Author’s Accepted Manuscript

Assessing Mozambique's exposure to coastal

climate hazards and erosion

Pedro Cabral, Gabriela Augusto, Adeoluwa

Akande, Anjos Costa, Nelson Amade, Sérgio
Niquisse, Ali Atumane, António Cuna, Kiana
Kazemi, Ramires Mlucasse, Rubini Santha
www.elsevier.com/locate/ijdr

PII: S2212-4209(16)30619-7
DOI: http://dx.doi.org/10.1016/j.ijdrr.2017.04.002
Reference: IJDRR533
To appear in: International Journal of Disaster Risk Reduction
Received date: 15 November 2016
Revised date: 10 April 2017
Accepted date: 12 April 2017
Cite this article as: Pedro Cabral, Gabriela Augusto, Adeoluwa Akande, Anjos
Costa, Nelson Amade, Sérgio Niquisse, Ali Atumane, António Cuna, Kiana
Kazemi, Ramires Mlucasse and Rubini Santha, Assessing Mozambique's
exposure to coastal climate hazards and erosion, International Journal of
Disaster Risk Reduction, http://dx.doi.org/10.1016/j.ijdrr.2017.04.002
This is a PDF file of an unedited manuscript that has been accepted for
publication. As a service to our customers we are providing this early version of
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form.
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
 Assessing Mozambique’s exposure to coastal climate hazards and erosion

Pedro Cabral1, Gabriela Augusto1, Adeoluwa Akande1, Anjos Costa1,2, Nelson Amade1,2, Sérgio
Niquisse1,2, Ali Atumane1,2, António Cuna1,2, Kiana Kazemi1, Ramires Mlucasse1,2, Rubini Santha1

1
NOVA IMS, Universidade Nova de Lisboa, 1070-312, Portugal
2
Universidade Católica de Moçambique, Beira, Moçambique

*Corresponding author, [email protected]

Abstract
An increasing number of people in the world are living in coastal areas characterized by high
geophysical and biophysical sensitivity. Thus, it is necessary to provide coastal planners with tools
helping them to design efficient management plans to mitigate the negative effects caused by a
growing number of coastal climate hazards that threaten life and property. We calculate an
Exposure Index (EI) for the coastline of Mozambique and assess the importance of the natural
habitats in reducing exposure to coastal climate hazards. We estimate, for year 2015, an increase of
276% in the number of people affected by a high, or very high, level of exposure when compared to
a “Without habitats” scenario, i.e. excluding the protective effects of sand dunes, mangroves, and
corals. The results of the EI are supported by the Desinventar Database, which has historic data
concerning loss and damage caused by events of geological or weather related origin. These results
also indicate where the most exposed areas are thereby providing useful information to design
effective coastal plans that increase resilience to climate hazards and erosion in Mozambique.

Key-words: InVEST; climate change; ecosystem services; coastal hazard; Desinventar

1. Introduction
The severity of the impacts of extreme and non-extreme weather and climate events depends
strongly on the level of vulnerability and exposure to these events [1]. It is widely accepted that
coasts will be exposed to increasing climate hazards in the coming decades due to climate change
and sea-level rise [2]. Exposure is defined as “the situation of people, property, systems, or other
elements present in hazard zones that are thereby subject to potential losses” [3] and vulnerability
refers to “the characteristics and circumstances of a community, system or asset that make it
susceptible to the damaging effects of a hazard” [2]. Exposure is also referred to as biophysical

1
vulnerability, site, and situation of the place, whereas vulnerability is referred to as social
vulnerability comprising the built environment and the social fabric [4,5].
Coastal regions are particularly exposed, although with different levels of vulnerability, to the
adverse effects of climate change and sea-level rise, leading to increased risks that include loss and
damage by climate hazards and coastal erosion, which may be exacerbated by increasing human-
induced pressures [6]. However, coastal areas are attracting an increasing number of people
because they often provide more subsistence resources than those found inland, more access points
to marine trade and transport, more recreational, and cultural opportunities and, for some, a special
sense of place at the land/sea interface [7]. This population growth, and the associated high
urbanization rates, are driven by fast economic development and migration toward the coast [8,9].
Consequently, the degree of exposure coastal populations is being boosted by the pressures added
to coastal systems that need to accommodate more and more people and associated infrastructures
[10,11], and by the expected increase in coastal flooding and sea level rise [12]. The exposure of
coastal areas depends greatly on their geomorphologic characteristics, and a significant reduction in
the exposure of coastal areas to climate hazards and erosion can be achieved through the
conservation of natural habitats that provide protection to the coast (e.g. dunes, mangroves, corals,
seagrasses, etc.), or by building coastal protection infrastructures [13–16]. The former is generally an
option for most countries, but the latter is only viable for countries that can afford to pay for
expensive engineering infrastructures [17].
The need exists to study coastal exposure and a considerable number of studies have been carried
out for many regions in the world using different methodologies [10,13,18–23]. However, with some
exceptions [24,25], these studies are scarce for Africa. Historic data availability for African countries,
and particularly on sea-level rise, is fairly limited when compared to other regions in the World [26].
In Mozambique, a country that has been identified as being particularly exposed to increased
flooding due to the occurrence of tropical storms [27,1,28], with the exception of a study for the city
of Beira [28], there is no study reported to date about exposure to coastal climate hazards and
erosion. Besides having one of the longest African coastlines (approximately 2,700km) this country
also has approximately 60% of its population located in coastal districts [29–31].
In this article we calculate the exposure index of the Mozambique coast to climate hazards and
erosion using a 1km2 spatial resolution. To assess the model pertinence, the results of the exposure
index (EI) are compared with the Desinventar Database [32], which has historical data concerning
loss and damage caused by events of geological or weather related origin. We also assess the
importance of the habitats in reducing exposure to coastal climate hazards and erosion.

