Politics of Climate-Inducted Displacement in Iraq

DRIVERS OF
CLIMATE-INDUCED
DISPLACEMENT
IN IRAQ
CLIMATE VULNERABILITY ASSESSMENT
OCTOBER 2023
The International Organization for Migration (IOM) is committed to the principle that humane and orderly
migration benefits migrants and society. As an intergovernmental organization, IOM acts with its partners in the
international community to assist in meeting the operational challenges of migration; advance understanding
of migration issues; encourage social and economic development through migration; and uphold the human
dignity and well-being of migrants.
The opinions expressed in the report are those of the authors and do not necessarily reflect the views of
the International Organization for Migration (IOM). The information contained in this report is for general
information purposes only. Names and boundaries do not imply official endorsement or acceptance by the
International Organization for Migration (IOM).
All maps in the report are for illustration purposes only. Names and boundaries on maps do not imply official
endorsement or acceptance by IOM.
This publication was issued without formal editing by IOM.
This publication has been issued without IOM Publication Unit (PUB) approval for adherence to IOM’s brand
and style standards.
International Organization for Migration

Address: UNAMI Compound (Diwan 2), International Zone, Baghdad/Iraq
Email: [email protected]
Website: https://iraqdtm.iom.int
The following citation is required when using any data and information included in this information product: ‘International
Organization for Migration (IOM), October 2023. Drivers of Climate-Induced Displacement in Iraq: Climate Vulnerability
Assessment. IOM, Iraq.’ For more information on terms and conditions of DTM information products, please refer to:
https://dtm.iom.int/terms-and-conditions.
© 2023 International Organization for Migration (IOM)
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted
in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without the prior
written permission of the publisher.
CONTENTS
ACRONYMS 2
EXECUTIVE SUMMARY 3
INTRODUCTION 7
METHODOLOGY 8
DEFINITIONS AND INDICATORS 8
CHALLENGES AND LIMITATIONS 9
FINDINGS 10
DISPLACEMENT DYNAMICS 10
DRIVERS OF CLIMATE-INDUCED DEPOPULATION 11
ENVIRONMENTAL HAZARDS AND WATER ACCESS 12
ACCESS TO SERVICES AND INFRASTRUCTURE 15
LIVELIHOODS AND MITIGATION MEASURES 16
TENSION AND CONFLICT 17
CHANGES IN VEGETATION HEALTH IN HOTSPOT LOCATIONS 18
CONCLUSION 20
RECOMMENDATIONS 20
AREA-BASED PROGRAMMING 20
AREAS FOR FURTHER DATA COLLECTION AND RESEARCH 20
ANNEX 21
REFERENCES 27
DRIVERS OF CLIMATE-INDUCED DISPLACEMENT IN IRAQ: CLIMATE VULNERABILITY ASSESSMENT
ACRONYMS
CHAID Chi-Square Automatic Interaction Detection
DTM Displacement Tracking Matrix
ET Emergency Tracking
HHs Households
IOM International Organization for Migration
NASA National Aeronautics and Space Administration
NDVI Normalized Difference Vegetation Index
NIR No Information Rate
RARTs Rapid Assessment and Response Teams
WFP World Food Programme
2 IOM IRAQ
EXECUTIVE SUMMARY
Climate change1 and environmental degradation2 have contributed to the displacement are those who rely on land and water resources for income, struggle to access basic
of at least 55,290 individuals in assessed locations in central and southern Iraq between services and experience tensions over natural resources.12 Ineffective water manage-
January 2016 and October 2022. 3, 4
This figure represents an estimated 15 per cent ment policies, broken or inefficient water infrastructure and damming or diversions
of the original population5 that used to reside in these locations. In other words, more by upstream governorates and countries also contribute to unequal access to avail-
than 1 in 10 people from these locations have been displaced in the past six years. able water resources.13, 14, 15 Together, these factors shape the ability of people to
Given the uptick in environmental migration observed in 2022, this trend is expected to withstand the environmental hazards confronting them in central and southern Iraq,
worsen, particularly in the absence of adequate mitigation and adaptation strategies.6, 7 such as drought, sand/dust storms, increased water salinity and reduced water levels.
