Environmental Vulnerability Mapping in Liberia _Unep 2006

Environmental Vulnerability Mapping in LiberiaWork in Progress (July 2006)

For the latest version of this document visit http://postconflict.unep.ch/liberia/displacement

United Nations Environment ProgrammePost-Conflict Branch

15, Chemin des AnémonesCH-1219 Châtelaine, Geneva

SwitzerlandTel.: +41 (0)22 917 85 30Fax: +41 (0)22 917 80 64http://postconflict.unep.ch

http://postconflict.unep.ch/

http://postconflict.unep.ch/liberia/displacement

Liberia Environmental Vulnerability Mapping

Table of Contents

1 Introduction.....................................................................................................................................1

1.1 Definitions............................................................................................................................... 2

2 Summary........................................................................................................................................ 3

3 Developing an Environmental Vulnerability Methodology.............................................................. 4

3.1 Context................................................................................................................................... 4

3.2 Understanding Environmental Vulnerability............................................................................4

3.3 Available Data Sources.......................................................................................................... 5

3.3.1 Terrain Morphology and Derived Hydrology................................................................... 6

3.3.2 Land Cover......................................................................................................................9

3.4 A Simple Model of Environmental Vulnerability....................................................................11

3.4.1 Likelihood of Impact – Proximity and Accessibility........................................................11

3.4.2 Extent – Degree of Loss of Value................................................................................. 14

3.5 Protected Areas.................................................................................................................... 20

3.6 Ecosystem Service Value Map ............................................................................................ 20

3.7 Combining Hazard Impact Potential and Value in a Vulnerability Model..............................22

3.7.1 Interpreting the Vulnerability Map................................................................................. 24

3.7.2 Vulnerability Map Examples..........................................................................................25

4 Running the Model.......................................................................................................................29

4.1 Step 1 – Data Preparation - Camp Locations.......................................................................29

4.2 Step 2 - Definition of Land Characteristics from Existing Sources....................................... 32

4.3 Step 3 – Calculating Proximity and Accessibility.................................................................. 32

4.4 Step 4 – Definition of Environmental Value – Dimension 2.................................................. 35

4.5 Step 5 Combining Hazard Impact Potential and Value in a Vulnerability Model.................. 36

5 Drainage and Catchment Analysis...............................................................................................38

5.1 Ecological Service Value by Catchment...............................................................................40

5.2 Drainage Concentration........................................................................................................41

6 Cartographic Vulnerability Maps.................................................................................................. 43

7 References................................................................................................................................... 50

i


Acronyms

Acronym MeaningCGIAR Consultative Group on International Agricultural ResearchCI Conservation InternationalDCW Digital Chart of the WorldIDP Internally Displaced PersonsFRAME Framework for Assessing, Monitoring and Evaluating the Environment

in Refugee-related Operations

FRM Foret Ressources ManagementGIS Geographical Information SystemsHIC Humanitarian Information Centre, Liberia,

http://www.humanitarianinfo.org/liberia/MEA Millennium Ecosystem AssessmentNHM Natural History Museum, LondonSRTM Shuttle Radar Topography MissionUNHCR United Nations High Commissioner for RefugeesUNEP United Nations Environment ProgrammeWB World Bank

ii

http://www.humanitarianinfo.org/liberia/


iii

Foreword This report was compiled as part of UNEP’s work on the environmental considerations of human displacement in Liberia. Within a broader project entitled ‘Strengthening Capacities for the Integration of the Environmental Dimension in Refugee and IDP Settlements and Flows in Angola, Liberia and Sierra Leone’, The Liberia work included a basic needs assessments and a review of existing literature, two capacity building workshops and the publication of UNEP’s Environmental Considerations of Human Displacement in Liberia: A Guide for decision makers and practitioners. The project was undertaken with financial assistance from the governments of Norway and Sweden. The environmental implications arising from refugee or IDP camps obviously depend on the context of the camps and the vulnerability of the environment around the camps. In order to assist with the appropriate siting of camps (and also of areas earmarked for organized resettlement) UNEP developed a conceptual model of environmental vulnerability in relation to camps and then used GIS to develop a spatial model that predicts both the location of potential impacts and their likely magnitude. This vulnerability model is based on two main sources of information: the value of the ecosystem services provided by the land at a given geographic location and the accessibility from a camp to that land as a measure of the likelihood of it being impacted. In the GIS analysis these two components (ecosystem services value and accessibility) have been combined into a single map indicating vulnerability of ecosystem services to impact from camps. The maps that were produced from this methodology were passed on to the United Nations Country Team as part of the contingency planning to deal with a potential influx of refugees from Cote d’Ivoire (in early 2006). This report provides an overview of the methodology adopted in the above process. As additional funding becomes available UNEP intends to further develop the methodology began in this report. Consequently the current report should be seen as a work in progress and future versions of the methodology will be made available at: (http://postconflict.unep.ch/liberia/displacement/). If you have any comments on this report please email them to [email protected]


iv

Acknowledgements As noted above this report was developed within the context of UNEP’s work on the environmental considerations of human displacement in Liberia and the vulnerability mapping team included: Kay Farmer, Programme Manager, UNEP Liberia, Tim Richards, Conservation Technology Ltd, Stone, David Senior Environmental Expert, UNEP Geneva, Richard Wood, Technical Coordinator, UNEP Geneva and Grant Wroe-Street, Project Coordinator, UNEP Geneva The contribution of data and technical support provided by Margaret Hall, Humanitarian Information Centre, OCHA, Liberia is gratefully acknowledged. In addition, the following people were asked to review an earlier version of the vulnerability mapping work: Jock Baker Care International, Einar Bjorgo, UNOSAT, Tyler Christie, Conservation International, James.Kamara, UNEP, Mengesha Kebede UNHCR, Christian.Lambrechts, UNEP, Andrew Mayne, UNHCR, Valentine Ndibalema, UNHCR and Luc St-Pierre, UNHCR. We gratefully acknowledge all comments received.