2
 2. The study area
Mozambique is located in southeast Africa and comprises a land surface of about 800,000 km2.
Maputo is the capital and the country’s largest city (Figure 1). The country had an estimated
population of 25 million inhabitants in 2014 [33]. About 70% of Mozambicans live in rural areas, and
40% of them are located in the districts of Nampula and Zambézia [33]. About 45% of the population
is under 14 years old, and the elderly represent about 3.5% of the total population [33]. Typically
Mozambicans live in large families, 53.9% of the families have more than 5 members, and the
presence of collateral related family, or unrelated elements in the household is relatively common
(15%) [34]. The purchasing power in Mozambique has been rising sharply, but on average each
Mozambican spends less than 21 USD per month, or about 14 USD per month if living in a rural area
[34]. Most lodgings are built of adobe; only 18% of the houses are made of bricks, and 30% of the
families live in houses made of precarious building materials [34]. Safe water supply reaches roughly
half of the households, and the electric utility network is poor or nonexistent in rural households.
Safe sanitation covers less than 1% of rural households. Nearly 45% of the Mozambican population
cannot write or read, and this percentage is higher among women (58%) [34].
Mozambique is divided into 11 provinces and shares borders with Tanzania, Malawi, Zambia,
Zimbabwe, South Africa, and Swaziland. After a devastating civil war that ended in 1992 [35], the
first democratic elections took place in 1994, and since then the country has enjoyed political
stability and rapid economic growth (7.4% of annual GDP growth in 2014) (World Bank, 2016). A
Gross Domestic Product (GDP) of US$16.961 billion and a Gross National Income (GNI) per capita of
US$600 in 2014 (World Bank, 2016) make Mozambique one of the poorest countries in the world.
This country ranked 180th out of 188 countries in the most recent Human Development Index (HDI)
[37].

3
 Figure 1 Study area

There are 104 identified river basins flowing into 11 major rivers, with the Zambezi and Limpopo
being the two largest, that drain the water from the Central African high plateau into the Indian
Ocean [38,39]. The majority of the rivers have a torrential, wet season regime with high waters
during 3-4 months, and low, dry season flows (sometimes totally dry) for the remainder of the year.
Up to 52% of the country is within international river basins [39]. The continental shelf is wider in
the center (approximately 145km, near Sofala), decreasing in both north and south [40,41].
The Mozambican coast is divided into three main areas: the sandy coastline in the south, the
estuarine coast in the center, and the coralline coast, with coral limestone in the north [40,42]. The
sandy coastline starts at the South African border and extends to the Save Delta. It is a lowland coast
dominated by sand dunes, lakes, lagoons, and bays receiving river estuaries, namely the Maputo
bay, which accepts the estuaries of four rivers [41,42]. This lowland coast is bordered by an
extensive dune system [41] dominated by wetlands of the deltas of large rivers in the estuarine
coastline. The estuarine coastline starts at the mouth of the Save (130km south of Beira), and
extends to Angoche, encompassing the Zambezi river delta. The coralline coastline covers over
700km from Angoche until the Tanzanian border. This is a narrow coastline with the bordering
plateau sloping into hills and highlands. The whole coast is also punctuated by numerous coral and

4
sandy islands, some very popular for recreation and leisure. The coast of Mozambique has 49
districts, belonging to different provinces, evenly distributed among the three natural areas.
Mozambique has a tropical climate with two distinct climatic seasons, a wet summer from October
to March and a dry winter from April to September. The tropical cyclone season in the South Indian
Ocean occurs between November and April, and over half of the storms that occur during a typical
season appear in January or February [43]. On average 1.16 tropical cyclones per year made landfall
in Mozambique, between 1980 and 2007 [44]. About half of those were formed in the Mozambique
Channel [45]. The frequency and intensity of tropical cyclones has increased since 1993, with seven
cyclones of category 3 to 5 hitting the coastline in the period 1994-2007, against only two in the
previous 12 years [43]. Of the 15 cyclones occurring on the Mozambican coast between 1980 and
2007, eight hit the estuarine coastline, four in the coralline coastline, and three in the sandy
coastline [46].
The Mozambique coast is largely dominated by natural habitats, although their conservation status
is not always optimal [40,47]. Sand dunes are common all along the coast. In the southernmost part
of the Mozambique coast, the sandy coastline consists of a low lying coastal plain separated from
the coast by an uninterrupted undulating cordon of sand dunes extending for 100km, up to 2km
width and rising to over 100m above sea level at some points [48]. This extensive dune system is
interrupted on the estuary coast in the center by wetlands, where the dominant natural habitat is
mangroves. In the narrow coralline coast north of Angoche the sand dunes are small, alternating
with cliffs and attached beaches[48]. Mangroves occur along the whole Mozambican coast. In the
coralline coast, where rivers are non-tidal, mangroves are mainly confined to the vicinity of river
mouths. In the estuarine coast, the estuaries and deltas of big rivers such as the Save, Zambezi,
Púnguè, and Buzi, allow for mangroves to extend for dozens of kilometers inland and hundreds of
kilometers along the coast. Mangroves are also found in the sandy coastline, again in the estuary of
rivers, notably in the Morrumbene estuary and on the bays of Maputo and Inhambane.
There are nine species of mangrove trees in Mozambique, and the estimated area of this forest in
the country is over 350,000ha [40]. The estimated area of coral reefs in Mozambique coast ranges
from 413 to 570km2 [38]. In the northernmost part of the coast, coral reefs constitute an almost
continuous fringing reef for more than 700km on the eastern shores of the islands and in the
exposed mainland coast. The turbidity of the coast around the main river mouths on the estuarine
coast does not allow coral reefs to thrive, but along the sandy southern coast, for about 850km,
coral reefs are again found, although more sparsely than in the north. Coral reefs are one of the
major attractions for the coastal tourism industry, and many infrastructures are already in place for
that activity [47].

5
Mozambique is likely to be severely affected by climate change and the rising sea level because of its
location in the inter-tropical conversion zone, its long coastline with large portions below sea level,
and because it is a catchment zone of large international river basins[31]. Coastal erosion is also a
pressing concern in Mozambique. South of Maputo the estimated average coastal retreat increased
from 0.11m/year in the period 1971-1975 to 1.1m/year in the period of 1999-2004 [49], caused by
climate change and land cover changes [50]. Coastal erosion is also a major concern in the mangrove
region, threatening the mangroves and forcing population displacement [50,51].