To better understand the root causes of these movements, the International

IMPACTS OF CLIMATE CHANGE AND
Organization for Migration’s (IOM) Displacement Tracking Matrix (DTM) devel-
ENVIRONMENTAL DEGRADATION
oped a tool to assess a location’s vulnerability to climate-induced displacement.
Adverse environmental conditions can lead to lower crop and fishing yields
The tool measured four dimensions: 1) environmental hazards and water
and a reduced capacity to feed livestock. In response, some families have aban-
access, 2) services and infrastructure, 3) livelihoods and mitigation measures8
doned agricultural, livestock and fishing livelihoods altogether.16 Additionally, land
and 4) tension and conflict. This section presents the key findings across these
degradation and increasingly scarce water resources can contribute to tension
dimensions. This assessment examined conditions in locations that have already
in communities. These tensions, in turn, make cooperative approaches towards
recorded climate-induced displacement,9 as measured by IOM DTM’s Climate
water management more challenging.17,18,19 Moreover, climate change and envi-
Emergency Tracking tool.10 Data collection for this assessment took place
ronmental degradation make locations less habitable by impeding the ability of
between August and October 2022 across nine governorates, 29 districts and
households to meet their most essential needs. For example, water shortages
262 locations. IOM’s Rapid Assessment and Response Teams (RARTs) collected
impact individual health outcomes while also increasing community-level health
these data through interviews with key informants (KIs) at the location level.
risks such as infectious disease outbreaks. In the face of these stressors, some
These interviews were triangulated using the normalized difference vegeta-
families send members of the household to different locations in search of a job,
tion index (NDVI) anomaly data, which measure changes in the ‘greenness’ of
while others reduce their expenses or sell assets, land or livestock. If conditions
vegetation cover and are used as a proxy to compare the density and health of
persist, families may displace to other locations.
vegetation over time. The NDVI anomaly data were extracted from the United
States Geological Survey/National Aeronautics and Space Administration (NASA) Accordingly, in contexts where climate change and environmental degradation
Landsat remote sensing data11 and analysed in partnership with the World Food are worsening living conditions and making traditional livelihoods less sustainable,
Programme’s Vulnerability and Analysis Mapping team. displacement from these areas serves as a strategy to cope with deteriorating envi-
ronmental conditions. Given the resources required to move, however, people who
PRE-EXISTING VULNERABILITIES are displaced by environmental factors are not necessarily those most impacted
Environmental degradation and climate change impact different groups in different or those with the greatest needs or vulnerability. Families who remain behind
ways and to differing degrees. The families who likely face the greatest challenges may lack the means to relocate, forming a potential ‘trapped population.’20, 21, 22
1 The United Nations Framework Convention on Climate Change defines climate change as ‘a change of climate which is attributed directly or indirectly to human activity that
alters the composition of the global atmosphere and which is in addition to other natural climate variability that has been observed over comparable time periods.’ IOM, Migration,
Environment and Climate Change: Evidence for Policy Glossary (Geneva, 2014).
2 The United Nations Office for Disaster Risk Reduction (UNDRR) defines environmental degradation as ‘[t]he reduction of the capacity of the environment to meet social and
ecological objectives and needs. […] Degradation of the environment can alter the frequency and intensity of natural hazards and increase the vulnerability of communities. The
types of human-induced degradation are varied and include land misuse, soil erosion and loss, desertification, wildland fires, loss of biodiversity, deforestation, mangrove destruction,
land, water and air pollution, climate change, sea level rise and ozone depletion.’ IOM, Migration, Environment and Climate Change: Evidence for Policy Glossary (Geneva, 2014).
3 International Organization for Migration, Displacement Tracking Matrix – Climate-Induced Displacement – Central and Southern Iraq (Baghdad, September 2022).