1 IntroductionThe ending of civil war in Liberia, the signing of the Accra Peace Agreement in 2003 and the recent

successful elections all give cause for hope for the normalisation of civil society and the opportunity

for the people of Liberia to start to rebuild their lives. However, many people remain in Internally

Displaced Persons (IDP) and Refugee camps and there continues to be the threat of refugee

influxes from neighbouring countries. Liberia is rich in natural resources and still has large, intact

natural ecosystems. These ecosystems are vulnerable, however, to degradation by threats

including direct resource exploitation by both industries and populations.

The necessary establishment of IDP and Refugee camps to meet the immediate needs of

potentially large numbers of displaced people has direct and indirect impacts on the environment.

This report describes work undertaken to map the vulnerability of the Liberian environment to the

impacts of IDP and Refugee camps on the country's natural environment.

Vulnerability implies both exposure to a threat and the potential loss of environmental assets or

value as a result of that threat. As such, Vulnerability Mapping is a task fraught with difficulties in

that the threat of new camps is not well defined or controllable and how environments can be

valued is the subject of much debate and discussion. Even areas that may have a environmental

or ecological service value that is perceived to be low can be adversely effected by over utilisation

of local resources.

This report attempts to develop a methodology that is both consistent and robust and which

provides useful practical information regarding the implications of camp location. Furthermore the

methodology should be generic and it should be possible for it to be applied elsewhere.

The location of camps is ideally determined by a combination of security, logistics, political, social

and environmental factors, consequently in this report we do not attempt to specify where camps

should be located. The UNHCR and CARE International have recently published a Framework for

Assessing, Monitoring and Evaluating the Environment in Refugee-related Operations (FRAME),

(UNHCR, 2005). This report further develops a Geographical Information System methodology for

identifying potential Refugee / IDP environmental impacts in Liberia, West Africa.

1


1.1 DefinitionsTable 1 contains a definition of terms used throughout the report. An attempt has been made in the

text of the report to be precise with the terms used. The key term is vulnerability and a generic

definition has been adopted from Tobin and Montz (1997). This is a useful definition that can be

successfully applied to environmental vulnerability.

Table 1: Definition of Terms

Threat The possibility of a Refugee or IDP camp being sited in a given locality.

Hazard Environmental degradation caused directly or indirectly by a camp.

Vulnerability “The extent to which a community, structure, service or geographic area is likely

to be damaged or disrupted by the impact of a particular hazard.” (Tobin &

Montz 1997).

Accessibility The accessibility of a given location from a camp or camps. Accessibility does

not necessarily imply proximity as barriers and impediments exist within the

landscape.

Value An attribute of a locality; based on a value system that values one

environmental attribute above another. The environment that will be lost or

damaged.

Cost Surface, or Friction Surface

A cost surface specifies the cost, in arbitrary units, of traversing one unit distance of the landscape.

Cost Distance The accumulated cost, in cost surface units, of moving across a cost-surface from A to B.

2


2 SummaryThe methodology can be summarised in the following 5 steps:

Table 2: Summary of Methodological Steps

Data

Preparation

Step 1 Definition of existing or potential camp locations

Step 2 Definition of land characteristics from existing sources:

• Landscape morphology and hydrology

• Land cover

Definition of

Accessibility to

the environment

from camp

locations

Step 3 • Definition of a cost-surface, based on land cover, which

defines impediments to moving across the landscape.

• Definition of a cost-distance map that estimates the cost to

move from A (camp or potential camp locations) to B (all

other locations) in cost-distance units.

• Analysis of downstream impacts of drainage concentrationDefinition of

Environmental

Value

Step 4 Definition of a measure of environmental value based on land cover,

ecological service value and protection status

Vulnerability

Mapping

Step 5 Combination of Hazard Impact Potential and Value in a Vulnerability

Model.

3


3 Developing an Environmental Vulnerability Methodology 3.1 ContextThe context of this analysis is the environmental

impact of Refugee and Internally Displaced Persons

(IDP) Camps in Liberia. Such camps may host

sizeable populations of up to 20,000 people and may

have a significant impact on the environment of the

camp and its environs. The impact of large camps is

clearly visible by satellites from space as vegetation

cover is removed from within the camp boundaries.

Depending on whether resources, such as food,

building materials and fuel wood, are brought into the

camp a camp's population will also have an a direct

impact on the surrounding environment and,

potentially, an indirect impact downstream from the

camp. Camps are visible on this Landsat image as

white geometric areas devoid of vegetation cover.

3.2 Understanding Environmental VulnerabilityA definition of vulnerability given by Tobin & Montz (1997) states that vulnerability is:

“the extent to which a community, structure, service or geographic area is likely to be

damaged or disrupted by the impact of a particular hazard”

In the context of Refugee / IDP camps in Liberia this definition can be reiterated as follows to clarify

its meaning.

• The geographic area is the area effected by the impact of the camp, either within the camp

boundary or outside the camp, or, indirectly, downstream.

• The extent refers to degree to which the geographic area may be damaged. A measure of

the extent of damage indicates that the value of the geographic area has been reduced in

some way.1

• The likelihood of damage refers to the probability of the camp having a negative impact

and can be equated with the accessibility of a given area from the camp and its sensitivity

to negative impacts.

• The impact refers to the degradation or damaged caused, for example through

1 Tobin and Montz (1997) use the word extent to mean an amount of damage, not the geographical extent

of the damage.

4

Figure 1: High density of camps to the

north of Monrovia. Camps are marked

as yellow stars.


unsustainable utilisation of resources.

• The hazard is the cause of the impact, in this case the siting of a camp and its population.

From the basis of this definition it is possible to construct a conceptual model of environmental

vulnerability in relation to camps and to build a spatial model, using Geographical Information

Systems (GIS), to estimate what those impacts actually are, or what they may be in a predictive

sense.

3.3 Available Data SourcesReadily available data sources for Liberia are limited. The data necessary to build a vulnerability

model are primarily concerned with accessibility and measures of environmental value.

Data Set Date DescriptionExisting camp locations

Camp locations taken from table provides by UNEP Liberia Mission

Contingency camp locations

The sites identified as potential crossing points for a possible influx from Cote d'Ivoire are those under discussion in the UN contingency planning process currently under way in Liberia. The responsibility for identifying future camp sites rests with the Government of Liberia. Contingency crossing points, way stations and camps should be considered as provisional.