3. Materials and methods

The Exposure Index

The relative exposure of the Mozambican coastline to coastal erosion and inundation was estimated
using the Coastal Vulnerability model available in InVEST [52]. InVEST is a free open-source tool
developed by the Natural Capital Project, available at www.naturalcapitalproject.org. It includes a
set of different ecosystem service models. Besides being an intuitive tool, it has also been designed
for data poor environments as is the case of this study. The Coastal Vulnerability model provides an
Exposure Index (EI) map by ranking biophysical variables for each coastline segment, or pixel (Eq. 1):

EI = (RGeomorphologyRReliefRHabitatsRWindExposureRWaveExposureRSurge)1/7 (Eq. 1)

where EI is the Exposure Index and Ri represents the ranking of the ith bio-geophysical variable used
to calculate the EI. The variables used in this model are the following [53]: RGeomorphology is the
rank of geomorphology for each coastline segment; RRelief is the relief ranking of each coastline
segment; RHabitats is a natural habitat ranking for each coastline segment; RWindExposure is the wind
exposure ranking of each coastline segment; RWaveExposure is the average height of adjacent seas, and
RSurge is the distance ranking between the coastline and the edge of the continental margin. Although
this model allows for the inclusion of a variable of sea-level rise (SLR), this variable was not
considered in this study due to the lack of data.

This model weights equally the variables that consider the effect of storms on exposure by
incorporating wind, waves, and surge potential data as well as data and models for habitat type,
geomorphology, relief, and sea level rise [13]. This model was selected because it is freely accessible
and because it provides spatially explicit and quantitative outputs, making it useful for coastal
management assessments [13,52]. Additionally, it requires a relatively low quantity of data and has

6
been successfully applied in different study areas [13,53,54]. The model also enables the estimation
of the population subject to coastal hazards and erosion by extracting population values from the
grid cell that overlaps the shoreline segment [52].

Habitat scenarios
The habitats considered in this study, i.e. mangroves, coral reefs and sand dunes, are generally
accepted as important factors in the protection of coastlines worldwide. The ability of mangroves to
protect population and infrastructure from storms and cyclones has been widely assessed in
numerous scientific studies [15,55]. Coral reefs have also been acknowledged for dissipating wave
energy and, therefore, helping to avoid the worst effects of sea storms on habitats and
infrastructures of nearby coasts [56,57]. Sand dunes are important in preventing coastal erosion and
flooding [58,59]. Despite the presence of other important habitats, such as seagrass beds, and their
importance in coastal protection [13], these were not included in the model due to data
unavailability.
To assess the role of coastal ecosystems in reducing exposure to sea-level rise and storms, two
scenarios were developed: (i) the “With habitats” scenario, which included the three habitats
(mangroves, coral reefs, and sand dunes) in the exposure index; and (ii) the “Without habitats”
scenario, which assumed that those habitats no longer provide protection. The “With habitats”
scenario corresponds to the “current state” of the system. The “Without habitats” scenario is an
extreme situation that is used to evaluate where, and to what extent, habitats play a significant role
in protecting people, and to determine where their loss would affect exposure from coastal hazards.
Future scenarios are expected to vary along this spectrum between the “current state” of the system
and a total degradation of the protective effect of habitats.

Data and preprocessing

A national-scale database was developed to assess the Mozambican coastal exposure. This included
data gathered from local (GIS Center of the Catholic University of Mozambique – UCM, in Beira) and
state agencies (CENACARTA, National Centre of Cartography, http://www.cenacarta.com), as well as
through the Internet to access global datasets produced by several organizations (Table 1). All the
data were preprocessed using ArcGIS software [60] and had, or were converted into the Universal
Transverse Mercator (UTM) projected coordinate system using the World Geodetic System datum
84 (WGS 84).

7
Table 1 Description of datasets used in the study

Dataset Description Source

Land polygon Geographic shape of the coastal area of interest, depicting the [61]
boundaries of land and seascape.
Relief The Digital Elevation Model (DEM) that was used to compute the [62]
relief ranking of each shoreline segment was the Advanced
Spaceborne Thermal Emission and Reflection Radiometer (ASTER)
which has a spatial resolution of 30 meters.
Geomorphology The geomorphology map was obtained using the shoreline map UCM GIS
from the Global Self-consistent, Hierarchical, High-resolution Center,
Geography Database (GSHHG) and a lithology map from CENACARTA
CENACARTA. and [61]
Natural habitats The dune and mangrove habitats were extracted from the CENACARTA
Mozambican land cover map of 1999 produced by CENACARTA. The and UCM GIS
coral reefs polygon ESRI shapefile was provided by the UCM GIS Center (1999)
Center.
Climate forcing This input was used to compute the Wind and Wave Exposure [63]
grid ranking of each shoreline segment. We used the full-spectral third-
generation wind-wave model WaveWatch III.
Continental shelf Depth contour used as an indicator for surge level. [52]
Bathymetry This input is used to compute the average depth along the fetch [64]
rays to determine the exposure of each shoreline segment and in
the computation of the surge potential. This grid has a spatial
resolution of 30 arc-second.
Population We used the estimates of the Gridded Population of the World [65]
(GPW) population data for year 2015 with a spatial resolution of 2.5
arc minutes.

The lithology map used to compute the RGeomorphology variable was reclassified from 1 (very low
erodibility) to 5 (very high erodibility) with the help of experts from the Portuguese National
Laboratory of Energy and Geology (LNEG). Alluvions and terraces were given the maximum value of
erodibility (i.e., 5). Undifferentiated deposits received value 4. Sandstones, clays, and related rocks
were classified with 3. Basic and carbonate rocks were classified with 2. Gneisses, migmatites,
granites, and related rocks were given value 1. Finally, a GIS procedure was used to assign the value
of the reclassified lithology polygons to the shoreline map.

8
The natural habitats were associated with an increasing exposure rank, from 1 (very low) to 5 (very
high) [13]. The coral reefs and mangroves were classified with value 1, and dunes with value 2. The
corals and mangroves protective distance considered was 2000m. The dunes’ protective distance
was set to 300m [52].