4 As of June 2023, DTM has recorded the displacement of 83,520 individuals in central and southern Iraq. International Organization for Migration, Displacement Tracking Matrix –
Climate-Induced Displacement – Central and Southern Iraq (Baghdad, July 2023).
5 The original population refers to people who have displaced from assessed locations as well as the population who continue to reside there.
6 IOM, Migration, Environment and Climate Change in Iraq (Baghdad, 2022).
7 Lisa Binder, Barbora Šedová, Lukas Rüttinger, Julia Tomalka & Stephanie Gleixner, Climate Risk Profile: Iraq, Potsdam Institute for Climate Impact Research & adelphi (Potsdam,
2022).
8 Mitigation measure is used here to refer to: ‘an action taken by individuals within a household when shocks, such as natural disasters and conflict, push them beyond the difficulties
faced in ‘normal’ times.’ The term ‘mitigation measure’ is used instead of ‘coping strategy’ as the latter term is commonly used as a measure of food security. Oxfam, Measuring
Household Stress: Introducing the multi-sector Coping Strategy Index for Afghanistan (Baghdad, May 2018).
9 DTM used its Climate Emergency Tracking tool to determine which locations have already recorded climate-induced displacement.
10 All of IOM DTM’s Climate Emergency Tracking reports can be found here.
11 Landsat Missions, Landsat Normalized Difference Vegetation Index, United States Geological Survey (n.d., Reston).
12 Roger Guiu, When Canals Run Dry: Displacement Triggered by Water Stress in the South of Iraq, Internal Displacement Monitoring Centre, Social Inquiry & Norwegian Refugee Council
(Geneva, 2020).
13 Action Contre la Faim & REACH, Climate Motivated Displacement: Baseline (Erbil, 2022).
14 Roger Guiu, When Canals Run Dry: Displacement Triggered by Water Stress in the South of Iraq, Internal Displacement Monitoring Centre, Social Inquiry & Norwegian Refugee Council
(Geneva, 2020).
15 World Food Programme and Social Inquiry, Prospects for Resilience Amid Fragility: Conflict Analysis of Al-Qurna and Al-Dair Districts in Basra Governorate (Baghdad, 2022).
17 Ibid.
18 World Food Programme & Social Inquiry, Prospects for Resilience Amid Fragility: Conflict Analysis of Al-Qurna and Al-Dair Districts in Basra Governorate (Baghdad, 2022).
19 World Food Programme & Social Inquiry, Improving Prospects for Peace and Stability in Vulnerable Communities in Southern Iraq. Thi-Qar Governorate Conflict Analysis (Baghdad, 2022).
21 IOM, Intersessional workshop on climate change, environmental degradation and migration: Background paper, International Dialogue on Migration (Geneva, 2011).
22 IOM defines trapped populations as ‘[p]opulations who do not migrate, yet are situated in areas under threat, […] at risk of becoming ‘trapped’ or having to stay behind, where
they will be more vulnerable to environmental shocks and impoverishment.’ IOM notes that trapped population may be ‘poorer households who may not have the resources to
move and whose livelihoods are affected.’ IOM, Migration, Environment and Climate Change: Evidence for Policy Glossary (Geneva, 2014).
3
OVERALL VULNERABILITY DISPLACEMENT DYNAMICS

The analysis for this section comes from a Random Forest classification algorithm. Nearly 62,000 families live in locations where climate-induced displacement
For a description of the model and methodology, please see the Methodology is taking place. In around two thirds of locations, over one in five people had
and Annex of this report. already left due to environmental issues at the time of the assessment (Map 1).