Land Use 2003 / 2004

World Bank / FRM, 2004, Current State of the Forest Cover in Liberia. Visual interpretation of Landsat imagery into 11 land use classes.

Roads Various Roads data derived from the Digital Chart of the World, HIC, UNEP and manual interpretation of Landsat imagery

Terrain•Slope•Stream lines•Catchments

SRTM 90m database from the CGIAR Consortium for Spatial Information (http://srtm.csi.cgiar.org/). Various derived data sets can be calculated from this data set, such as stream lines, drainage basins, etc.

Value of ecosystem services

1997 The estimated value of 17 ecosystem services, published by Costanza et al., 1997.

Protected Areas 2005 World Database on Protected Areas, Conservation International, Liberia.

Of these data sets the land use and terrain data merit further explanation.

5

http://srtm.csi.cgiar.org/


3.3.1 Terrain Morphology and Derived HydrologyThe morphology of the landscape is available from Shuttle Radar Topographic Mission (SRTM)

data. These data were acquired by the NASA Space Shuttle in February 2000 using a dual radar

systems to extract terrain heights using interferometry. The data have a nominal 30m horizontal

resolution and are available in the public domain at 90m horizontal resolution and 1m vertical

resolution (< 16m vertical error). The data utilised are from the CGIAR Consortium for Spatial

Information (http://srtm.csi.cgiar.org/) which has pre-processed the data to interpolate missing

values. An evaluation of SRTM data quality, in comparison with other sources of terrain data such

as 1:50,000 topographic maps, is provided by Jarvis et al. (2004).

This data set provides a consistent and accurate terrain model at the national level. A number of

derived data layers can be generated from these data that are of use in vulnerability mapping,

including:

• Major drainage basins

• Minor drainage basins

6

Figure 2: Example of SRTM terrain data, northern Liberia. The level of detail

of the SRTM data is apparent. In this image white areas have a higher

altitude than dark areas.

http://srtm.csi.cgiar.org/


• Stream lines

• Flow direction

• Flow accumulation

• Hill shading

• Contours

A national Liberian SRTM data set has been compiled and cleaned to make it hydrologically

consistent – i.e. without sinks that do not drain. The derived data sets are used in place of more

traditional data such as Digital Chart of the World (DCW). In very flat areas near the coast, where

it is not possible to accurately derive stream channels using automatic techniques, the location of

the channels has been corrected to Landsat imagery.

7

Figure 3: Terrain of Liberia Figure 4: Terrain and major stream channels


Other derived terrain measures, such as slope, are also utilised in the following analysis.

The derived terrain measures were compiled using HydroTools for ArcView, Schäuble, H, (2003).

8

Figure 5: Terrain and minor catchmentsTerrain Legend in

metres.


3.3.2 Land CoverThe most recent land cover information for Liberia which is both consistent and that provides full

national coverage is the World Bank / Foret Ressources Management report (WB/FRM, 2004) on

the Current State of the Forest Cover in Liberia which contains a land cover map of Liberia

compiled by visual interpretation of Landsat imagery dating from 2003 and 2004. The map was

digitised at a nominal scale of 1:100,000 with a minimum mapping unit of 1,000 ha.

These authors define a land cover classification for mapping forest cover consisting of the following

land cover classes:

Table 3: World Bank / FRM Land Cover Classes

Class No Class1 Urban (both urban and rural settlements)

2.1 Predominantly rural agricultural domain2.2 Agricultural area with small forest presence2.3 Mixed agriculture and forest3.1 Agriculture degraded forest3.2 Open dense forest3.3 Closed dense forest5 Free water6 Savannah or bare soil7 Coastal (littoral) ecosystem complex8 Agro-industrial plantation

These classes are represented in the map illustrated in Figure 6. The map covers all of Liberia

and is the most recent and consistent land cover map of the country. The map was produced for

the purpose of assessing forest cover in Liberia and is also useful for the purposes of

environmental vulnerability mapping. However, it is not ideal and it is likely that both the land cover

classification and the minimum mapping scale could be adjusted to produce a map more directly

relevant to vulnerability mapping related to IDP / Refugee camps.

9


10

Figure 6: World Band / FRM Land Cover Map of Liberia, 2004 (Source HIC 2004)


3.4 A Simple Model of Environmental VulnerabilityBased on the definition of vulnerability, see Table 1, the vulnerability model is based on two

information dimensions – 1) the extent or degree of loss at a given geographic location and 2) the

likelihood of a geographic area being impacted – related to the accessibility from a camp to that

location. The model is extended further later in this report through comparison with water

catchments.

3.4.1 Likelihood of Impact – Proximity and AccessibilityThe likelihood of a geographic area being damaged or disrupted by the impact of a camp is

proportional to its proximity to a camp and the accessibility from a camp to the geographic area.

Proximity does not necessarily imply accessibility. In principle an area that is close to a camp is

more likely to be impacted than an area far away from a camp, however, even if a geographic area

is in close proximity to a camp it may not be possible to access the area because of an impediment

in the landscape such as an impassable river.

In this example location B is accessible from camp A but

location C is not directly accessible because of a river

barrier, even though B and C are roughly the same distance

from the camp (have the same proximity). To get to C from A

it is necessary to cross the river by the bridge at D which

involves a journey of approximately twice the distance as to

get to B.

Rather than using the distance “as the crow flies” (Euclidian

Distance) from A to C it is more useful to use the Cost-

Distance or the estimated cost involved in getting from A to B

or C. The Cost-Distance can be calculated by assigning

costs (frictions) to land cover types according to the ease of

traversing them. For example it is easy to travel along a road

whereas it is very difficult, or even impossible, to cross the

river. However it is possible to cross the river on a road, i.e.

over a bridge.

The Cost-Distance from all of the camps illustrates how this

concept can be used to estimate accessibility from one point

(or many) to any other according to the ease or difficulty with

which it is possible to traverse the landscape. In this

illustration short cost-distances are shown in light tones

whereas long cost-distances are shown in dark tones. The

cost to get from A to C is approximately twice the cost to get

11

A

BC


from A to B.

Cost-Distance is a term used in the GIS community and may imply an economic cost

in monetary units. If a person is driving a car then there are standard monetary costs

that can be applied for driving a given distance – fuel consumption, depreciation, etc.