Model validation
To validate the ability of the exposure index to capture the exposure to coastal climate hazards, we
used the publicly available Desinventar Database [32]. This database compiles records of disaster
loss and damage from events of geological or weather related origin in 82 countries, including
Mozambique. The data contained in the Desinventar Database are allocated by district, so we report
the exposure index of individual cells also using the district level. The districts have different
numbers of cells, each with its own computed and classified exposure index, and for this reason we
decided to classify as “Exposed” any district in which 10% or more of its cells belonged to the “High”
or “Very High” classes of the exposure index. Using this criterion, data concerning loss and damage
from coastal climate events from the Desinventar Database were compared to investigate if
“Exposed” districts had a significantly different record from the districts considered “Less exposed”
by the model concerning climate hazards typically associated with the coast: tropical depression,
surge, strong wind, storm, rains, floods, flash flood, and cyclone. A total of 2003 records of damage
and loss due to these events occurred in the 128 districts of Mozambique, 711 of these were located
in 48 coastal districts.

4. Results
4.1 The Exposure Index (EI)
The exposure index (EI), which is a relative index resulting from the InVEST Coastal Vulnerability
model, was calculated for the “with” and “without” habitats scenarios. The values of the cells of the
two resulting rasters (one for each scenario) were combined together and classified into five
quintiles as follows: Very low” (<2.54); “Low” (2.54 to 2.90); “Moderate” (2.90 to 3.22); “High” (3.22
to 3.53); and “Very high” (> 3.53). Figure 2 shows the relative exposure to coastal hazard and erosion
of each 1km2 compared with all other segments of Mozambique in the “With habitats” scenario.

9
 Figure 2 Exposure Index for Mozambique in the “With habitats” scenario

The most important concentration of cells classified as “High” and “Very high” exposure are located
in the Provinces of Zambézia and Sofala, which correspond to lowlands with unfavorable
geomorphology.

4.1.1 The habitats’ importance in coastal protection

About 23.4% of the cells of the resulting EI for the Mozambique coast were classified with “High” or
“Very high” exposure to coastal climate events and erosion in the “With habitats” scenario. This
value increases to 57.8% in the “Without habitats” scenario; 26 coastal districts (out of a total of 49)
have 10% of more of their cells classified as “High” or “Very high” exposure level in the “With
habitats” scenario. In the “Without habitats” scenario this number would increase to 41 (Figure 3).
The absence of the protective effect of the habitats in Mozambique coast would expose about
577,448 people to coastal climate hazards and erosion, representing an increase of 276% in the
number of exposed people. The average EI for the Mozambique coast would increase from 2.81 in
the “With habitats” scenario to 3.26 if the protective effect of habitats was removed, i.e., an
increase of 16%.

10
 Figure 3 Exposed districts in the “With habitats” (A, map on the left) and in the “Without habitats”
(B, map on the right) scenarios

4.2 Comparison of EI with recorded history of coastal climate events causing loss and damage
For validation purposes, we considered only the exposed districts of the “With habitats” scenario, as
this is assumed to be the “current state” of the system. The coastal districts in this scenario were
split into two sets regarding coastal climate hazards types according to the Desinventar Database,
one referring to the districts deemed exposed by the model, and the other set with the districts that
the model considers less exposed (Figure 3). On average, each exposed district suffered 15.8 climate
related events causing loss or damage, whereas the less exposed districts suffered on average 12.7
events for the same period of time (Table 2). This difference is not statistically significant (p-value=
0.199).

11
Table 2 Coastal climate events causing damage and/or loss and human casualties between 1975 and
2015 (Source: Desinventar Database, 2016 [32])

Average in exposed Average in less p-value

districts exposed districts
Coastal climate events causing 15.8 12.7 0.199
damage and/or loss
Coastal climate events causing 5.28 2.48 0.012
human casualties

However, the human casualties occurring as a direct consequence of climate related events are still
the most important indicator of the severity of the event. Thus, the most pertinent variable to be
considered in the model assessment is the number of events causing fatalities, as this is the first
consequence to be avoided. According to the Desinventar Database, there were 147 climate related
events that caused deaths in the coastal districts of Mozambique between 1975 and 2015. Cyclones,
floods, and flash floods accounted for the large majority of these events (Figure 4). On average, the
exposed districts suffered 4.66 climate related events causing death among their population,
whereas the less exposed districts suffered 2.65 deadly climate related events. This difference is
statistically significant (p-value of 0.012).

Coastal climate hazards causing human casualties in

Mozambique coastal districts (1979-2015)

Average number of events

TROP. DEPRESS. with human casualties in less
Exposed Districts
THUNDER STORM
Average number of events
STRONG WIND
with human casualties in
STORM Exposed Districts
RAINS
FLOOD
FLASH FLOOD
CYCLONE

0 0.5 1 1.5

12
 Figure 4 Coastal climate hazards causing human casualties in Mozambique’s coastal districts (1979-
2015) (Source: Desinventar Database, 2016)

These results support the pertinence of the EI indicating that exposed districts are more likely to
suffer damage and loss from climatic events, and are also more likely to suffer fatalities from
cyclones, floods, and storms. As the two sets of data are totally unrelated except for the geographic
location, and the Desinventar Database records the historic “ground truth” of coastal climate
hazards, it is therefore reasonable to assume that the EI is a pertinent measure of coastal exposure
to climate coastal events and erosion on the Mozambique coast.

5. Discussion

An Exposure Index for Mozambique

The Mozambique coast is attracting considerable interest for development projects in agriculture,
aquaculture, energy crops, tourism, and related infrastructures [31,66]. Whereas most of these
projects are desirable and may help to improve the life of a growing population, the local and
national decision makers should account for all ecosystem services whenever a natural habitat is
replaced by some other land use required for these projects. The protection of the coast is one of
the ecosystem services seldom accounted for. The EI map proposes a first glance regarding the
exposure to climate hazards and erosion of the whole Mozambique coast. Moreover, it provides
valuable hints regarding the importance of habitat conservation for the protection of coastal
populations from climate hazards and coastal erosion.
Results point to the particularly high exposure of districts belonging to the central estuarine region,
where 14 out of 15 have more than 10% of their coastal area classified as “High” or “Very high” by
the EI. The removal or extreme deterioration of mangrove in this region would place all the districts
of the estuarine region into the situation of highly exposed. Coincidently, the estuarine coast is also
the region where cyclones hit twice as often as in the north or the south [46], and is the coastal
region more prone to suffer from floods due to the transnational and torrential nature of its big
rivers. Moreover, mangroves due to the extraordinary range of goods and services provided to the
local population [51,55], attract a great number of inhabitants who tend to dwell very close to the
coastline. As a consequence, the number of people affected by coastal hazards is greater in this
region. It should also be noted that while providing protection to the coast mangroves themselves
are affected by climate hazards [67]. The importance of mangrove conservation and management is
clearly expressed by the Mozambique government national strategy and action plan for mangrove