• The strongest predictor of the rate of depopulation (that is, the portion • The districts recording the highest levels of climate-induced displacement
of the original population displaced due to environmental factors) is the are:
presence of multiple water-related issues, such as reduced rainfall, lower
• Qal’at Saleh (Missan Governorate), where half of the population in
water allocation and broken or inefficient water infrastructure.23 As certain
assessed locations has been displaced (1,728 families);
water-related problems are widespread, not all issues are associated with high
rates of depopulation. In particular, a reduction in irrigation sources, a decrease • Al-Rifa’i (Thi-Qar Governorate), where three in five households in
in rainfall and a decline in water quality are common to nearly all locations, assessed locations have been displaced (1,321 families);
regardless of depopulation rate. On the other hand, the cost of water trucking
• Nassriya (Thi-Qar Governorate), where nearly one in five households
and damming or river diversions are specific water issues associated with
in assessed locations have been displaced (1,257 families).
higher rates of depopulation.
• Ten locations have been fully abandoned. Most of these locations are in
• Families struggling to meet their basic food needs is the second strongest
Thi-Qar Governorate, including five locations in Nassriya and three in Suq
predictor of the depopulation rate. This suggests that those being displaced
Al-Shoyokh.
are in a particularly vulnerable economic position.
• As of December 2022, 4,360 more families (26,160 individuals) were displaced
• The third strongest predictor of the depopulation rate is difficult access to
by climate change and environmental degradation in 2022 compared to 2021,
services or infrastructure. This highlights the vulnerability of families residing
representing an eight-fold increase (+803%).
in remote rural locations, where access to education, health care and markets
may be more challenging. • This increase is partially attributable to growing trends of climate-induced
displacement in Diyala, Babylon, Thi-Qar and Wassit governorates. More than
• Reliance on land for livelihoods is the fourth strongest predictor of
three quarters of climate-induced displacement in these governorates took
displacement. Changes in the environment have a greater impact on farmers,
place in 2022. The primary drivers of displacement in these governorates
livestock herders, fishers and transhumant pastoralists,24 as they depend on
in 2022 appear to be low rainfall and low water levels in the rivers and
favourable environmental conditions to make a living.
tributaries. Additional aggravating factors include low levels of groundwater,
• The fifth strongest predictor of displacement is the adoption of mitigation water salinization, disputes over water allocation and restrictions by the
measures by families. Reliance on mitigation measures indicates that families Government on the use of water.
have been negatively impacted by changes in the environment and are taking
steps to overcome these challenges. The most common mitigation measure, Figure 2. Districts with highest portion of population displaced
that is, sending households members to another location to make money,
61%
points to the lack of alternative livelihood opportunities in the area. However,
51%
this measure also enables families to raise money to remain or relocate. Other
44%
42%
tactics, such as borrowing money, reducing meal size or taking children out
32%
of school, undermine the well-being and resilience of families. If the situation
persists, families may be forced to leave the area.
Figure 1. Top five predictors of climate-induced depopulation
Al-Rifa'i Qal'at Saleh Afaq Hamza Amara

1 Water issues 10.09
Thi-Qar Missan Qadissiya Qadissiya Missan
2 Families struggling to meet food needs 8.64
3 Service and infrastructure issues 8.28
4 Reliance on land for livelihoods 7.73
5 Adoption of mitigation measures 6.48
Note: Based on mean decrease in Gini coefficient from Random Forest analysis.
23 Water-related issues include: 1) a reduction in irrigation water supply, 2) damming or river diversions, 3) reduced rainfall patterns, 4) broken or inefficient water infrastructure,
5) reduced water allocation, 6) reduced water quality (e.g. salinity or pollution), 7) population growth/intensive agriculture and 8) rising costs of water trucking.
24 Transhumance is defined as the ‘seasonal movement of people with their livestock between pastures (typically between mountain and lowland pastures) often over long distances,
and sometimes between borders. The term is often used as a synonymous [sic] of pastoralism.’ IOM, International Migration Law No. 34, Glossary on Migration (Geneva, 2019).
