derived due to the annual occurrence of the
events during the months of May or June. These events usually impact the
western coastline of the island and are probably caused by a combination of high
tides and strong wind waves. Impacts from udha events are usually restricted to
within 20m of the western coastline. Due to the comparatively high coastal ridges
on the western coastline, the effects of udha incidents are further controlled.
The udha phenomena needs to be further explored based on long term wave and
climatological data of the Indian Ocean. Udha events could prove to be a major
hazard in the face of climate change since these events are very frequents, have
a direct link to climate patterns and sea level.
Processes controlling water levels around Gan
Waves undergo extreme and rapid transformations as they interact with reef crest, which
control the character of hydrodynamic processes on adjacent reef flat. One of the
10
products of such transformations is the water level setup created at the reef edge and
currents generated by the wave setup. Current records made for various studied over
reef flats (Aslam, 2004) have shown low frequency oscillations in the current speed.
These low frequency oscillations in the current speed have been attributed to surf beat,
edge wave and shear waves.
The degree to which wave energy is transformed or "filtered" by the process of wave
breaking on the reef depends on several factors, including overall reef geometry, water
depth at the reef crest, uniformity of depth along and across the reef, width of the reef
flat and depth of the reef flat (Gourlay, 1994, Gourlay, 1996 ).
Strong winds can cause higher incident waves to break on the reef and the sea-level can
rise locally due to shear force of wind on the water surface. The rise in water level due
the shear force of winds and the wave setup created as a result of breaking waves on
the reef edge can produce high water level set up on the reef flat. Similarly surges or
swell waves beyond significant wave heights of 9m can cause water levels to rise 3.0m
on the reef flat (based on (Department of Meteorology, 2007)). When such rises in water
level are combined with high tides there could be strong surges of water across the reef
flat. Due to the low elevation of Gan coastline, such waves have the potential to create
flooding.
Kench and Brander (2006) reported a relationship between wave energy propagation
across a reef flat and, reef width and depth. Using their proposed Reef Energy Window
Index, the percentage of occurrence of gravity wave energy at Gan reef flat is
approximately 40%.
Historical surge related flood impacts
The common flooding area as a result of surges at present on the island is identified to
be on the oceanward (eastern) coastline of the island. The inland extent of flooding is
greatest towards the northern wetland. The reason could be attributed to the
topographically lower elevations and absence of natural ridge system.
11
wave propagation patterns around Gan
Figure 2.1 Historical flood events and probable wave propagation patterns in Gan and its
reef flat.
The highest wave height reported on the island during flooding events was 1.0m (3.0ft).
This height is consistent with flood heights reported from swell or surge related waves in
Maldives.
Future event prediction
It is known that Gan is exposed to abnormal swell waves originating from the Southern
Indian Ocean. Due to its location in the southern half of the country, this should be
12
considered amongst the most serious hazards facing the island. The exposure swell
waves are mainly from south-easterly to southerly direction. There is also a low
probability of storms in Bay of Bengal to generate swell waves. Events beyond these
arcs may not influence Gan or could have reduced impact due to the protection offered
by the southern and western rim of the atoll.
Historic storm events 1945 - 2007
Possible range of swell wave direction in L.Gan: SE to S & NNE to NE
Figure 2.2 Historical storm tracks (1945-2007) and possible direction of swell waves for
Gan Island
At present, it is very difficult to forecast the exact probability of swell hazard event and
their intensities due to the unpredictability of swell events and lack of research into their
impacts on Maldives. However, since the hazard exposure scenario is critical for this
study a tentative exposure scenario has been developed based on the historical events.
In this regard there is a probability of major swell events occurring every 15 years in Gan
with probable water heights (on land) of 1.0m and every 8 years with probable water
heights of 0.5-0.75m. Events with water heights less than 0.5m and greater than 0.2m
13
are likely to occur once every 5years. The timing of swell events is expected to be
predominantly between April to October, based on historic events and storm event
patterns (see Table 2.2).
Table 2.2 Variation of Severe storm events in South Indian Ocean between 1999 & 2003 (source: (Buckley and Leslie (2004)) Severe wind event
variation
30 °E to 39 °E 12.5 17
40 °E to 49 °E 7.5 10
50 °E to 59 °E 7.5 26
60 °E to 69 °E 6 14
70 °E to 79 °E 6 6
80 °E to 89 °E 12 6
90 °E to 99 °E 12 8
100 °E to 109 °E 8 3
110 °E to 119 °E 15 7
120 °E to 130 °E 13.5 2
The intensity of flooding in the inland areas may have been increased by improper
wetland reclamation. The reclaimed areas are considerably lower than the existing
island causing flood water to run-off towards the island more frequently.
2.2.2 Heavy Rainfall
The rainfall pattern in the Maldives is largely controlled by the Indian Ocean monsoons.
Generally the NE monsoon is dryer than the SW monsoon. Rainfall data from the three
main meteorological stations, HDh Hanimaadhoo, K. Hulhule and S Gan shows an
increasing average rainfall from the northern regions to the southern regions of the
country (Figure 2.3). The average rainfall at S Gan is approximately 481mm more than
that at HDh Hanimadhoo.
14
0
500
1000
1500
2000
2500
3000
3500
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003
Year
Figure 2.3 Mean annual rainfall across the Maldives archipelago.
The closest meteorological station to L.Gan is Kadhoo airport which became operational
in 1986. Unfortunately this study does not have access to Kadhoo data. Moreover,
Kadhoo data may be limited for long term trend observation due smaller number of
detailed observation years. Hence, to resolve the issue, data from Hulhule’ has been
used. It is recommended that further assessment be made once Kadhoo data becomes
available.
The mean annual rainfall of Hulhule’ is 1991.5mm with a Standard Deviation of 316.4mm
and the mean monthly rainfall is 191.6mm. Rainfall varies throughout the year with mean
highest rainfall during October, December and May and lowest between February and
April (See Figure 2.4 below).
15
Figure 2.4 Mean Monthly Rainfall in Hulhule’(1975-2004).
Historic records of rainfall related flooding on the island of Gan indicates that this island
is often flooded and its intensity is high in certain areas of the island. Records for all
incidents have not been kept but interviews with locals and research into newspaper
reports show that localised levels of flooding within sections of Thundi, Mathimaradhoo
and Mukurimagu. These areas usually correspond to wetland edges in Thundi and
Mathimaradhoo settlement, and reclaimed wetland areas in Mukurimagu settlement.
Moreover, substantial topographic variations exist within the Gan Island, as is common
on larger islands of Maldives. Settlement expansion into and along the edges of these
low lying areas have exposed them to flood impact. Furthermore, to remedy flooding on
roads, they were levelled and relevelled with extra sand without considering the flooding
implications for surrounding houses. At present some houses are about 0.3m lower than
the adjacent roads in all three settlements. With no artificial drainage system for the
roads, the surrounding houses in the low areas are at constant risk of flooding. Heavy
rainfall related flooding has been reported to reach up to 0.35m above the ground level
in Thundi and Mukurimagu. In addition, construction of the ‘main road’ along the length
of the island has caused blockages for water runoff towards the northern wetland areas.
As a result the areas on either side of the main road are usually flooded during heavy
rainfall.
16
The impacts of flooding so far reported has not been disastrous, but has had continued
impacts on the community such as damage to personal belongings, crops and
disruptions to daily life.
It would be possible to identify threshold levels for heavy rainfall for a single day that
could cause flooding in Gan, through observation of daily rainfall data in Kadhoo.