13
protection and rehabilitation of mangrove forests [30]. It should be stressed though that the regular
flow of freshwater, crucial for the conservation status of mangroves [30,40], requires the
management of river basins well outside of Mozambique.
The importance of the wooded dunes in the sandy coast is also clearly highlighted by the EI results.
The absence of this habitat would expose nine more districts to coastal climate hazards and erosion.
Without sand dunes, 14 of the 15 districts in the southern coast would be highly and very highly
exposed to climate events and erosion. Wooded sand dunes are also under considerable pressure
due to the development of tourism and increasing reclamation of land for agriculture, and this
region of the coast is particularly subject to coastal erosion due to sea dynamics [50].
The absence of habitats in the coral coast would also increase the number of exposed districts by
58%. The deficient management of fishing is one of the main threats to coral reefs on the
Mozambique coast [47], while inland pollution and the deterioration of mangroves trapping
sediments from inland are also a concern [31]. It must be noted however that many of the coral
reefs of Mozambique are located in the dozens of islands that were not included in this study.

Model validation
Model validation is an important aspect to consider when producing exposure maps, or any kind of
modelling output, in order to demonstrate the correctness of the theoretically presumed indications
of exposure and the validity of the indicators [68,69]. In this paper we used the publically available
Desinventar Database [32] to assess the accuracy of our model. This freely available database has
proved to be an extremely valuable asset to be used in exposure studies. The validation method
developed in this study shows how to take advantage of this database in a very straightforward way.

Limitations
A number of limitations of this modeling approach are identified in the InVEST model manual
including the simplification of the exposure index calculation into the geometric mean of seven
variables and exposure categories [52]. In addition to those, we would like to stress some features,
such as the quality and conservation status of the habitats that were not considered in our model.
Other identified limitations include data scarcity, which was an important hindrance we had to deal
with. For instance, data about the sea level rise for Mozambique come from a single station in
Maputo. If more observations were available, it would be possible to estimate the trend of sea level
variation by fitting the available observations to a linear regression and producing different sea level
rise scenarios. For this reason, we did not create scenarios of sea level rise in this study.

14
The habitats dataset was produced in 1999, which makes it outdated. Nevertheless, we must accept
its important impact on the results. Future versions of this model should use more recent maps of
habitats, as well as consider additional habitats that are important for coastal protection (e.g.
seagrass beds, kelp forests, others) [13], mostly because there has been a strong upturn in tourism
and other economic activities in Mozambique coastal areas. We used estimates of Mozambican
population for year 2015, which may also inaccurately represent the number of people actually living
in the coastal districts of Mozambique. Additionally, and regarding the importance of casualties
caused by climate related events, the lack of estimates for population in Mozambique districts in the
period covered by the database, precludes further investigation. During that time period the
Mozambique population more than doubled, and there were massive unmonitored population
displacements during the civil war. Another point to address concerns the coastal morphology.
South of mouth of the Zambezi river the coast is a large lowland plain where the land morphology
provides no physical obstacles to storm progression, and therefore the 1km inland distance may not
be the appropriate measure to estimate the population and infrastructure likely to be affected by
cyclones or floods. This is especially the case for the estuarine region, where mangroves encourage
the settlement of more population than sand dunes or beaches. Thus, the affected population is
likely to be largely underestimated in these areas. Bearing in mind that most lodgings in the country
are of light construction, cyclones do cause loss and damage hundreds of kilometers inland.
Finally, it is important to stress that this is not a vulnerability model for the Mozambique coast [4].
The EI map only addresses exposure, or the “biophysical vulnerability”. Studies concerning the social
vulnerability to climate hazards in Africa will undoubtedly be of utmost importance and interest,
especially in highly dynamic demographic regions, as is the case of Mozambique.

6. Conclusions
Material losses due to climate events in developing countries, such as Mozambique, have severe
implications in the lives of the populations [51,70]. Climate related events often cause the loss of
crops and infrastructures, endangering the lives and health of the population through famine, water
contamination, and blocking much needed aid for the population during and long after the disaster
[51,70]. Mozambique is demonstrating an astonishing capacity to prepare and adapt its population
to climate hazards with very limited resources and capacities [71]. Some challenges demand
international support as pressure on coastal areas will increase with the economic development of
an increasing population and ongoing climate change. This support is especially crucial in the
management of the large rivers that the country shares with neighboring states: more than the
impact of the level of fresh water flow on the welfare of the local population and the conservation of

15
coastal habitats, we should note that according to the Desinventar Database, the main events
causing casualties, loss, and damage in Mozambique are droughts and floods.
This study presented a methodology for obtaining an EI for the Mozambique coastline. The results
are statistically supported by independent validation data from the Desinventar Database. The EI is a
valid indicator to assess the coastal exposure and anticipate the effects of land cover changes in the
exposure of coastal infrastructures and populations to coastal climate hazards and erosion. The
model provides a consistent framework to assess and analyze coastal exposure in a country which
needs to deal with coastal hazards regularly, and where coastal erosion is a pressing challenge.
Additionally, the study reveals the importance of preserving the natural habitats to save lives from
coastal hazard events. This is particularly important for poor countries that do not have the financial
means to build and maintain coastal protection infrastructures. Finally, it also demonstrates the
importance of having updated datasets for building better exposure models. The model’s accuracy
can be easily improved with the update of the multiple underlying datasets and some local
adjustments to expand affected areas according to coastal morphology. The findings of this study
may help decision makers to prioritize conservation efforts to avoid the worst effects of climate
coastal hazards and coastal erosion in Mozambique.