4 IOM IRAQ
Map 1. Portion of original population displaced due to environmental factors

Tikrit
Ra'ua Kifri
Samarra Al-Daur
Portion of original population displaced due to
Haditha Salah Al-Din Khanaqin
environmental factors
Al-Khalis
Al-Thethar
Balad
Al-Muqdadiya 1 - 10%
Heet
Al-Fares Diyala 11 - 20%
Ana Tarmia
Ba'quba 21 - 30%
Baladrooz
Kadhimia 31 - 40%
Adhamia
w 2
Tha awra
1
ra
Th
Ramadi Abu Ghraib
Karkh
Al Resafa 41 - 50%
BaghdadMada'in >50%
Anbar Falluja
Mahmoudiya
Al-Azezia Badra
Al-Suwaira
Wassit
Kerbala Al-Musayab
Al-Mahawil
Ain Al-Tamur
Kerbala Al-Hindiya Babylon Al-Na'maniya
Kut
Ali Al-Gharbi
Hashimiya
Hilla
Al-Hai
Al-Rutba
Kufa
ISLAMIC REPUBLIC
Diwaniya Amara
OF IRAN
Afaq
Al-Manathera Qadissiya
Al-Shamiya Al-Rifa'i
Missan
Al-Kahla
Hamza
Al-Rumaitha
Al-Maimouna
Al-Mejar Qal'at Saleh
Al-Shatra
Al-Kabir
Najaf Al-Samawa
Thi-Qar
Al-Khidhir
Nassriya
Najaf Al-Midaina
Suq
Al-Shoyokh Al-Qurna
Al-Chibayish Shatt
Al-Arab
Muthanna Basrah
Al-Salman Basrah Abu

Al-Zubair Al-Khaseeb
SAUDI ARABIA Fao
KUWAIT Persian Gulf
ENVIRONMENTAL HAZARDS AND WATER ACCESS SERVICES AND INFRASTRUCTURE

• Many locations are confronting multiple extreme weather and slow-onset • Access to basic services26 is a challenge in most locations. The service most
events. Nearly three quarters of locations experienced 6–8 types of hazards. difficult to access was water. In just under two thirds of locations, less than 75
The most common hazards, reported in over 85 per cent of locations, per cent of households had enough water for drinking or domestic purposes.
are droughts and sand/dust storms. Additionally, most locations reported
• Across governorates, Thi-Qar reported the worst access to services, with
increased water salinity, soil degradation and changing rainfall patterns.
nearly half of locations lacking access to all or nearly all services assessed. At
• Almost all locations reported a decrease in irrigation water supply in the the district level, Al-Rifa’i in Thi-Qar, Baladrooz in Diyala, Amara in Missan
past 12 months. KIs primarily attributed this reduction to reduced rainfall and Al-Samawa and Al-Rumaitha in Muthanna reported the highest portions
patterns, as well as reduced water quality, making water sources unsuitable of locations lacking access to all or nearly all services.
for irrigation. Additionally, at least half of locations pointed to the impact
• Support to these communities remains limited, despite the widespread
of water allocation and damming/river diversions, connected to a lack of
challenges they face. Only 5 per cent of locations reported receiving aid or
international water sharing agreements and inequitable distribution within
assistance27 from the Government of Iraq, humanitarian organizations, local
the country.25 Furthermore, just over one third of locations cited ineffective
charities, relatives and friends or other sources.
water management, in the form of broken or inefficient water infrastructure,
as a cause of this dwindling supply.
25 Nussaibah Younis, Early warning: How Iraq can adapt to climate change, European Council on Foreign Relations (Berlin, 2022).
26 Services assessed include electricity, water for drinking and domestic use, reliance on water trucking, primary and secondary education, health care, markets and access to aid or
assistance.
27 If KIs responded affirmatively that households in the locations had received any aid or assistance, they were asked the source of this aid (e.g. government assistance, humanitarian
organizations, charity from the local community, relatives and friends or other). They were then asked what form of assistance had been received. Options included: housing
reconstruction, temporary shelter, agricultural input subsidies, subsidized water supply, cash assistance/social safety net, in-kind food aid, psychosocial care, water, sanitation and
hygiene assistance, education assistance, agricultural assistance or other.