Unfortunately, we were unable to acquire daily historical data. However, available limited
severe weather reports shows that Kadhoo received a maximum precipitation of
110.8mm for a 24 hour period on 21th November 2004 (DoM, 2005). Based on interviews
with locals, this event caused minor to moderate levels of flooding in all three
settlements. Damages in Thundi and Mukurimagu settlements were reported for
personal property, backyard crops and some open field crops. Flood heights in the
northern half was reported at 0.2-0.35m. The worst affected area was the southern and
western part of the Thundi, western part of Mathimaradhoo and northern part of
Mukurimagu island, at low lying areas close to wetlands. Schools in Thundi Island were
closed due to flood waters. Similarly an event during January 2003 caused 78.mm of
rainfall during a 24 hour period and led minor damages to personal property in
Mukurimagu island.
The probable maximum precipitations predicted for Hulhule’ and S.Gan by UNDP (2006)
are as follows:
Table 2.3 Probable Maximum Precipitation for various Return periods in Hulhule’ and Gan Station Return Period
50 year 100 year 200 year 500 year
Hulhule’ 187.4 203.6 219.8 241.1 Gan 218.1 238.1 258.1 284.4
Given the high variations in rainfall in Kadhoo, these figures may vary. Based on the field
observations and correlations with severe weather reports from Department of
Meteorology ((DoM, 2005) the following threshold levels were identified for flooding.
These figures must be revised once historical daily rainfall data becomes available
(Table 2.4).
Table 2.4 Threshold levels for rainfall related flooding in Gan Threshold level (daily rainfall)
Impact
17
50mm Puddles on road, flooding in low houses, occasional minor damage to household goods in most vulnerable locations, disruption to businesses and primary school in low areas.
100mm Moderate flooding in low houses; all low lying roads flooded; moderate damage to household items especially in the backyard areas
150mm Widespread flooding on roads and low lying houses. Moderate to major damage to household goods, School closure.
200mm Widespread flooding on roads and houses. Major damages to household goods, sewerage network, backyard crops, School closure, gullies created along shoreline, possible damage to road infrastructure.
230+mm Widespread flooding around the island. Major damages to household goods and housing structure, schools closed, businesses closed, damage to crops, damage to road infrastructure, sewerage network and quay wall.
Quite often heavy rainfall is associated with multiple hazards especially strong winds and
possible swell waves. It is therefore likely that a major rainfall event could inflict far more
damages those identified in the table.
2.2.3 Wind storms and cyclones
Maldives being located within the equatorial region of the Indian Ocean is generally free
from cyclonic activity (Figure 2.5). There have only been a few cyclonic strength
depressions that have tracked through the Maldives, all which occurred in the northern
and north central regions. According to the hazard risk assessment report (UNDP, 2006)
Gan falls within the second least hazardous zone for cyclone related hazards and has a
maximum predicted cyclonic wind speeds of 56 Kts (see figure below). There are no
such records for the southern region, although a number of gale force winds have been
recorded due to low depressions in the region. Winds exceeding 35 knots (gale to strong
gale winds) were reported as individual events in Kadhoo annually between 2002 and
2006, all caused by known low pressure systems near Maldives rather than the
monsoon (DoM, 2005). The maximum wind speed in Kadhoo during this period was
approximately 46 kts.
probable maximum cyclone wind speed (kts)
Figure 2.5 Cyclone hazard zones of the Maldives as defined by UNDP (2006).
Historic records for Gan have indicated that near gale force winds (see Table 2.5) have
caused minor damage to property and trees on the island. Hence during the high winds
between 2002 and 2005, a number of minor to moderate damages were reported to
vegetation and backyard crops. Gan does have lush vegetation dominated by larger
trees species, which acts to minimise the direct exposure of properties.
In order to perform a probability analysis of strong wind and threshold levels for damage,
daily wind data is crucial. However, such data was unavailable for this study.
The threshold levels for damage are predicted based on interviews with locals and
housing structural assessments provided by risk assessment report (UNDP, 2006), as
summarized in Table 2.6.
Table 2.5 Beaufort scale and the categorisation of wind speeds
Beau- fort No Description Cyclone
category
Specifications for estimating speed over land
0 Calm Less than 1 less than 1 Calm, smoke rises vertically.
1 Light Air 1 -3 1 - 5
Direction of wind shown by smoke drift, but not by wind
vanes.
Wind felt on face; leaves rustle; ordinary wind vane moved
by wind.
Leaves and small twigs in constant motion; wind extends
light flag.
4
Moderate
breeze 11 - 16 20 - 28 Raises dust and loose paper; small branches moved.
5 Fresh breeze 17 -21 29 - 38
Small trees in leaf begin to sway; crested wavelets form on
inland waters.
Large branches in motion; whistling heard in telegraph
wires; umbrellas used with difficulty.
7 Near gale 28 - 33 50 - 61
Whole trees in motion; inconvenience felt when walking
against the wind.
8 Gale Category 1 34 - 40 62 - 74 Breaks twigs off trees; generally impedes progress.
9 Strong gale Category 1 41 - 47 75 - 88
Slight structural damage occurs (chimney pots and slates
removed).
Seldom experienced inland; trees uprooted; considerable
structural damage occurs.
Very rarely experienced; accompanied by widespread
damage.
12 Hurricane Category 3,4,5 64 and over 118 and over Severe and extensive damage. Table 2.6 Threshold levels for wind damage based on interviews with locals and available meteorological data Wind speeds Impact 1-10 knots No Damage 11 – 16 knots No Damage 17 – 21 knots Light damage to trees and crops 22 – 28 knots Breaking branches and minor damage to
open crops, some weak roofs damaged 28 – 33 knots Minor damage to open crops and vegetation 34 - 40 knots Minor to Moderate to major damage to
houses, crops and trees 40+ Knots Moderate to Major damage to houses, trees
falling, crops damaged
2.2.4 Tsunami
UNDP (2006) reported the region where Gan is geographically located to be a very high
tsunami hazard zone. The tsunami of December 2004 had devastated a number of
islands in the eastern rim of Laamu atoll along with parts of Gan. According to the
official estimates, 50% of the island was flooded during this event. Field surveys and
aerial photographs immediately after the event revealed that approximately 70% of the
island was flooded. Flood waters travelled approximately 1km inland in the northern half
while much of the southern end was flushed from east to west. Hence, all the 3
20
settlements were flooded. The settlements of Mathimaradhoo and Mukurimagu along
with the ‘industrial zone’ were entirely flooded. Flooding in Thundi settlement was limited
to 30%. The significantly high exposure of Mathimaradhoo and Mukurimagu are due to
the close proximity to the oceanward coastline.
There were extensive damage to properties in Mathimaradhoo and Mukurumagu and a
significant percentage of the population of the island lost much of their livelihood due to
the damage to crops and businesses. The tsunami run-up height at the eastern
shoreline of the island was reported to be approximately 4m above MSL reducing to
0.3m inland. The severest damage to the houses and structures were limited within
approximately 150m from the eastern shoreline. The decay of the flood water for this
tsunami showed a logarithmic decay function. Tsunami induced tide level within the
lagoon predicted using the tide data from the nearest tide station at Hulhule’ shows why
the island was not flooded from its lagoonward side (Fig 2.6 and Fig 2.7). It is evident
that the tide level within the atoll lagoon did not rise above the elevation of the island.
200 400 600 800 1000 1200 1400 1600m
P1P2 P3
(December 2004 tsunami)
Distance from oceanward shoreline
m )
Fig 2.6. Maximum water level caused by tsunami of December 2004 plotted across the island profile of Gan near Thundi settlement evidently showing the reason why the island did not get flooded from the lagoonward side. Graph also shows the logarithmically decaying flood water level.