However, the cost does not have to be represented in monetary terms and can also be

thought of as the amount of effort or energy needed to move around. This is more

appropriate if a person is walking, as most IDP / Refugee camp inhabitants are likely to

be doing.

To calculate Cost-Distance it is necessary to start with a Cost map that estimates the relative costs

necessary to traverse the landscape. A set of six classes has been used that categorises the cost

of movement a unit distance from Very Easy (low cost) to Impossible (very high cost).

Table 4: Relative Costs to Traverse a Unit Distance

Weight Movement Category1 Very Easy2 Easy3 Not Easy4 Hard5 Very Hard

100 Impossible

By relating the cost of movement across the landscape to land cover classes it is possible to

compile a travel Cost Map2. In turn this can be used to calculate Cost-Distances between points or

areas of interest.

The road network is used to define road surfaces that can be crossed with a unit cost, i.e. 1. The

river network, derived from the SRTM data, is used to define rivers that are impossible to be

crossed, other than by a bridge. For this purpose the river network was extracted using a flow

accumulation threshold of 10,000 pixels to extract only the major rivers. In this way minor rivers

and streams were excluded. The costs assigned to traversing the landscape are shown in Table 2.

2 Cost Values are also known as Friction Values

12


The cost units are arbitrary and defined as 1 cost unit to traverse 92m along a road3.

Table 5: Landscape Traversal Cost Values

Class No Class1 Urban (both urban and rural

settlements)2

2.1 Predominantly rural agricultural domain

2

2.2 Agricultural area with small forest presence

3

2.3 Mixed agriculture and forest 33.1 Agriculture degraded forest 33.2 Open dense forest 43.3 Closed dense forest 55 Free water 1006 Savannah or bare soil 17 Coastal (littoral) ecosystem

complex5

8 Agro-industrial plantation 2

3 92m is the horizontal spatial resolution of the SRTM data.

13

Figure 7: Cost Surface indicating the ease of traversing the landscape relative to walking on a road.


Farrow and Nelson (2001) provide a review and tools for accessibility modelling4.

The more detailed the spatial information available the better the cost-distance measure will model

the actual movement of people.

3.4.2 Extent – Degree of Loss of ValueTo measure the extent to which a geographic area may be effected by the impacts of a camp

presupposes 1) a value system that attributes worth to the area's environment and 2) the ability to

measure the degree of loss or degradation compared either to itself or to another area. Valuing the

environment is a complex and potentially contentious issue which is difficult to resolve. However, to

assess vulnerability and the associated extent or potential degree of loss of value it is necessary to

base our analysis on fixed assumptions of ecosystem service value.

Concepts of value, value systems and valuation have a long history, dating back to Aristotle, and

having meaning in a variety of disciplines (Farber et al. 2005). Conventional economic theory has

generally undervalued or even ignored environmental and ecological service values (Chichilnisky,

1996). As the Earth's natural capital is being expended the issue of ecological valuation has come

to the fore and is now an active area of research. In particular the Millennium Ecosystem

Assessment (MEA, 2005) has recently made much use of the concepts of ecosystem valuation in

the context of ecosystems and human well-being (MEA, 2005).

MEA presents the framework of Total Economic Value (TEV) after Pearce and Warford (1993), as

shown in Figure 9. In this framework TEV is comprised of Use Value, including Direct Use Value,

Indirect Use Value and Option Value, and Non-Use or Existence Value.

4 Note that software tools have been developed for this specific study and Farrow and Nelson's tools have

not been utilised.

14

Figure 8: Cost Surface Detail

http://www.maweb.org/


• Direct use values are ecosystem services used for both consumptive and non-consumptive

purposes.

• Indirect use values are intermediate inputs for final goods and services, such as water, soil

nutrients, control services, etc.

• Option values are values attributed to the option of utilising the ecosystem in the future.

(Source MEA, 2005)

The transformation of land cover from one type to another (or one ecosystem to another) involves

a corresponding change in value. Figure 10 illustrates the value changes involved in deforestation.

15

Figure 9: Total Use Value (Source: MEA, 2005)


As indicated above - in developing an environmental vulnerability map it is necessary to assess the

extent, or amount, of potential damage and associated potential conversion from one ecosystem to

another. This implies both that the value of the geographic area is known, or that it is known

relative to another geographic area, and that the value system defining the valuation is also known.

This challenge must be confronted in order to produce a measure of vulnerability that is more than

just a statement of accessibility.

Knowledge of accessibility to an area is not sufficient to make a judgement as to whether that area

is more or less vulnerable than another equally accessible area. The second component of our

vulnerability model is a measure of environmental value in order to make this judgement. Costanza

et al. (1997) state that “although ecosystem valuation is certainly difficult and fraught with

uncertainties, one choice we do not have is whether or not to do it”. These authors have taken on

this task and have estimated the economic value of 17 ecosystems services based on a synthesis

of results from more that 100 separate studies. The 17 ecosystem services considered are shown

in Table 6 and their estimated values are given in Table 7.

16

Figure 10: Ecosystem Conversion (Source: MEA, 2005)


Table 6: Ecosystem Services and Function (Source Costanza et al. 1997)

No Ecosystem Service

Ecosystem Functions

1 Gas regulation Regulation of atmospheric chemical composition

2 Climate regulation Regulation of global temperature, precipitation, and other biological mediated climatic processes at global or local levels

3 Disturbance regulation

Capacitance, damping and integrity of ecosystem response to environmental fluctuations

4 Water regulation Regulation of hydrological flows

5 Water supply Storage and retention of water

6 Erosion control and sediment retention

Retention of soil within an ecosystem

7 Soil formation Soil formation processes

8 Nutrient cycling Storage, internal cycling, processing and acquisition of nutrients

9 Waste treatment Recovery of mobile nutrients and removal or breakdown of excess or xenic nutrients and compounds

10 Pollination Movement of floral gametes

11 Biological control Trophic-dynamic regulations of populations

12 Refugia Habitat for resident and transient populations

13 Food production That portion of gross primary production extractable as food

14 Raw materials That portion of gross primary production extractable as raw materials

15 Genetic resources Sources of unique biological materials and products

16 Recreation Providing opportunities for recreational activities

17 Cultural Providing opportunities for non-commercial uses

(Source Costanza et al. 1997)

Table 7: Average Global Value of Annual Ecosystem Services

Ecosystem Annual Local Value (US$/ha/year, 1994 values)

Coastal 4,052

Open Ocean 252

Wetlands 14,785

Tropical Forest 2,007

Lakes, Rivers 8,498

Temperate/Boreal Forest 302

Grasslands 232

Cropland 92

Estuaries 22,832

Tidal marsh / mangrove 9,990

Swamps / floodplains 19,580

(Source Costanza et al. 1997)

17


The ecosystem service values given in Table 7, together with the World Bank land cover maps

described below, are used as the basis of a Liberian ecosystem valuation system as an input to the

vulnerability assessment. The land cover map is used to spatialise the values of the ecosystem

services.