Acknowledgements
The authors would like to acknowledge the contributions of Aurete Pereira and Tomás Oliveira from
LNEG for helping with information about the lithology map. We would also like to thank Gregg
Verutes from the Natural Capital Project, who helped us decisively in calibrating and interpreting
some results of the InVEST coastal vulnerability model.

16
References

[1] O.-D. Cardona, M.K. van Aalst, J. Birkmann, M. Fordham, G. McGregor, R. Perez, et al.,
Determinants of Risk: Exposure and Vulnerability, in: C.B. Field, V. Barros, T.F. Stocker, Q.
Dahe (Eds.), Manag. Risks Extrem. Events Disasters to Adv. Clim. Chang. Adapt., Cambridge
University Press, Cambridge, 2012: pp. 65–108. doi:10.1017/CBO9781139177245.005.

[2] C.B. Field, V.R. Barros, D.J. Dokken, K.J. Mach, M.D. Mastrandrea, eds., Coastal Systems and
Low-Lying Areas, in: Clim. Chang. 2014 Impacts, Adapt. Vulnerability, Cambridge University
Press, Cambridge, 2014: pp. 361–410. doi:10.1017/CBO9781107415379.010.

[3] UNISDR, Terminology on Disaster Risk Reduction. United Nations International Strategy for
Disaster Reduction, (2009). http://www.unisdr.org/we/inform/publications/7817 (accessed
January 11, 2016).

[4] S.L. Cutter, B.J. Boruff, W.L. Shirley, Social Vulnerability to Environmental Hazards, Soc. Sci. Q.
84 (2003) 242–261. doi:10.1111/1540-6237.8402002.

[5] V.N. Martins, D.S. Silva, P. Cabral, Social vulnerability assessment to seismic risk using
multicriteria analysis: the case study of Vila Franca do Campo (São Miguel Island, Azores,
Portugal), Nat. Hazards. 62 (2012) 385–404. doi:10.1007/s11069-012-0084-x.

[6] R.J. Nicholls, P.P. Wong, V.R. Burkett, J.O. Codignotto, J.E. Hay, R.F. McLean, et al., Coastal
systems and low-lying areas. Contribution of Working Group II to the Fourth Assessment
Report of the Intergovernmental Panel on Climate Change., in: M.L. Parry, O.F. Canziani, J.P.
Palutikof, P.J. van der Linden, C.E. Hanson (Eds.), Clim. Chang. 2007 Impacts, Adapt.
Vulnerability, Cambridge University Press, Cambridge, UK, 2007: pp. 315–356.
https://www.ipcc.ch/pdf/assessment-report/ar4/wg2/ar4-wg2-chapter6.pdf.

[7] B. Neumann, A.T. Vafeidis, J. Zimmermann, R.J. Nicholls, Future coastal population growth
and exposure to sea-level rise and coastal flooding--a global assessment., PLoS One. 10
(2015) e0118571. doi:10.1371/journal.pone.0118571.

[8] G. McGranahan, D. Balk, B. Anderson, The rising tide: assessing the risks of climate change
and human settlements in low elevation coastal zones, Environ. Urban. 19 (2007) 17–37.
doi:10.1177/0956247807076960.

[9] K. Smith, We are seven billion, Nat. Clim. Chang. 1 (2011) 331–335.
doi:10.1038/nclimate1235.

[10] V.N. Martins, R. Pires, P. Cabral, Modelling of coastal vulnerability in the stretch between the

17
 beaches of Porto de Mós and Falésia, Algarve (Portugal), J. Coast. Conserv. 16 (2012) 503–
510.

[11] EEA, The changing faces of Europe’s coastal areas, Copenhagen, Denmark, 2006.
http://www.eea.europa.eu/publications/eea_report_2006_6.

[12] IPCC, Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing
Team, R.K. Pachauri and L.A. Meyer (eds.)], Geneva, Switzerland, 2014.

[13] K.K. Arkema, G. Guannel, G. Verutes, S.A. Wood, A. Guerry, M. Ruckelshaus, et al., Coastal
habitats shield people and property from sea-level rise and storms, Nat. Clim. Chang. 3 (2013)
913–918. doi:10.1038/nclimate1944.

[14] E.B. Barbier, Valuing the storm protection service of estuarine and coastal ecosystems,
Ecosyst. Serv. 11 (2015) 32–38. doi:10.1016/j.ecoser.2014.06.010.

[15] S. Das, A.-S. Crépin, Mangroves can provide protection against wind damage during storms,
Estuar. Coast. Shelf Sci. 134 (2013) 98–107. doi:10.1016/j.ecss.2013.09.021.

[16] S. Das, The role of natural ecosystems and socio-economic factors in the vulnerability of
coastal villages to cyclone and storm surge, Nat. Hazards. 64 (2012) 531–546.
doi:10.1007/s11069-012-0255-9.

[17] S. Brown, R.J. Nicholls, C.D. Woodroffe, S. Hanson, J. Hinkel, A.S. Kebede, et al., Sea-Level Rise
Impacts and Responses: A Global Perspective, in: Springer Netherlands, 2013: pp. 117–149.
doi:10.1007/978-94-007-5234-4_5.

[18] P.D. Kunte, N. Jauhari, U. Mehrotra, M. Kotha, A.S. Hursthouse, A.S. Gagnon, Multi-hazards
coastal vulnerability assessment of Goa, India, using geospatial techniques, Ocean Coast.
Manag. 95 (2014) 264–281. doi:10.1016/j.ocecoaman.2014.04.024.

[19] M.E. Hereher, Vulnerability assessment of the Saudi Arabian Red Sea coast to climate change,
Environ. Earth Sci. 75 (2016) 30. doi:10.1007/s12665-015-4835-3.

[20] G.S. Dwarakish, S.A. Vinay, U. Natesan, T. Asano, T. Kakinuma, K. Venkataramana, et al.,
Coastal vulnerability assessment of the future sea level rise in Udupi coastal zone of
Karnataka state, west coast of India, Ocean Coast. Manag. 52 (2009) 467–478.
doi:10.1016/j.ocecoaman.2009.07.007.