21
Fig 2.7. Water level recordings from the tide gauge at Hulhule’ indicating the wave height of tsunami 2004 (source: University of Hawai’i Sea Level Centre, http://ilikai.soest.hawaii.edu/uhslc/iot1d/male1.html) Comparatively higher exposure of Laamu Atoll may be partially due to the refraction of
the wave caused by the Indian Ocean bathymetry as it travelled westwards Maldives
(Ali, 2005). The Indian Ocean bathymetry (Fig 2.8) shows shallower water depths
extended far offshore at around the central region of the Maldives (at around the atolls of
Laamu – Meemu). This shallower area caused the wave to bend away from the
southern atolls and became focused towards the central region of the country. It is likely
that a similar pattern may persist in any future event if the waves originate from the
northern Sundra trench.
22
Fig 2.8. Submarine topography around Maldives archipelago and modelled wave refraction for the December 2004 tsunami (source: Ali (2005)). The predicted probable maximum tsunami wave height for the area where Gan is
located is 3.2 – 4.5m (UNDP, 2006). Examination of the flooding that will be caused by
a wave run-up of 4.5m for the island of Gan indicates that such a magnitude wave will
flood the entire island from coast to coast. The first 150-200m from the shoreline will be
a severely destructive zone (Fig 2.9). The theoretical tsunami flood decay curve was
plotted for a wave that is applied only for the direct wave from the oceanward side of the
island. It also is well understood that the tsunami wave will also travel into the atoll
lagoon which will cause the water level in the atoll lagoon to rise. This could cause
flooding of the island from the lagoonward side of the island, if the water level rises
above the height of the island. The maximum tsunami wave induced water level height
predicted for the atoll lagoon near Gan is 1.7m. This could flood the island of Gan not
23
just from the oceanward side of the island but also from the lagoonward side and the
entire island will be flooded.
200 400 600 800 1000 1200 1400 1600m
P1P2 P3
Distance from oceanward shoreline
Threshold level for flooding for severe
strucutral damage
Fig 2.9. Tsunami induced tide level within the atoll lagoon Gan will flood the entire island. Graph also shows the flooding decay curve and maximum impact zone for the maximum predicted tsunami at Gan.
2.2.5 Earthquakes
There hasn’t been any major earthquake related incident recorded in the history of Gan
or even Maldives. However, there have been a number of anecdotally reported tremors
around the country.
The Disaster Risk Assessment Report (UNDP 2006) highlighted that Laamu Atoll is
geographically located in the highest seismic hazard zone rated 2 out of 5, based on the
entire country. According to the report the rate of decay of peak ground acceleration
(PGA) for the zone 2 in which Gan is located has a value less than 0.05 for a 475 years
return period (see table below). PGA values provided in the report have been converted
to Modified Mercalli Intensity (MMI) scale (see column ‘MMI’ in table 3.9 table below).
The MMI is a measure of the local damage potential of the earthquake. See table 3.10
for the range of damages for specific MMI values. Limited studies have been performed
to determine the correlation between structural damage and ground motion in the region.
The conversion used here is based on United States Geological Survey findings. No
24
attempt has been made to individually model the exposure of Gan Island as time was
limited for such a detailed assessment. Instead, the findings of UNDP (2006) were used.
Table 2.7 Probable maximum PGA values in each seismic hazard zone of Maldives (modified from UNDP, 2006). Seismic hazard zone
PGA values for 475yrs return period
MMI2
1 < 0.04 I 2 0.04 – 0.05 I 3 0.05 – 0.07 I 4 0.07 – 0.18 I-II 5 0.18 – 0.32 II-III
Table 2.8 Modified Mercalli Intensity description (Richter, 1958).
MMI Value
Shaking Severity
Description of Damage
I Low Not felt. Marginal and long period effects of large earthquakes.
II Low Felt by persons at rest, on upper floors, or favourably placed.
III Low Felt indoors. Hanging objects swing. Vibration like passing of light trucks. Duration estimated. May not be recognized as an earthquake.
IV Low Hanging objects swing. Vibration like passing of heavy trucks; or sensation of a jolt like a heavy ball striking the walls. Standing motor cars rock. Windows, dishes, doors rattle. Glasses clink. Crockery clashes. In the upper range of IV, wooden walls and frame creak.
V Low Felt outdoors; direction estimated. Sleepers wakened. Liquids disturbed, some spilled. Small unstable objects displaced or upset. Doors swing, close, open. Shutters, pictures move. Pendulum clocks stop, start, change rate.
VI-XII Light - Catastrophe
Light to total destruction
According to these findings the threshold for damage is very limited even in a 475 year
return earthquake. It should however be noted that the actual damage may be different
in Maldives since the masonry and structural stability factors have not been considered
at local level for the MMI values presented here. Usually such adjustments can only be
accurately made using historical events, which is almost nonexistent in Maldives.
2.2.6 Climate Change
2 Based on KATZFEY, J. J. & MCINNES, K. L. (1996) GCM simulation of eastern Australian cutoff lows.
Journal of Climate, 2337-2355.
25
The debate on climate change, especially Sea Level Rise (SLR) is far from complete.
Questions have been raised about SLR itself (Morner et al., 2004, Morner, 2004) and
the potential for coral island environments to naturally adapt (Kench et al., 2005,
Woodroffe, 1993). However the majority view of the scientific community is that climate
is changing and that these changes are more likely to have far reaching consequences
for Maldives. For a country like Maldives, who are most at risk from any climate change
impacts, it is important to consider a cautious approach in planning by considering worst
case scenarios. The findings presented in this section are based on existing literature.
No attempt has been made to undertake detailed modelling of climate change impacts
specifically on the island due to time limitations. Hence, the projection could change with
new findings and should be constantly reviewed.
The most critical driver for future hazard exposure in Maldives is the predicted sea level
rise and Sea Surface Temperature (SST) rise. Khan et al. (2002, Woodroffe, 1993)
analysis of tidal data for Gan, Addu Atoll shows the overall trend of Mean Tidal Level
(MTL) is increasing in the southern atolls of Maldives. Their analysis shows an
increasing annual MTL at Gan of 3.9 mm/year. These findings have also been backed
by a slightly higher increase reported for Diego Garcia south of Addu Atoll (Sheppard,
2002). These calculations are higher than the average annual rate of 5.0 mm forecasted
by IPCC (2001), but IPCC does predict a likely acceleration as time passes. Hence, this
indicates that the MTL at Gan by 2100 will be nearly 0.4m above the present day MTL.
Similarly, Khan et al. (2002) reported air temperature at Addu Atoll is expected to rise at
a rate of 0.4C per year, while the rate of rise in SST is 0.3C. Although no specific studies
have been done for Laamu Atoll, the findings from Addu Atoll could be used as a guide
to predicted changes.
Predicted changes in extreme wind gusts related to climate change assumes that
maximum wind gusts will increase by 2.5, 5 and 10 per cent per degree of global
warming (Hay, 2006). Application of the rate of rise of SST to the best case assumption
indicates a 15% increase in the maximum wind gusts by the year 2010 in southern
Atolls.
The global circulation models predict an enhanced hydrological cycle and an increase in
the mean rainfall over most of the Asia. It is therefore evident that the probability of
occurrence and intensity of rainfall related flood hazards for the island of Gan will be
26
increased in the future. It has also been reported that a warmer future climate as
predicted by the climate change scenarios will cause a greater variability in the Indian
monsoon, thus increasing the chances of extreme dry and wet monsoon seasons (Giorgi
and Francisco, 2000). Global circulation models have predicted average precipitation in
tropical south Asia, where the Maldives archipelago lies, to increase at a rate of 0.14%
per year (Figure 2.10).
0
2
4
6
8
10
12
Year
% )
Fig 2.10 Graph showing the rate of increase of averaged annual mean precipitation in
tropical south Asia (Adger et al., 2004)
There are no conclusive agreements over the increase in frequency and intensity of
Southern Indian Ocean Storms. However, some researchers have reported a possible
increase in intensity and even a northward migration of the southern hemisphere storm
belt (Kitoh et al., 1997) due rise in Sea Surface Temperatures (SST) and Sea Level
Rise. If this is to happen in the Southern Indian Ocean, the frequency of and intensity of
storms reaching Gan Island coastline will increase and thereby exposing the island more
frequent damages from swell waves. The increase in sea level rise will also cause the
storms to be more intense with higher flood heights.