There are differences of opinion as to the monetary value of ecosystems (Costanza et al., 1997,

Pimentel 1997), however the valuation can be used as relative indicators of their value and provide

an objective way of compiling an environmental value map of Liberia.

The values given for each ecosystem are in US$ per hectare per year. The WB/FRM land cover

classification is shown in Table 8 together with more convenient normalised annual local values.

As there is not yet consensus about the precise monetary value of ecosystem services and that in

creating the vulnerability map it is not necessary to assign absolute values in monetary terms to

the land cover classes, relative weights have been used instead.

The WB/FRM classification has more sub-classes per ecosystem than Costanza (1997) and

estimated values are given for these classes are given below in Table 8.

Table 8. Land Cover Class Environmental Value Weights

Land Cover Class Value WeightUrban 1Predominantly rural agricultural domain 8Agricultural area with small forest

presence

10

Mixed agriculture and forest 33Agriculture degraded forest 50Open dense forest 50Closed dense forest 66Free water 281Savannah or bare soil 8Coastal ecosystem complex 134Wetlands 489Agro-industrial plantation 28

(Shaded land cover classes indicate those that are considered to be natural and

whose values are increased by also being located in current or proposed

protected areas.)

The land cover classes of the WB/FRM map are assigned these values on a per pixel basis,

resulting in a map of the relative ecosystem service values provided by each land cover class.

Note that the analysis has been undertaken at the spatial resolution of the SRTM data which is

nominally 92 m, or 0.85 hectares.

18


There is currently much debate about the need to include the value of ecosystem services in policy

and planning decision making, however, there is as yet little published information to go on, in

Liberia and elsewhere. As ecological services values for Liberia were not available to the authors

the values published by Costanza et al. (1997) have been used instead. With this in mind the

valuation has been used as an indicator of the relative values of ecosystem services rather than as

an absolute measure of their values in monetary units.

The task of ecological valuation in respect of environmental vulnerability mapping is further

complicated by issues of scale. At the national level Liberia is rich in natural resources and has a

range of ecosystems that provide high value services to society. However, at the scale of the IDP

or refugee camp only a limited range of the national ecosystem spectrum is likely to be present

locally. It follows that only a limited range of ecosystem services may be available at the camp level

and that what may seem a low value environment within the national context is actually a high

value environment in the local context. In this report the terms “lower-value” and “higher-value” are

used in terms of the national spectrum of ecosystem service values . As we have seen what may

be termed “lower-value” at the national scale could be termed “higher-value” at the local scale,

however this distinction is not generally made in the text.

It should also be noted that whilst Liberia is rich in ecosystems, society is poor and needs to

develop economically, and cannot necessarily afford to forego converting some natural ecosystems

into more economic forms of production including IDP and Refugee camps. There is therefore a

trade-off to be made between retaining the “value” of natural ecosystems, and the services they

provide, and the benefits of converting the ecosystem to another use. Informed planning decisions

must balance this trade-off of values. Uninformed planning, or the absence of planning altogether,

can lead to the spectre of a third alternative where ecosystem degradation is such that the

environment can support neither ecosystem services nor economic activity. The objective of the

vulnerability mapping is to highlight those areas where the trade-off, or loss of value, is

disproportionally high with respect to IDP and Refugee camps.

19


3.5 Protected AreasFurther information pertaining to the perceived value of the land cover classes is available in the

form of the location of established and proposed protected areas and biodiversity priority zones.

These maps indicate areas that have been identified as being of high value for the conservation of

natural resources and have been developed primarily by Conservation International and the

Natural History Museum of London (NHM).

These protected area maps are incorporated into the analysis by doubling the ecosystem service

value weights inside the protected areas for those classes that represent natural ecosystems, i.e.

excluding agriculture, degraded forest, industrial plantations and urban areas.

The protected areas used are those listed by WB/FRM (2004), namely:

Existing Protected Areas

• Sapo National Park

• East Nimba Nature Reserve

Proposed Protected Areas

• Lake Piso

• Lofa-Mano

• Wologizi East

• Wonegizi West

• West Nimba

• Cesto-Gbi

• Cestos-Sehnkwekn

• Grebo

• Marshall

Existing and proposed protected areas are given equal weight. The biodiversity priority areas maps

has not been utilised because it is very correlated with the protected areas data and has very

general boundaries. Also the National Forest areas are not given any additional value weight.

3.6 Ecosystem Service Value Map The Ecosystem Service Value Map is shown in Figure 12. Areas are valued according to Table 8

and coloured according to the vulnerability classification described below. Areas are coloured with

increasing value from cyan (values 0 to 50), through magenta (values 51 to 100), to red (101 to

150) and lastly green (151 and above). From Table 8 it can be seen that cyan covers mainly

20

Figure 11: Protected Areas (green) &

Proposed Protected Areas (cyan)


modified ecosystems such as agriculture, magenta covers mainly natural undisturbed ecosystems

such as closed dense forest and red covers protected undisturbed ecosystems. Very high values,

shown in green, are protected wetlands.

In addition to providing an input to the vulnerability assessment the ecosystem service value map

could be used for other purposes such as estimating the value of Liberia’s ecosystems by county

or by watershed.