[21] R.S. Mahendra, P.C. Mohanty, H. Bisoyi, T.S. Kumar, S. Nayak, Assessment and management

18
 of coastal multi-hazard vulnerability along the Cuddalore–Villupuram, east coast of India
using geospatial techniques, Ocean Coast. Manag. 54 (2011) 302–311.
doi:10.1016/j.ocecoaman.2010.12.008.

[22] T. Wu, X. Hou, Q. Chen, Coastal economic vulnerability to sea level rise of Bohai Rim in China,
Nat. Hazards. 80 (2016) 1231–1241. doi:10.1007/s11069-015-2020-3.

[23] P. Sheik Mujabar, N. Chandrasekar, Coastal erosion hazard and vulnerability assessment for
southern coastal Tamil Nadu of India by using remote sensing and GIS, Nat. Hazards. 69
(2013) 1295–1314. doi:10.1007/s11069-011-9962-x.

[24] K.A. Addo, Assessing Coastal Vulnerability Index to Climate Change: the Case of Accra –
Ghana, J. Coast. Res. 165 (2013) 1892–1897. doi:10.2112/SI65-320.1.

[25] I.O. Adelekan, Vulnerability of poor urban coastal communities to flooding in Lagos, Nigeria,
Environ. Urban. 22 (2010) 433–450. doi:10.1177/0956247810380141.

[26] P. Woodworth, A. Aman, T. Aarup, Sea level monitoring in Africa, African J. Mar. Sci. 29 (2007)
321–330. doi:10.2989/AJMS.2007.29.3.2.332.

[27] R.J. Nicholls, R.S.J. Tol, Impacts and responses to sea-level rise: a global analysis of the SRES
scenarios over the twenty-first century, Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 364
(2006) 1073–1095. doi:10.1098/rsta.2006.1754.

[28] D. Chemane, H. Motta, M. Achimo, Vulnerability of coastal resources to climate changes in

Mozambique: a call for integrated coastal zone management, Ocean Coast. Manag. 37 (1997)
63–83. doi:10.1016/S0964-5691(97)00073-2.

[29] INE, Recenseamento Geral da População e Habitação, (2016).

http://www.ine.gov.mz/operacoes-estatisticas/censos/censo-2007/rgph-2007.

[30] República de Moçambique, Estratégia e plano de acção nacional para restauração de mangal
2015-2020. Ministério da Terra, Ambiente e Desenvolvimento Rural, Centro de
Desenvolvimento Sustentável para as Zonas Costeira, (2015) 43.

[31] República de Moçambique, Estratégia Nacional de Adaptação e Mitigação de Mudanças

Climáticas 2013-2025, Ministério para a Coordenação da Acção Ambiental, (2015) 71.
http://www.cgcmc.gov.mz/pt/politicas-estrategias/94-estrategia-nacional-de-adaptacao-e-
mitigacao-de-mudancas-climaticas.

[32] UNISDR, DesInventar, Disaster Information System, United Nations Office for Disaster Risk

19
 Reduction, (2016). http://www.desinventar.net/.

[33] INE, Estatísticas e Indicadores Sociais, Maputo, Mozambique, 2014.

http://www.ine.gov.mz/estatisticas/estatisticas-demograficas-e-indicadores-
sociais/estatisticas-e-indicadores-sociais/estatisticas-e-indicadores-sociais-2014/view.

[34] INE, Relatório final do inquérito ao orçamento familiar - IOF-2014/15, Maputo, Mozambique,
2016. http://www.ine.gov.mz/operacoes-estatisticas/inqueritos/inquerito-sobre-orcamento-
familiar/relatorio-final-do-inquerito-ao-orcamento-familiar-iof-2014-15/view.

[35] R. Brouwer, M.P. Falcão, Wood fuel consumption in Maputo, Mozambique, Biomass and
Bioenergy. 27 (2004) 233–245. doi:10.1016/j.biombioe.2004.01.005.

[36] World Bank, Mozambique, (2016). http://data.worldbank.org/country/mozambique

(accessed March 19, 2016).

[37] UNDP, Human Development Report 2015 - Work for Human Development, 2015.
http://hdr.undp.org/sites/default/files/2015_human_development_report_1.pdf.

[38] C. Wong, M. Roy, A.K. Duraiappah, Focus on Mozambique, 2005.

[39] F. Tauacale, Water Resources of Mozambique “and the situation of the shared rivers,” in:
River Basin Inf. Syst. Meet., Nairobi, Kenya, 2002.
http://www.unep.org/DEWA/water/RiverBasin/documents/Presentations/Francisco.doc.

[40] F.M.A. Barbosa, C.C. Cuambe, S.O. Bandeira, Status and distribution of mangroves in
Mozambique, South African J. Bot. 67 (2001) 393–398. doi:10.1016/S0254-6299(15)31155-8.

[41] S.J. Armitage, G.A. Botha, G.A.T. Duller, A.G. Wintle, L.P. Rebêlo, F.J. Momade, The formation
and evolution of the barrier islands of Inhaca and Bazaruto, Mozambique, Geomorphology.
82 (2006) 295–308. doi:10.1016/j.geomorph.2006.05.011.

[42] MICOA, Centro de Desenvolvimento Sustentável para as Zonas Costeiras, Ministério para a
Coordenação da Acção Ambiental, (2016).
http://www.zonascosteiras.gov.mz/rubrique.php3?id_rubrique=2O/;Jny 2001. 67 393-398
(accessed June 12, 2016).

[43] A. Mavume, L. Rydberg, M. Rouault, J. Lutjeharms, Climatology and Landfall of Tropical

Cyclones in the South- West Indian Ocean, West. Indian Ocean J. Mar. Sci. 8 (2010).
doi:10.4314/wiojms.v8i1.56672.

[44] P. Bettinger, K.L. Merry, Relative vulnerability of the forests along southeastern African coasts

20
 to cyclones, Singap. J. Trop. Geogr. 33 (2012) 320–334. doi:10.1111/sjtg.12000.

[45] C.J. Matyas, Tropical cyclone formation and motion in the Mozambique Channel, Int. J.
Climatol. 35 (2015) 375–390. doi:10.1002/joc.3985.