The above discussed predicted climate changes for Gan and surrounding region is
summarised below in Table 2.9. It should be cautioned that the values are estimates
based on most recent available literature on Maldives which themselves have a number
of uncertainties and possible errors. Hence, the values should only be taken as guide as
it existed in 2006 and should be constantly reviewed. The first three elements are based
climate change drivers while the bottom three are climatological consequences.
27
Table 2.9 Summary of climate change related parameters for various hazards. Element Predicted
rate of
SLR 3.9-5.0mm /yr
Yr 2050: +0.2m
Yr 2100: +0.4m
Yr 2050: +0.4m
Yr 2100: +0.88m
Air Temp 0.4°C / decade
Yr 2050: +1.72°
Yr 2100: +3.72°
Yr 2050: +1.29°
Yr 2100: +2.79°
Increase in storm surges and swell wave related flooding, Coral bleaching & reduction in coral defences
Rainfall +0.14% / yr (or +32mm/yr)
Yr 2050: +1384mm
Yr 2100: +2993mm
Wind gusts 5% and 10% / degree of warming
Yr 2050: +3.8 Knots
Yr 2100: +8.3 Knots
Swell Waves
Increase in swell wave related flooding.
2.3 Event Scenarios
Based on the discussion provided in section 2.2 above, the following event scenarios
have been estimated for Gan Island (Tables 2.10-12).
28
Hazard Max
Low Moderat e
Swell Waves
(wave heights on reef flat – Average Island ridge height +1.7m above reef flat)
NA < 2.0m
4.5m < 2.0m
Heavy Rainfall
284mm <60m m
> 60mm >175m m
High Moderate Low
Table 2.11 Slow onset flooding hazards (medium term scenario – year 2050)
1.0.1. H
1.0.4. 1.0.5. L
1.0.12. < 2.0m
1.0.20. < 2.0m
1.0.28. < 60mm
1.0.29. > 60mm
1.0.30. > 175mm
Hazard Max
Low Moderate Severe Low Moderate Severe
Wind storm NA <30 knts
> 30 knts > 45Knts
Unlikely none
2.4 Hazard zones Hazard zones have been developed using a hazard intensity index. The index is based
on a number of variables, namely historical records, topography, reef geomorphology,
vegetation characteristics, existing mitigation measures and hazard impact threshold
levels. The index ranges from 0 to 5 where 0 is considered as no impact and 5 is
considered as very severe. In order to standardise the hazard zone for use in other
components of this study only events above the severe threshold were considered.
Hence, the hazard zones should be interpreted with reference to the hazard scenarios
identified above.
2.4.1 Swell waves and SW monsoon high Waves
Swell waves higher than 3.0m on reef flat are predicted to reach the eastern coastline of
the island. These waves may penetrate 200 to 700m inland (Figure 2.11). The wave
height on coastline is on average estimated to be 1.0 m and with rapid decline as it
moves inland. The runoff on to the island is facilitated by the low topography towards the
centre and due to the absence of high coastal ridges along much of the coastline (sees
physical environment section).
The western side of the island is relatively protected due to the cumulative effects of
higher elevation of the area and the lower drainage basin on the east. Effects on the
western side may be felt from wind waves and high seas (Udha), but will be limited to
20-50m from the coastline.
30
Amongst the individual settlements, the Thundi settlement is relatively protected from
impacts of swell waves due to the presence of a 1000m wide uninhabited land on
eastern side. The settlements of Mathimaradhoo, Mukurimagu and the industrial zone
remain exposed due to their location close to the eastern coastline. Structures within the
first 100m of the coastline along Mathimaradhoo and Mukurimagu are particularly
exposed to severe intensity should an event in the severe category strike. Moreover the
entire footprint of these two settlements is exposed to varied intensities from a severe
category event.
1,000
Mathimaradhoo
Thundi
Contour lines represent intensity index based on a severe event scenario (+3.0m on reef flat &
+1.3m to +0.3m on land)
Mukurimagu
Figure 2.11 Hazard zoning map for swell waves and southwest monsoon high seas.
31
2.4.2 Tsunamis
When a severe threshold of tsunami hazard (>3.0m on reef flat) is considered, the entire
island predicted to be effected (Figure 2.12). If the waves reach beyond 4.0m on reef flat
the entire island is highly likely to be flooded due the prevalent tide levels. High intensity
waves will flush through the island from the eastern side while tide related surges will
occur within the atoll, flooding the western coastline. The intensity of flood waters will be
highest 200-250m from the shoreline. Intensity could also be high up to 500m inland
owing to the downward slope existing along the length of the island, 300m from the
eastern coastline. Impact beyond 500m is still considered to be moderate considering
the possible surge from atoll lagoon due to rise in tide level.
The effected zone is dependent on the distance from coastline and minor variations in
topography as it advances inland. Wave height around the island will vary based on the
original tsunami wave height, but the areas marked as low intensity is predicted to have
proportionally lower heights compared to the coastline.
32
Low
Mukurimagu
Mathimaradhoo
Thundi
metres
Intensity Index
Industrial Zone
Figure 2.12 Hazard zoning map for tsunami flooding.
The settlements of Mathimaradhoo and Mukurimagu will receive high intensity waves
due its proximity to coastline. The entire settlement foot prints of these two settlements
are considered a hazard zone for severe category tsunamis. The settlement of Thundi is
protected from the direct impact of the waves although parts of the settlement could
experience strong wave flushing form the north east. Thundi is predicted to be mostly
exposed to flooding caused by rise in tsunami related. Impacts will therefore be lower
than the settlements on the eastern coastline.
33
2.4.3 Heavy Rainfall
Heavy rainfall above the severe threshold is expected to flood parts of all three
settlements (Figure 2.13). The areas predicted for severe intensity are the wetland areas
and the topographic lows on the eastern half of the island. These areas act as drainage
basins for the surrounding higher areas and due the large size of the island the
‘catchment area’ is considerable for surface runoff during heavy rainfall.
In Thundi settlement the high intensity zone is expected to be located in the southern
and western areas, especially where the new resettlement project is being undertaken.
The natural depression and occasional wetlands patches in this area performs a
drainage function for this section of the island. Similarly, the central wetland area close
to the Mathimaradhoo settlement is also expected to lead to flooding on the western
parts of the settlement. Similar topographic lows and old wetland areas in the
Mukurimagu settlement will also cause moderate to severe flooding.
34
Mukurimagu
Intensity Index
Contour lines represent intensity index based on a severe event
scenario (+3.0m to +0.5m on land)
Note: White areas represent areas
with no data
Figure 2.13 Hazard zoning map for heavy rainfall related flooding. The rainfall hazard zones are approximate and based on the extrapolation of
topographic data collected during field visits. The white areas represent areas with no
field surveys due to poor accessibility. A comprehensive topographic survey is required
before these hazard zones could be accurately established.
2.4.4 Strong Wind
The intensity of the strong wind across the island is expected to remain fairly constant.
Smaller variations may exist between the west and east side where by the west side
receives higher intensity due to the predominant westerly direction of abnormally strong
35
winds. The entire island has been assigned an intensity index of 4 for strong winds
during a severe event.
2.4.5 Earthquakes
The entire island is a hazard zone with equal intensity. An intensity index of 1 has been assigned. 2.4.6 Climate Change
Establishing hazard zones specifically for climate change is impractical at this stage due
to the lack of topographic and bathymetric data. However, the predicted impact patterns
and hazard zones described above are expected to be prevalent with climate change as
well, although the intensity is likely to slightly increase.