21

Figure 12: Ecosystem Service Value Map excluding Protected Area weighting


3.7 Combining Hazard Impact Potential and Value in a Vulnerability ModelTo complete the Environmental Vulnerability Map the Cost-Distance and Ecosystem Service Value

maps must be combined to show the range of combinations of potential exposure to the hazard

and ecosystem service values.

Plotted on a graph it is clear to see that at the four corners:

• A – has high ecological service value and is near to a camp

• B – has high ecological service value and is far from a camp

• C – has low ecological service value and is far from a camp

• D – has low ecological service value and is near to a camp

The corner that is most vulnerable is A. The corner that is least vulnerable is C. The line from C to

A represents an axis of increasing environmental vulnerability.

The vulnerability colour key colours the vulnerability axis from green (C) through cyan, blue

and magenta to red (A).

At right-angles to the vulnerability axis is an axis of equal vulnerability. For example, a line from D to B represents areas that have medium vulnerability. Areas that have very low ecological service

value and that are near a camp (D) have a medium level environmental vulnerability; as do areas

22

Figure 13: Vulnerability Model


that have very high ecological service value but which are very far from a camp (B). In the colour

key all values along this line have the same colour (blue) but the shade of the colour changes from

dark blue at D to light blue at B. The complete colour key is shown below in Figure 14 and covers

all possible combinations of the data.

23

Figure 14: Vulnerability Model - vulnerability axis and

equal vulnerability zones 1 - 8

Figure 15: Vulnerability Model


3.7.1 Interpreting the Vulnerability MapThe colour table in Figure 15 may seem rather abstract, however, with the aid of an example it is

quite simple to interpret.

This example shows a camp in an agricultural zone with

low forest presence and that has a relatively low ecological

services values. The camp is located at the red dot at the

centre of the vulnerability zone. Near the camp the

vulnerability map is dark blue, indicating a medium level of

vulnerability with low ecological service values near the

camp. Further away from the camp the vulnerability map is

green, indicating low ecological service values that are far

from the camp and which have a low vulnerability rating.

The colour table should be used to interpret the

vulnerability classes. The greener the colour is the lower

the vulnerability; the redder the colour is the greater the

vulnerability.

This example is in an environment of mixed agriculture and

forest, and closed dense forest. There are no green areas

and the blues, cyans and magenta tones indicate medium

to high vulnerability ratings.

Note that low ecological service value does not mean no ecological service value, just that it is

lower than the other categories.

As mentioned above protected areas are another source of information relating to the perceived

value of an area. Protected areas may be located where they are for a range of reasons, but they

generally contain features that are highly valued for other than economic reasons, such as

biodiversity, aesthetic value, etc. They are given additional weight in the vulnerability analysis by

multiplying their ecological service values by a factor of two. In the vulnerability map the high

ecological service values are generally associated with either protected ecosystems or wetlands.

24


3.7.2 Vulnerability Map ExamplesThe environs of Monrovia have a dense

concentration of existing IDP and Refugee camps,

both around the city and strung out to the north east

along the road to Gbarnga. To the north of the city

and the St Paul river there are nine camps within a

radius of just 4 kilometres, constrained on three

sides by the sea to the west, the St Paul river to the

south and another smaller river to the north. The

other camps are quite evenly spread along the road

between 4 and 10 kilometres apart.

The close proximity of these camps means that there is an extensive contiguous vulnerability zone

stretching from the coast for about 120 kilometres inland. This zone passes through rural

agricultural areas and an extensive area of industrial plantation. Near Gbarnga it also skirts around

areas of mixed agriculture and forest.

The map on the following page shows the vulnerability map superimposed on top of the land cover,

together with the existing camps. The vulnerability zones range from low vulnerability in green,

through medium vulnerability areas in blues, to some small areas of higher vulnerability in magenta

and red.

In this part of the country the highest ecosystem service values are found along the coast. The

Buchanan IDP camp is the only camp located in isolation in the south east of the map. This camp

is in close proximity to a coastal area of high ecosystem service values that is coloured red and

classed as highly vulnerable.

What may not be immediately apparent from the vulnerability map, but which can be seen from the

river network, is the fact that 15 of the existing camps are upstream of the proposed Marshall

protected area located on the coast.

This map illustrates the point that in order to minimise environmental impacts camps should ideally

be well spaced in the landscape to prevent their environmental footprints converging and imposing

increased environmental pressures. Furthermore, attention should be paid to the location of camps

within the drainage network as impacts may be off site, particularly where drainage networks may

converge.

Figure 17 shows an example vulnerability map from the north east of Liberia, containing the five

contingency planning sites at B'hai Niko, Bin Sawmill, Janzon Town, Zleh Town and Pohan.

25


26

Figure 16: Vulnerability Map of the environs of Monrovia


In contrast to the previous example these sites have

been identified as possible sites in contingency

planning and as yet are not yet in use. The eastern

part of the country is more forested and as a

consequence has higher ecosystem service values

that the previous example. Furthermore, the sites

are in the vicinity of the proposed Zwedru protected

area. The Zleh, Pohan and B'hai Niko sites are all

located in small agricultural areas, whereas Bin

Sawmill and Janzon Town within either open or

closed dense forest.

Compared with the previous map there are few low vulnerability areas in green and more areas of

cyan, magenta and red, indicating higher environmental vulnerability. This is a direct consequence

of the higher ecosystem service values of this generally forested region. Along roads and where

the land cover is more open the vulnerability zones spread out over a wider area, in particular

around Pohan and Zleh.

The Bin Sawmill site is located in close proximity to the proposed Zwedru protected area. The

medium to high vulnerability immediately surrounding the site becomes high vulnerability within the

boundary of the proposed protected area. Note that the road network enables accessibility into the

proposed protected area from both the north-west and from the south. This combination of close

proximity of contingency sites to the proposed protected area and the existence of roads means

that the protected area is vulnerable to encroachments and related environmental impacts.

The Janzon Town contingency site shows high vulnerability in its immediate vicinity because of its

location within closed dense forest. Other sites exist nearby, for example the area of rural

agriculture to the north east of the Janzon Town vulnerability zone, that may be more suitable from

an environmental perspective for the location of a contingency site.

Of these sites, Pohan and Zleh Town have generally lower environmental vulnerability and may be

prioritised ahead of the other more vulnerable sites.