[46] K. Asante, R. Brito, G. Brundrit, P. Epstein, A. Fernandes, M. M.R., et al., Main Report: INGC
Climate Change Report: Study on the impact of climate change on disaster risk in
Mozambique, (2009) 55.
http://www.mz.undp.org/content/mozambique/en/home/library/environment_energy/stud
y_on_the_impact_of_climate_change_on_disaster_risk_management.html.

[47] H. Motta, M. Pereira, M. Gonçalves, T. Ridgway, M. Schleyer, Coral reef monitoring in

Mozambique. II: 2000 report. MICOA/CORDIO/ORI/WWF, (2002) 31.

[48] J. Palalane, M. Larson, H. Hanson, D. Juízo, Coastal Erosion in Mozambique: Governing

Processes and Remedial Measures, http://dx.doi.org/10.2112/JCOASTRES-D-14-00020.1.
(2016).

[49] M.E. Moreira, A dinâmica dos sistemas litorais do Sul de Moçambique durante os últimos 30
anos, Finisterra. 40 (2005). doi:10.18055/FINIS1495.

[50] J.V.Q. Langa, Problemas na zona costeira de Moçambique com ênfase para a costa de
Maputo, Rev. Gestão Costeira Integr. 7 (2007) 33–44. doi:10.5894/rgci8.

[51] E.A. Massuanganhe, C. Macamo, L.-O. Westerberg, S. Bandeira, A. Mavume, E. Ribeiro,

Deltaic coasts under climate-related catastrophic events – Insights from the Save River delta,
Mozambique, Ocean Coast. Manag. 116 (2015) 331–340.
doi:10.1016/j.ocecoaman.2015.08.008.

[52] H. Tallis, T. Rickets, A. Guerry, S. Wood, R. Sharp, E. Nelson, et al., InVEST 3.1.1 User’s Guide:
Integrated Valuation of Environmental Services and Tradeoffs. The Natural Capital Project,
(2015). http://ncp-dev.stanford.edu/~dataportal/invest-
releases/documentation/current_release/.

[53] M.G. Chung, H. Kang, S.-U. Choi, Assessment of Coastal Ecosystem Services for Conservation
Strategies in South Korea, PLoS One. 10 (2015) e0133856.
doi:10.1371/journal.pone.0133856.

[54] T. Hopper, M.S. Meixler, Modeling Coastal Vulnerability through Space and Time, PLoS One.
11 (2016) e0163495. doi:10.1371/journal.pone.0163495.

21
[55] E.B. Barbier, The protective service of mangrove ecosystems: A review of valuation methods,
Mar. Pollut. Bull. 109 (2016) 676–681. doi:10.1016/j.marpolbul.2016.01.033.

[56] B.T. van Zanten, P.J.H. van Beukering, A.J. Wagtendonk, Coastal protection by coral reefs: A
framework for spatial assessment and economic valuation, Ocean Coast. Manag. 96 (2014)
94–103. doi:10.1016/j.ocecoaman.2014.05.001.

[57] M.B.S.F. Costa, M. Araújo, T.C.M. Araújo, E. Siegle, Influence of reef geometry on wave
attenuation on a Brazilian coral reef, Geomorphology. 253 (2016) 318–327.
doi:10.1016/j.geomorph.2015.11.001.

[58] M.E. Hanley, S.P.G. Hoggart, D.J. Simmonds, A. Bichot, M.A. Colangelo, F. Bozzeda, et al.,
Shifting sands? Coastal protection by sand banks, beaches and dunes, Coast. Eng. 87 (2014)
136–146. doi:10.1016/j.coastaleng.2013.10.020.

[59] J.P. Doody, Sand Dune Conservation, Management and Restoration, Springer Netherlands,
Dordrecht, 2013. doi:10.1007/978-94-007-4731-9.

[60] ESRI, ArcGIS, (2014). www.esri.com.

[61] NOAA, Global Self-consistent, Hierarchical, High-resolution Geography Database (GSHHG),

National Oceanic and Atmospheric Administration, Natl. Ocean. Atmos. Adm. (2015).
http://www.ngdc.noaa.gov/mgg/shorelines/ (accessed March 3, 2016).

[62] NASA/METI, ASTER Global DEM data (GDEM V2), (2012). http://earthexplorer.usgs.gov/.

[63] H.L. Tolman, User Manual and System Documentation of WAVEWATCH III version 3.14
Technical Note, 2009.

[64] GEBCO, General Bathymetric Chart of the Oceans (GEBCO), (2014). http://www.gebco.net/.

[65] CIESIN, Center for International Earth Science Information Network (CIESIN), Columbia
University; and Centro Internacional de Agricultura Tropical (CIAT). Gridded Population of the
World Version 3 (GPWv3). Palisades, NY: Socioeconomic Data and Applications Center,
(2005). doi:http://sedac.ciesin.columbia.edu/plue/gpw.

[66] OECD, Investment Policy Reviews: Mozambique, (2013) 29.

https://www.oecd.org/daf/inv/investment-policy/IPR-Mozambique-Oct2013-Summary.pdf.

[67] C.C.F. Macamo, E. Massuanganhe, D.K. Nicolau, S.O. Bandeira, J.B. Adams, Mangrove’s
response to cyclone Eline (2000): What is happening 14 years later, Aquat. Bot. 134 (2016)
10–17. doi:10.1016/j.aquabot.2016.05.004.

22
[68] A. Fekete, Validation of a social vulnerability index in context to river-floods in Germany, Nat.
Hazards Earth Syst. Sci. 9 (2009) 393–403. www.nat-hazards-earth-syst-sci.net/9/393/2009/
(accessed August 11, 2016).

[69] V.N. Martins, P. Cabral, D. Sousa e Silva, Urban modelling for seismic prone areas: the case
study of Vila Franca do Campo (Azores Archipelago, Portugal), Nat. Hazards Earth Syst. Sci. 12
(2012) 2731–2741. doi:10.5194/nhess-12-2731-2012.

[70] C.J. Matyas, J.A. Silva, Extreme weather and economic well-being in rural Mozambique, Nat.
Hazards. 66 (2013) 31–49. doi:10.1007/s11069-011-0064-6.

[71] N. Viana, Coping with floods: Mozambique and Britain, Contemp. Rev. 290 (2008) 493.
http://connection.ebscohost.com/c/articles/36321574/coping-floods-mozambique-britain.

23