2.4.7 Composite Hazard Zones
A composite hazard zone map was produced using a GIS based on the above hazard
zoning and intensity index (Figure 2.14). The coastal zone approximately 200m from the
oceanward coastline and the northern wetland areas are predicted to be the most
intense regions for multiple hazards. The eastern side is particularly identified as a
hazard zone due to the exposure to swell waves, tsunamis and wind damage.
36
0
Intensity Index
metres
Contour lines represent intensity index based on a severe event
scenarios
37
2.5 Limitations and recommendation for future study The main limitation for this study is the incompleteness of the historic data for different
hazardous events. The island authorities do not collect and record the impacts and
dates of these events in a systematic manner. There is no systematic and consistent
format for keeping the records. In addition to the lack of complete historic records there
is no monitoring of coastal and environmental changes caused by anthropogenic
activities such as road maintenance, beach replenishment, causeway building and
reclamation works. It was noted that the island offices do not have the technical
capacity to carryout such monitoring and record keeping exercises. It is therefore evident
that there is an urgent need to increase the capacity of the island offices to collect and
maintain records of hazardous events in a systematic manner.
The second major limitation was the inaccessibility to long-term meteorological data from
the region. Historical meteorological datasets atleast as daily records are critical in
predicting trends and calculating the return periods of events specific to the site. The
inaccessibility was caused by lack of resources to access them after the Department of
Meteorology levied a substantial charge for acquiring the data. The lack of data has
been compensated by borrowing data from alternate internet based resources such as
University of Hawaii Tidal data. A more comprehensive assessment is thus
recommended especially for wind storms and heavy rainfall once high resolution
meteorological data is available.
The future development plans for the island are not finalised. Furthermore the existing
drafts do not have proper documentations explaining the rationale and design criteria’s
and prevailing environmental factors based on which the plan should have been drawn
up. It was hence, impractical to access the future hazard exposure of the island based
on a draft concept plan. It is recommended that this study be extended to include the
impacts of new developments, especially land reclamations, once the plans are finalised.
The meteorological records in Maldives are based on 5 major stations and not at atoll
level or island level. Hence all hazard predictions for Gan are based on regional data
rather than localised data. Often the datasets available are short for accurate long term
prediction. Hence, it should be noted that there would be a high degree of estimation
and the actual hazard events could vary from what is described in this report. However,
the findings are the closest approximation possible based on available data and time,
38
and does represent a detailed although not a comprehensive picture of hazard exposure
in Gan.
References ALI, S. (2005) December 26 2004 Tsunami Impact Assessment and a Tsunami
Risk Assessment of the Maldives. School of Civil Engineering and the Envrionment. Southampton, United Kingdom, University of Southampton, .
BINNIE BLACK & VEATCH (2000) Enviromental / Technical study for dredging / reclamation works under Hulhumale' Project - Final Report. Male', Ministry of Construction and Public Works.
BUCKLEY, B. W. & LESLIE, L. M. (2004) Preliminary climatology and improved modelling of South Indian Ocean and southern ocean mid-latitude cyclones. International Journal of Climatology, 24, 1211-1230.
DEPARTMENT OF METEOROLOGY (2007) The unsually strong swell, tidal waves hit Maldives Islands [sic]. Male', Maldives, Department of Meterorology.
DEPARTMENT OF METEOROLOGY (DOM) (2005) Severe weather events in 2002 2003 and 2004. Accessed 1 November 2005, <http://www.meteorology.gov.mv/default.asp?pd=climate&id=3>, Department of Meteorology, Male', Maldives.
DHI (1999) Physical modelling on wave disturbance and breakwater stability, Fuvahmulah Port Project. Denmark, Port Consult.
ENVIORNMENT AND DREDGING CONSULTANCY (EDC) (2006) Envrionmental Impact Assessment of Construction of Safe Island Viligilli, Gaafu Alifu Atoll, Maldives. Male', Maldives, Ministry of Planning and National Development.
GIORGI, F. & FRANCISCO, R. (2000) Uncertainties in regional climate change prediction: a regional analysis of ensemble simulations with HadCM2 coupled AOGCM. Climate Dynamics, 16, 169-182.
GODA, Y. (1998) Causes of high waves at Maldives in April 1987. Male', Asia Development Bank.
GOURLAY, M. R. (1994) Wave transformation on a coral reef. Coastal Engineering, 23, 17-42.
GOURLAY, M. R. (1996 ) Wave set-up on coral reefs. 2. Set-up on reefs with various profiles. Coastal Engineeting, 28, 17-55.
HAY, J. E. (2006) Climate Risk Profile for the Maldives. Male', Ministry of Envrionment Energy and Water, Maldives.
IPCC (2001) Climate Change 2001: The Scientific Basis, New York, Cambridge, United Kingdom and New York, NY, USA.
KATZFEY, J. J. & MCINNES, K. L. (1996) GCM simulation of eastern Australian cutoff lows. Journal of Climate, 2337-2355.
KENCH, P. S., MCLEAN, R. F. & NICHOL, S. L. (2005) New model of reef-island evolution: Maldives, Indian Ocean. Geology, 33, 145-148.
39
KHAN, T. M. A., QUADIR, D. A., MURTY, T. S., KABIR, A., AKTAR, F. & SARKAR, M. A. (2002) Relative Sea Level Changes in Maldives and Vulnerability of Land Due to abnormal Coastal Inundation. Marine Geodesy, 25, 133–143.
KITOH, A., YUKIMOTO, S., NODA, A. & MOTOI, T. (1997) Simulated changes in the Asian summer monsoon at times of increased atmospheric CO2. Journal of Meteorological Society of Japan, 75, 1019-1031.
MANIKU, H. A. (1990) Changes in the Topography of Maldives, Male', Forum of Writers on Environment of Maldives.
MORNER, N.-A. (2004) The Maldives project: a future free from sea-level flooding. Contemporary South Asia, 13, 149-155.
MORNER, N.-A., TOOLEY, M. & POSSNERT, G. (2004) New perspectives for the future of the Maldives. Global and Planetary Change, 40, 177-182.
RICHTER, C. F. (1958) Elementary Seismology, San Francisco, W.H. Freeman and Company.
SHEPPARD, C. R. C. (2002) Island Elevations, Reef Condition and Sea Level Rise in Atolls of Chagos, British Indian Ocean Territory. IN LINDEN, O., D. SOUTER, D. WILHELMSSON, AND D. OBURA (Ed.) Coral degradation in the Indian Ocean: Status Report 2002. Kalmar, Sweden, CORDIO, Department of Biology and Environmental Science, University of Kalmar.
WOODROFFE, C. D. (1993) Morphology and evolution of reef islands in the Maldives. Proceedings of the 7th International Coral Reef Symposium, 1992. Guam, University of Guam Marine Laboratory.
YOUNG, I. R. (1999) Seasonal variability of the global ocean wind and wave climate. International Journal of Climatology, 19, 931–950.
40
3.1 Environment Settings
3.1.1 Terrestrial Environment
Topography
The topography of Gan was assessed using three island profiles (see Figure 3.1). Given
below are the general findings from this assessment.
The island is generally low lying with an average elevation of +0.9 m MSL along the
surveyed topographic profiles. This finding was reconfirmed from the shallow depths of
ground water table around the island (on average approximately 1 m at median tide).
As characteristic of large islands, considerable variations in topography were observed
in Gan. The island does not have high elevations comparable to other large islands in
Maldives but there are substantial tracts of low areas developed during island formation.
These include a major wetland area in the northern half of the island. As can be seen
from general trends in low areas shown in Figure 3.5, a low depression extends almost
along the length of the island. It has to be noted that the map shows predicted low areas,
based on topographic profiles (see Figure 3.2-4) and general field observations. The
roads used for the assessment, except profile 3 (Figure 2.4) had been modified during
road maintenance. To accommodate this limitation, additional field assessments were
undertaken in surrounding unmodified areas during level surveys.