27


Figure 17: Vulnerability map of contingency sites in eastern Liberia

28


4 Running the ModelTo calculate the Cost Distance it is necessary to have a starting point and and end point. The

starting point is the camp(s) locations(s) and the end point is every other geographic location. With

regard to camp location there are three possible scenarios for which the model can be run, namely

using:

• The location of existing camps

• The location of zones for camp locations that satisfy the UNHCR FRAME criteria

• The location of camps identified as part of contingency planning for future population

displacements

Clearly the vulnerability model is dependent upon the location of refugee / IDP camps because by

default an absence of camps means an absence of vulnerability5. This means that there is no

unique vulnerability map that can be made. All vulnerability maps must be in response to an

identified threat – or distribution of camps.

4.1 Step 1 – Data Preparation - Camp LocationsFor the analysis camp locations may be either:

• Existing camps

• Potential locations for new camps using a camp location methodology such as UNHCR-

FRAME

• Potential zone locations for new camps based on contingency planning

Any of these, or combinations of these, can be used as a starting point for the analysis.

In this example the location of existing camps, crossing points and way points are used to seed the

vulnerability model.

5 In the absence of a camp a geographic area may, of course, be vulnerable to some other type of threat.

29


Figure 18, above, shows the location of existing IDP and Refugee camps in Liberia. These camps

can be used as an input to the vulnerability model.

Figure 19 outlines the UNHCR-FRAME approach, the last frame showing the potential locations for

camps in green. Note that existing known sites are included in this analysis and are therefore

shown as being unsuitable for further camp development.

30

Figure 18: Location of Existing Camps and Contingency Sites


Figure 19: UNHCR FRAME Site Selection Approach

31


4.2 Step 2 - Definition of Land Characteristics from Existing SourcesLand cover and terrain data should be prepared as describe in the previous sections in order to

generate cost-surface and ecological service value maps.

4.3 Step 3 – Calculating Proximity and AccessibilityGiven a set of camp locations, or potential camp location zones, it is possible to calculate the Cost-

Distance from the camps to all other locations, using the camp locations and Cost-Surface

described in Section 3.4.1 and illustrated in Figure 7.

Figure 20 shows the Cost-Distance of moving across the landscape from existing camps.

The influence of a camp is obviously localised and calculation of the cost-distance from a camp to

every location in Liberia is of limited use because camp inhabitants will have a local rather than

national impact. The cost-distance should therefore be limited to the immediate vicinity of the

camp(s). The FRAME methodology uses a threshold of 15 km (15,000m) as the distance that

camp members are likely to walk in a day.

32

Figure 20: Cost-Distance base on existing camps, contingency sites and crossing points,

blue = higher cost, yellow = lower cost


Figure 21 illustrates the cost-distance surface within 15k cost-distance units of the existing camps.

The pattern is clearly highly influenced by the road network and reflects that the ease of traversing

the landscape by road.

The low cost-distance values at location A and B, in Figure 22, illustrates that the barrier of the

rivers means that movement from camps on the south side of the river is forced to go the long way

around over the bridge, making the blue area at A and B relatively inaccessible from the camps on

both the north and south sides of the river.

33

Figure 21: Cost-Distance from camps < 15,000 cost-distance units


34

Figure 22: Detail of cost-distance analysis around Monrovia.

A

B


4.4 Step 4 – Definition of Environmental Value – Dimension 2Figure 24 shows a detail of the environmental value for the same area shown in Figure 22, the

boundary of the 15k cost-distance threshold is also shown. It is quite clear that the existing camps

are generally located in areas of low ecological service value.

Despite the existing camps generally being located in an area of relatively low ecosystem service

value this does not imply that they are in an area that has no ecosystem service value. These

relatively low values must be seen in the context that Liberia contains some areas of very high

value, such as intact tropical forest habitats.

35

Figure 23: Ecosystem Service Value in relation to 15k cost-distance footprint


4.5 Step 5 Combining Hazard Impact Potential and Value in a Vulnerability ModelTo complete the Environmental Vulnerability Map the cost surface and ecosystem service value

maps must be combined to show those areas that are a) potentially exposed to the hazard and b) have a high ecosystem service value. This is done by subdividing the two dimensional data space

into a matrix and colour coding the vulnerability map according to which cell in the matrix a pixel's

values occupy in the data space.

Figure 24 shows the vulnerability of the environment to existing camps. Please refer to Figure 13

for a full explanation of the colour coding. Cyan tones indicate low ecological service values with

increasing proximity to a camp from light to dark tones. Similarly orange and red tones indicate

high environmental values with increasing proximity to camps from light to dark tones. It is clear

from Figure 24 that existing camps are located in a region of generally low ecological service

values, except near the coast. It should, however, be noted that low ecological service values does

not mean no ecological service value.

Figure 25 shows the environmental vulnerability map for the entire country based on existing

camps and also the contingency sites and entry points identified in relation to a potential influx of

refugees from Cote D'Ivoire. The spatial distribution of vulnerability intensity is apparent from the

36

Figure 24: Vulnerability of the environment to existing camps in the environs of Monrovia


distribution of colours.

37

Figure 25: Vulnerability of the environment to existing camps and contingency sites and entry

points


5 Drainage and Catchment AnalysisWater catchments provide a subdivision of the landscape into units that have a physical basis and

which are closely associated with hydrological processes and which are directly affected by land

management and land cover changes. In the landscape a hierarchy of drainage catchments exists

from major drainage basins to micro-catchments.

The SRTM terrain data provides a consistent information layer from which a number of drainage

parameters can be extracted. HydroTools for ArcView has been used to correct the SRTM data to

remove sinks that have no drainage outlet, which then enables the stream network and

catchments to be extracted. A number of derived hydrologic measures can be extracted from the

SRTM data including:

• flow direction – the direction of flow from a cell

• flow accumulation – the total accumulation of flow into a grid

• flow length – the distance from the farthest upstream point in a catchment

• catchments / watersheds – contributing area

• stream order - the numerical order of a stream branch in the stream network

• stream network – the network of stream channels

Using these derived measures the a set of catchments and the drainage relationships between

them can be calculated. Catchments can be identified at different levels in the national catchment

hierarchy. For the purposes of this study a threshold of 1,000 cells (approximately 1,000 ha) was

used to define the number of contributing cells to identify a stream, which in turn defines the size of

catchment extracted. Figures 26 and 27 show the stream network6 and the associated

catchments.