The low areas dictate the drainage system of the island. During heavy rainfall, low areas
are regularly flooded. Although much of these areas are located outside the settlements,
there are small patches of low areas that fall within the settlements. Hence, parts of
Thundi, Mathimaradhoo and Mukurumagu settlements experience occasional rainfall
related flooding. In general, larger islands are less exposed to the impacts of ocean
induced flooding due their width and the presence of buffer land. However, drainage
system prevalent on Gan and the general low elevation of the island appears to negate
this advantage. At present the low ridges on the oceanward side and the general
gradient towards the centre can cause flood waters around 2.0 m above MSL to flood
half of the island. The tsunami of 2004 appeared to prove this trend.
The coastal ridges of Gan are quite low averaging 1.5 m and increasing in height
southwards. It is interesting to note that the coastline facing northeast was the lowest
41
while the coastline facing southwest was highest. This pattern tends suggest that the
impacts of NE monsoon may not be that prominent and that there is an alternate wave
energy source effecting the southwest coastline. Observations by Nasser (2003) in a
Maldives wide wave energy study tends to suggest that wave energy approaching from
southeast Indian Ocean has a higher wave power and could effect reefs facing
southeast. The generally low elevations along the oceanward coastline also tend to
suggest that the overall wave energy is low. In any case the low ridges expose much of
Gan to the effects of ocean induced flooding.
Mukurimagu
Thundi
Mathimaradhoo
0 °E
42
1m
0
Note: Profile modified due to road maintenance activities
Lagoonward Side
Harbour quaywall
Ridge system generally low (+1.1m)
P1P2 P3
0 200 400 600 800 1000 1200 1400
Main Road
Oceanward Ridge
Note: Profile modified due to road maintenance activities
Lagoonward Side
P2 P3
43
0 50 100 150 200 250 300 350 400 450
1m
0
Elevation +0.7m
Main Road
Elevation +0.5m
Llagoonward Ridge
1,000500
metres
45
The overall vegetation cover in Gan Island is very high compared to other inhabited
islands, primarily due its large size (Figure 3.6). The settlement footprints of the three
main settlements and industrial zone cover just 15% of the total land area. Vegetation
cover within the settlements is low and is mainly restricted to medium sized backyard
fruit trees.
Majority of the vegetation cover on the island comprise of a wide ranging medium to low
species. There are patches of larger trees distributed across the island. Much of the
larger trees have been reported as cleared for forestry. The most prominent patch of
large trees (coconut palms) is located south of Thundi settlement. At present a large
proportion of this patch has been cleared for tsunami resettlement.
The coastal vegetation around the island is dense and well established. Coastal
vegetation around the settlements has been largely depleted, however. Coastal
vegetation in more than 75% of the coastline around the settlement of Mukurimagu has
been cleared while over 50% of Mathimaradhoo Island has been cleared. On average
the coastal vegetation belt is 30m wide within the settlement areas in Mathimaradhoo
and over 80 m wide around the industrial development zone.
Coastal vegetation in the northwest and the northeast corner of the island is relatively
young and sparsely distributed. The tsunami of 2004 seemed to have effected the
coastal vegetation, but has since recovered by 2007.
Ground Water and Soil
Gan Island has a substantial layer of fresh water (MPND, 2005). Water lens depth varies
across the island based on topography. Generally the water table could be reached with
less than 1m at median tide in all areas. This could decrease to 0.5 m during spring high
tides or more during heavy rainfall, especially in low lying areas.
Gan’s ground water is generally in good conditions and no traces of contamination were
reported (MPND, 2005). There were no shortages of potable water in the past due to the
good quality of ground water and availability of rainfall reserves.
The soil conditions were not assessed across the island due to time limitation. Gan is
reported to have some of the most fertile soils in the atoll (MPND, 2005).
46
due to harbour construction
Figure 3.7 Coastal Features
Figure 3.7 summarises the coastal characteristics of Gan Island. The coastline is
predominantly affected by monsoon wind driven waves and long distance swell waves
originating from south east Indian Ocean. On the ocean ward side, wave activity is more
prominent during the north east monsoon. On the lagoon wardside the impact of
47
monsoon wind generated waves are controlled by the closed nature of the reef, but
nonetheless creates enough conditions to generate waves within the atoll. Impacts from
waves originating within the atoll are expected to be limited.
The oceanward coastline does not resemble a high energy coastline, based on the
dominant geomorphic features. Furthermore, the coastline facing north and northeast
was observed to have less exposure to strong wave energy than the stretch of coastline
facing southwest. The net result is a low coastline with the no prominent ridge systems.
The exceptions are the southern half of the oceanward coastline which increases in
height, although still lower than other larger islands like Seenu Atoll Hithadhoo or
Gnaviyani Fuvahmulah.
The lengthy coastline along the oceanward and lagoonward sides has meant that the
effects of longshore drift are prominent. Sediment was observed to be transported along
long distances and was seen to vary across the two monsoon seasons. The lagoonward
coastline has a larger sand budget and tends to move more quantities. This extent of
transport may have been limited due to the presence of solid structures and dredged
areas on both ends of the island.
Gan was observed to be expanding northwards especially on its northeast and
northwest corners. This process has been hampered over the last 4 years due to the
harbour development in Thundi settlement. Large areas are now seen to be eroding
from the north possibly due to sand supplies being restricted during the southwest
monsoon.
Coastal erosion has been reported as a major issue by all three settlements. These
events appear to be seasonal or long-term cyclic changes occurring to the coastline. The
trends and patterns in erosion are difficult to forecast due to the recent introduction of
coastal modifications. Over the past 40 years, 3 ha of land have been eroded from the
island, especially the northern end of the island. During the same period over 2.5 ha of
new land have been added. Hence the net loss is insignificant, considering the size of
the island. However, coastal erosion is predicted to become a major environmental
issue, atleast in the medium-term, due to irreversible changes being brought to the
coastal environment by human activities.
48
General historical changes to reef conditions were assessed anecdotally, through
interviews with a number of fishermen. The general agreement amongst the
interviewees was that the quality of reef areas on both sides of the island has declined
considerably over the past 50 years. Much of these changes were reported to be in the
form of excess sedimentation in the lagoonward side and general coral decline on the
oceanward side. The construction of causeways may have played a role in this reduction
of quality.
Sea grass overgrowth is a major problem on both the lagoonward and oceanward
lagoon. The case of the lagoonward side is more prominent due to the slow currents and
coastal modifications blocking sediment transport. The condition in the southwest corner
of the island has deteriorated over the past few years with foul smell becoming a major
issue.
Coastal Modifications
As noted earlier, much of the coastal modifications have been undertaken on the
eastern shoreline of the island (Figure 3.8). Below is a summary of major
modifications.
• The two islands of Gan and Maandhoo have been joined together through land
reclamation. Water flow through the lagoon pass between the islands has been
halted permanently. There appears to be major localised changes to the coastal
sediment transport and erosion patterns around the region, following the
reclamation activity.
• A harbour was developed in the eastern coastline of Thundi settlement in 2006.
This included dredging activities and construction of solid structures
perpendicular to the shoreline, blocking sediment flow along the eastern
coastline. The area north of the harbour is highly mobile as the island was
continually growing in the region. Currently the sediment supply to the area has
been restricted and is very likely to cause erosion in the long term.
49
• There two areas with abandoned harbour work: east of Mukurimagu and north
Thundi Settlement. These areas have dredged harbour basins close to shoreline
and remnants of disposed dredge material. In Mukurimagu, along with the
dredged area, there is a 20m long pile of dredge material located perpendicular
to the shoreline. In thundi settlement the, dredge material no longer exists and
the dredged area has become considerably shallow, possibly during the tsunami
of 2004. These modifications cause disruptions to sediment flow and have
implication for localised erosion and accretion, and possibly to the sediment flow
regime around the island.
• Illegal sand mining is a major problem facing Gan Island. During the field visit,
small scale commercial mining was observed around the island. The islanders
reported that these activities were common and that the authorities were unable
to monitor these activities due limitations in regulation enforcement. Continued
sand mining especially from the oceanward coastline will facilitate severe erosion
if the net production of settlement exceeds the rate of natural erosion and sand
mining.
• Breakwaters have been constructed to mitigate coastal erosion in Mukurimagu
and Thundi settlement. The structures in Mukurimagu settlements are
constructed using sand-cement bags and coral pieces. It was constructed to
protect an historical site which is now within a few meters of the shoreline.
• The harbour constructed in Maandhoo Island comprise of 4m deep dredge area
and a breakwater perpendicular to the shoreline, extending to the eastern reef
edge. Since the reclamation of land between Gan and Maandhoo Island, the
coastline of the two islands is now merged. The presence of the Maandhoo
harbour has essentially blocked all sediment flows along region leading excess
sedimentation on the southern side and severe erosion on the northern side. The
excess growth of seagrass can also be partly attributed to the reduction in
current flows due to harbour construction.
50
beach areas
0 °E
Terrestrial Modifications
• The overall terrestrial environment of the island is in relatively good condition, but
that of settlement areas has been considerably modified.
51
• The vegetation within the settlement areas are limited to large ‘shade’ and
backyard fruit trees.
• Much of the coastal vegetation on the island is intact, but that of settlement areas
have been highly modified. Coastal vegetation cover is highest in the Thundi and
Mathimaradhoo, and lowest in Mukurimagu settlement with just a few meters of
vegetation. While Mathimaradhoo settlement has a strong coastal vegetation
system, it has been considerably modified due to clearing, road development and
sand mining.
• Land reclamation of wetland areas without considering the elevations and
impacts on drainage systems has caused such areas to flood during heavy
rainfall. This is most prominent in Thundi and Mukurimagu Settlements.
• Large scale sand mining, using heavy machinery, has been undertaken close to
the oceanward shoreline near the Mathimaradhoo Settlement. Much these
mining activities were undertaken during the road development activities for the
main road and the new industrial zone. Most of these areas are within 10-30m if
the coastline and involved complete clearing of vegetation. In some parts near
Mathimaradhoo, the width of the vegetation is barely 5m and has the potential to
be breached in the event of abnormal erosion. Since the mined areas are almost
to the low tide level, such a breach may cause permanent land loss.
3.3 Environmental mitigation against historical hazard events.
3.3.1 Natural Adaptation
Gan Island has signs natural adaptation to varying climatic conditions in the past. The
adjustment of ridges, coastal processes and drainage patterns are evident from initial
assessments and require further empirical assessments to understand the adaptation
processes. The defensive mechanisms established for the storms, especially in the
southern half of oceanward shoreline, are critical for minimising impacts of sea induced
hazards as well. It has to be noted that the extent of natural defence systems in Gan are
small compared to other large islands in Maldives such as Seenu Hithadhoo, Haa
Dhaalu Kulhudhuffushi or Fuvahmulah. It may be due to the lack of major natural
hazards, especially ocean induced natural hazards, on the island.
52
3.3.1 Human Adaptation
Gan has very few modifications undertaken to directly prevent natural hazards. The main
activities include construction of breakwaters to prevent erosion in the historical site near
Mathimaradhoo settlement and breakwaters to protect the harbour in Thundi settlement.
More adaptation activities have been undertaken on land to prevent rainfall related
flooding including rising of roads and houses to prevent flooding. The lack of both natural
and human adaptation measures could be taken as crude indicator of the historical
exposure of the island to natural hazards, specifically climate related hazards.
3.4 Environmental vulnerabilities to natural hazards
3.4.1 Natural Vulnerabilities
• The low elevation of the island is the main natural vulnerability of Gan. Similar to
the overall low topography of the island, the ridges around the island are low
compared to the other large islands assessed under this project. This can cause
surges and waves 1.5m above MSL to overtop the ridges and flood.
• The island is located in a high tsunami impact zone due to the ocean floor
topography off the Laamu atoll eastern rim. (Shifaz, 2004).
• It’s location on the eastern rim generally exposes the island to flooding events
and tsunamis (UNDP, 2005)
• As is the case with large islands studied in this project, Gan has substantial
variations in topography. These variations have exposed all three settlements to
occasional rainfall induced flooding. Amongst the three settlements, Mukurimagu
and Thundi reported higher frequency of rainfall flooding. Profiles taken across
the Mukurimagu settlement confirms the presence of major low area with less
than 0.5m from the water table. Similar cases were observed during the
assessment in Thundi settlement.
• The low areas with in the island, especially on the western side of the island
appear to facilitate inundation during flooding events most likely due to the
downward gradient. During the 2004 tsunami it was reported that flood waters
reached 800m inland. Analysis of topographic variation against flood extent
53
revealed the substantial effects of a low area on the flood extents. These
patterns were observed along two main east-west oriented roads of the island
and around Mathimaradhoo Settlement which lies adjacent to a wetland area.
Hence, the topographic variation may play a major role in exposing Gan to both
rainfall induced and ocean induced flooding events.
• The North-south orientation of the island along with low elevation exposes the
majority of the island’s eastern shoreline to ocean induced flooding hazards.
• The narrow width of Gan around Mukurimagu settlement coupled with low
elevation and low ridges exposes the settlement to severe flooding hazards. The
impact of tsunami was most heavily felt on Mukurimagu.
• Reef width appears to play an important role increasing or decreasing the
impacts of ocean induced wave activity. The proximity of Gan Island coastline to
reef edge may increase the exposure of the island to certain sea induced
Hazards. Implications of the existing distance needs to be studied further to
establish a concrete relationship.
3.4.2 Human induced vulnerabilities
• The main human impacts on natural hazards have largely been a result of
present land use.
o In the island of Mathimaradhoo and Mukurimagu the settlement is located
close to the eastern coastline. Coastal vegetation clearing has been
undertaken to accommodate this expansion. In Mathimaradhoo, the
extent of clearing has been controlled, but the coastal vegetation of much
of Mukurimagu is all but removed. It was reported that a large part of
coastal vegetation was affected during the tsunami. However, it is more
likely that the root cause of coastal vegetation loss was due to human
clearing, especially the undergrowth.
o Large areas of the densely vegetated land have been cleared for
agriculture and forestry. Such clearings close to the ocean ward coastline
facilitates to increase extent of flooding. A large proportion of the
agricultural crops located close to the eastern coastline were lost during
the tsunami.
54
• Large areas of land has been cleared and mined for sand close to the eastern
coastline. Sand mining was carried out to obtain material for road development
during the establishment of industrial zone development. Since then the main
roads have also been maintained using similar sand mining methods. The areas
mined are left as large holes, in some areas with barely 5m of coastal vegetation
separating the mined area and high tide line. These mining activities seriously
undermine the coastal defensive system of the mined area. If the narrow stretch
of vegetation separating the sea is breached by wave action, it is highly likely that
the mined areas would get flooded during high tide. It was also reported by the
islanders that the mined areas acted as a drainage zone to mitigate the impact of
the tsunami. While it is likely that the localised effect of tsunami might have been
slightly reduced, the long term effects of such mining activities could be
disastrous in terms of exposure to natural hazards.
• Sand mining has also been undertaken in the beach areas for construction
activities of Gan. There have been reports that sand mining from beach areas
are undertaken at a commercial scale especially following the numerous
construction activities on the island. Sand mining from inhabited island beaches
have been banned, and continued mining would lead to serious implications in
the future especially in terms of coastal erosion.