6 Figure 26 shows both the major and minor stream networks calculated using 10,000 and 1,000

contributing cells respectively.

38


39

Figure 26: Major and minor stream network

derived from SRTM data

Figure 27: Minor catchments extracted from SRTM data


The model outlined in previous sections uses per-pixel values of accessibility and ecosystem

service value to define a two dimensional data space which is subdivided into vulnerability classes.

The catchment maps however are a type of isopleth map and identifies contiguous areas with the

same arbitrary value (catchment id) and cannot easily be considered as an additional data axis in

the model. The following section considers how the catchment and drainage information derived

from the SRTM data can add value to the vulnerability model.

5.1 Ecological Service Value by CatchmentFigure 12 shows the ecological service value map, constructed through the combination of the

FRM land cover map and Costanza's ecological service value table. In Figure 12 the value of a

patch of landscape is governed by the ecosystem boundaries, many of which are artificially

imposed on the landcape by anthropogenic modification of the natural land cover, for example the

boundary between closed dense forest and open dense forest is a function of logging intensity.

Catchments however provide a natural subdivision of the landscape that are directly associated

with landscape morphology, hydrological processes and land cover.

40

Figure 28: Ecological service value by catchment (protected area weighting not included)


Within each catchment the sum of the ecological service value, according to Figure 12, has been

calculated and is shown in Figure 28. The general division of the country into north-western,

central and south-eastern zones is apparent, although this is also clear from the basic land cover

map. Where this map is useful is in identifying high value catchments that are intact and,

conversely, lower value catchments where camps may be more suitably located – in particular in

the north-west and the south-east of the country. The catchment value and the land cover maps

should be viewed together rather than individually to get the maximum benefit from them both.

The location of camps in lower value catchments may reduce the absolute impact on the

environment, however, higher value downstream catchments may be indirectly affected.

5.2 Drainage ConcentrationThe downstream impacts of pollution from camps can be estimated by looking at the drainage

catchments in which camps occur. For example, of the 37 known existing camps in the vicinity of

Monrovia, 15 of them, with existing populations totalling 100,895 people, are upstream of a single

outfall on the coast, illustrated by “B” in Figure 29.

The pollution of watercourses near a camp will only result in contamination of the watercourse

downstream and not upstream. Because the drainage from micro catchments contributes to larger

drainage basins there is the possibility that pollution from a number of camps will concentrate to a

single point, even if the camps are apparently well distributed in the landscape. Where stream

channels are well defined, contamination of the watercourse will be confined and will effect only the

channel (its water quality, flora and fauna) and not the surrounding area. However, where the

channel is less well defined, or where the stream flows into a wetland or lake, the contamination

41

Figure 29: Drainage of 15 camps to single outfall

at B.

B


will be dispersed more widely. A procedure is being developed to add the downstream impacts of

pollution to the vulnerability analysis. Note that in some parts of the country the pollution of

watercourses could have cross border implications.

42


6 Cartographic Vulnerability MapsThe plates on the following pages illustrate the vulnerability mapping applied to the ecological

service value and accessibility maps described in the preceding sections. The Plates show the

vulnerability analysis run on the known existing IDP and Refugee camps and potential camp and

transit camps identified in contingency planning in early 2006. The existing camps are mostly in the

environs of Monrovia, whilst the contingency planning camps are generally in the east and north of

the country. Please refer to the key to the Vulnerability legend shown in Figure 13 for an

explanation of the colours.

43

44

45

46

47

48

49

7 ReferencesChichilnisky, G., 1996, The economic value of the Earth's resources, Trends in Ecology and

Evolution, 11:3 135-140.

Costanza, R., d'Arge, R., de Gros, R., Faber, S., 1997, The value of the world's ecosystem

services and natural capital. Nature 387: 253-260.

Farrow, A., Nelson, A, 2001, Accessibility modelling in ArcView 3 – An extension for computing

travel time and market catchment information, International Centre for Tropical Agriculture.

Farber, S., Costanza, R., Wilson, A., 2005, Economic and ecological concepts for valuing

ecosystem services, In Environmental Values, Kalof, L., Satterfield, T. (Eds).

Foret Ressources Management, June 2004, Current state of the forest cover in Liberia – forest

information critical to decision making, Report to the World Bank

Jarvis, A., Rubiano, J. , Nelson, A., Farrow, A., Mulligan, M, 2004, Practical use of SRTM data in

the tropics – Comparisons with digital elevation models generated from cartographic data, Centro

Internacional de Agricultura Tropical (CIAT), Working Document No 198, 32p.

http://srtm.csi.cgiar.org/PDF/Jarvis4.pdf

Millenium Ecosystem Assessment, 2005, Ecosystems and human well-being: a framework for

assessment

Natural History Museum, 2005, Measuring Biodiversity Value, www.nhm.ac.uk

Pearce and Warford (1993), World Without End: Economics, Environment, and Sustainable

Development, Oxford University Press.

Primack, R., Essentials of Conservation Biology, 1998, Sinauer Associates

Pimentel, D, Wilson, C., McCallum, C., 1997, Economic and environmental benefits of diversity,

BioScience 47: 747-757.

National Research Council, 2005, Valuing ecosystem services – towards better environmental

decision-making, National Academies Press, Washington DC

Schäuble, H., 2003, HydroTools 1.0 for ArcView 3.x, Institute of Applied Geosciences, Technical

University of Darmstadt, www.terracs.de

Tobin, G.A., Montz, B.E., 1997, Natural Hazards, Explanation and Integration, The Guildford Press,

New York, London

UNHCR, CARE International, 2005, Geographic Information System (GIS) Applications for

Environmental Management in Refugee-related Operations, UNHCR.

http://www.nhm.ac.uk/

http://srtm.csi.cgiar.org/PDF/Jarvis4.pdf

Date post:	02-May-2017
Category:	Documents
Upload:	viet-huong
View:	217 times
Download:	0 times

Environmental Vulnerability Mapping in Liberia _Unep 2006

Documents