• Coastal modification activities around the island have caused localised exposure
to natural hazards. In the island of Mukurimagu, coastal structures have been
built in response to coastal erosion around a historical site. On the eastern side
of the same settlement harbour works were started and abandoned leaving a
dredged area and improperly disposed dredge material. Such activities are often
associated with undesired changes to the surrounding coastal environment, most
often coastal erosion.
• Past continuous road maintenance activities on the island to mitigate rainfall
flooding has caused the road to be raised higher than the surrounding housing
plots. As a result, in some area of the island, especially in Mukurimagu
settlement, the houses have become the drainage area for the roads causing
flooding in unlevelled houses during heavy rainfall.
3.5 Environmental assets to hazard mitigation
55
• Gan is the largest island in Maldives and its size is considered the biggest asset
against ocean induced flooding events. However, the advantage of size is
available primarily to the Thundi settlement as it is located on the western
coastline along widest area of the island.
• Generally strong coastal vegetation exists right around the island, especially on
the ocean ward side. The only exception in the Mukurimagu ward where the
coastal vegetation strip is narrow and degraded.
• The vegetation within the island is amongst the densest that can be found in any
inhabited island. Much of the island is therefore protected from the effects of
strong winds. However, there is a noticeable lack of larger trees in the island.
• The coastal processes along the western coastline of the island appear to be
functioning well without much human intervention. The blockage of water and
sediment flow between Gan and Maandhoo due to land reclamation may have
had considerable effect on the southern half of Gan but is expected to have
minimal impact on the rest of the coastline.
• Although a number of modifications have been undertaken on he western side of
the island, the fact that the coastal processes on the eastern side of the island
operates independently allows the natural adaptive functions of the island to
natural hazards remain intact.
• There is a well established drainage system, dominated by wetland areas in the
east and west restricting the impact of rainfall related flooding to these areas.
3.6 Predicted environmental impacts from natural hazards
The natural environment of Gan and islands Maldives archipelago in general are appear
to be resilient to most natural hazards. The impacts on island environments are usually
short-term and insignificant in terms of the natural or geological timeframe. Natural
timeframes are measured in 100’s of years which provides ample time for an island to
recover from impacts from major events such as tsunamis. The recovery of the Gan
Island environment, especially vegetation, ground water and geomorphologic features
following the tsunami is a good example of such rapid recovery. Different aspects of the
natural environment may differ in their recovery. Impacts on marine environment and
coastal processes may take longer to recover as their natural development processes
56
are slow. In comparison, impacts on terrestrial environment, such as vegetation and
groundwater may be more rapid. However, the speed of recovery of all these aspects
will be dependent on the prevailing climatic conditions.
The resilience of coral islands to impacts from long-term events, especially predicted sea
level rise is more difficult to predict. On the one hand it is generally argued that the
outlook for low lying coral island is ‘catastrophic’ under the predicted worst case
scenarios of sea level rise (IPCC 1990; IPCC 2001), with the entire Maldives predicted
to disappear in 150-200 years. On the other hand new research in Maldives suggests
that ‘contrary to most established commentaries on the precarious nature of atoll islands
Maldivian islands have existed for 5000 yr, are morphologically resilient rather than
fragile systems, and are expected to persist under current scenarios of future climate
change and sea-level rise’ (Kench, McLean et al. 2005). A number of prominent
scientists have similar views to the latter (for example, Woodroffe (1993), Morner
(1994)).
In this respect, it is plausible that Gan may continue to naturally adapt to rising sea level,
especially with a relatively unmodified eastern coastline. There are two scenarios for
geological impacts on Gan. First, if the sea level continues to rise as projected and the
coral reef system keep up with the rising sea level and survive the rise in Sea Surface
Temperatures, then the negative geological impacts are expected to be negligible,
based on the natural history of Maldives (based on findings by Kench et. al (2005),
Woodroffe (1993)). Second, if the sea level continues to rise as projected and the coral
reefs fail to keep-up, then their could be substantial changes to the land and beaches of
Gan (based on (Yamano 2000)). The question whether the coral islands could adjust to
the latter scenario may not be answered convincingly based on current research.
However, it is clear that Gan stands to undergo substantial change (even during the
potential long term geological adjustments), due to potential loss of land through erosion,
increased inundations, and salt water intrusion into water lens (based on Pernetta and
Sestini (1989), Woodroffe (1989), Kench and Cowell (2002)).
Gan has particular vulnerability to sea level rise due to the presence of wetland areas.
Since wetland areas in coral islands are linked to the tide and sea level, an increase in
sea level may result in increase in size of such areas and a subsequent reduction in land
(Woodroffe 1989).
57
As noted earlier, environmental impacts from natural hazards will be apparent in the
short-term and will appear as a major problem in inhabited islands due to a mismatch in
assessment timeframes for natural and socio-economic impacts. The following table
presents the short-term impacts from hazard event scenarios predicted for Gan.
Hazard Scenario Probability at Location
Potential Major Environmental Impacts
Tsunami (maximum scenario) 4.5m Low • Widespread damage to coastal vegetation
(Short-term)
• Long term or permanent damage to selected inland vegetation especially common backyard species such as mango and breadfruit trees.
• Salt water intrusion into wetland areas and island water lens causing loss of some flora and fauna.
• Contamination of ground water if the sewerage system is damaged or if liquid contaminants in the industrial zone such as diesel and chemicals are leaked.
• Salinisation of ground water lens to a considerable period of time causing ground water shortage. If the rainwater collection facilities are destroyed, potable water shortage would be critical in Mathimaradhoo and Mukurimagu Settlement.
• Widespread damage to crops (short-term)
• Widespread damage to coastal protection infrastructure
• Short-medium term loss of soil productivity
• Moderate damage to coral reefs (based on UNEP (2005))
Storm Surge (based on UNDP, (2005)) 0.60m (1.53m
storm tide) Low • Minor damage to coastal vegetation
• Minor loss of crops
• Minor geomorphologic changes in the northern oceanward shoreline and lagoon
1.32m (2.30m storm tide)
Very Low • Moderate damage to coastal vegetation
• Long term or permanent damage to selected inland vegetation especially common backyard species such as mango and breadfruit trees.
• Salt water intrusion into wetland areas and island water lens causing loss of some flora and fauna.
58
Potential Major Environmental Impacts
• Contamination of ground water if the sewerage system is damaged or if liquid contaminants in the industrial zone such as diesel and chemicals are leaked.
• Salinisation of ground water lens to a considerable period of time causing ground water shortage. If the rainwater collection facilities are destroyed, potable water shortage would be critical in Mathimaradhoo and Mukurimagu Settlement.
• Loss of crops
• Minor-moderate geomorphologic changes in the oceanward shoreline and lagoon
• Minor-moderate damage to coral reefs Strong Wind
28-33 Knots Very High • Minor damage to very old and young fruit trees
• Debris dispersion near waste sites.
• Minor damage to open field crops 34-65 Knots Low • Moderate damage to vegetation with falling
branches and occasionally whole trees
• Debris dispersion near waste sites.
• Moderate-high damage to open field crops
• Minor changes to coastal ridges 65+ Knots Very Low • Widespread damage to inland vegetation
• Debris dispersion near waste sites.
• Widespread damage to open field crops
• Minor changes to coastal ridges Heavy rainfall
187mm Moderate • Minor to moderate flooding in low areas, including roads and houses.
240mm Very Low • Widespread flooding but restricted to low areas of the island.
Drought • Minor damage to backyard fruit trees Earthquake • Minor-moderate geomorphologic changes Sea Level Rise by year 2100 (effects of single flood event)
Medium (0.41m)
• Loss of land due to erosion.
• Loss of coastal vegetation
• Major changes to coastal geomorphology.
• Saltwater intrusion into wetland areas and salinisation of ground water leading to water shortage and loss of flora and fauna.
• Minor to mo