5 IOM IRAQ
LIVELIHOODS climate-smart infrastructure and awareness-raising campaigns around

water conservation.29
• Agriculture and livestock rearing are among the top livelihood activities
in most locations assessed. Nearly all locations indicated loss of crop • Additionally, communities in central and southern Iraq require greater
production, livestock deaths or reduced fishing yields as a result of support to diversify livelihood options and expand the provision of basic
environmental factors. A similar portion struggled to feed livestock. As services, especially access to water for drinking and domestic use.
a result, large-scale abandonment of agricultural, livestock and fishing
• As tensions and conflict over natural resources tend to revolve around water,
activities is taking place. In 71 per cent of locations, over half of households
support should be provided to communities to develop monitoring systems
no longer practice these trades.
related to water scarcity and platforms for dialogue and mediation. Further
• Changes to the environment are also causing households to adopt mitigation research and analysis should be conducted to understand the relationship
measures, as cited by KIs in 75 per cent of locations. Among locations between environmental stress and tensions or conflict in Iraq, particularly in
where families are adopting mitigation measures, nearly all reported sending areas of displacement for climate migrants.
household members to another location to make money. Other common
• Many locations are experiencing wide-ranging challenges across the domains
mitigation measure include conserving water and reducing household
considered. However, based on a composite measure of all issues assessed,
expenditure.
10 districts have shown particularly high levels of vulnerability including:
• Qadissiya Governorate: Afaq and Hamza districts;

TENSION AND CONFLICT
• Missan Governorate: Al-Maimouna, Al-Mejar Al-Kabir, Amara, Qal’at
• Tension and conflict related to natural resources appear to be a localized
Saleh and Al-Kahla;
issue, reported by only seven districts including Nassriya, Kut and Suq
Al-Shoyokh districts. • Thi-Qar Governorate: Al-Rifa’i, Nassriya and Suq Al-Shoyokh districts.
• Disputes over natural resources led to the displacement of households in • Policies related to disaster risk reduction, climate change and development
three districts, namely Al-Qurna, Nassriya and Kut districts. should take into account the vulnerability and needs of those who have
displaced and may displace in the future.30, 31
• In locations reporting tension or conflict over natural resources, key
informants in more than four in five locations said such tensions have • Further data collection and research aimed at assessing and monitoring the
increased in the past year. impact of climate change and environmental degradation on living conditions
are needed. While this assessment collected data at the location level, more
• Tensions or conflict over natural resources primarily revolves around water,
granular assessments disaggregated by sex and age could shed light on
grazing land, livestock and arable land. Triggers for tension or conflict often
the different consequences faced by different identity groups. Moreover,
involved disagreements over water sharing, especially the perception that farmers
monitoring activities should integrate the perspectives of impacted families
were exceeding their water quota.28 Disputes tend to arise between members of
with remote sensing and satellite data on key environmental indicators.32
the same tribe or ethno-religious group sharing the same livelihoods.
Given the potential for greater climate-induced displacement in the future,
researchers and programmatic actors should further develop early warning
RECOMMENDATIONS or predictive models which could signal shifting trends in the short- to
• Water-related issues are the top driver of climate-induced displacement in long-term, building on existing examples.33, 34 Local authorities and the
Iraq. Addressing these challenges will require greater cooperation on the central government should be provided with technical assistance on data
distribution of water resources between Iraq and neighbouring countries, collection and analysis to build their capacity to assess, monitor and address
the development of more equitable water management policies within the consequences of climate change and environmental degradation based
Iraq, improvement of existing water infrastructure and acquisition of on their data.
28 This information was collected as part of a follow-up for qualitative information with the Rapid Assessment Response Teams.
30 Ibid.
31 IOM, People on the Move in a Changing Climate – Linking Policy, Evidence and Action (Geneva, 2022).
33 Ibid.
34 IOM, Somalia – Drought Related Displacement Analysis (December 2021) (Mogadishu, 2022).
6 IOM IRAQ
INTRODUCTION
Migration in response to environmental degradation and natural hazards35 is a including the extent of environment degradation, the impact of this degradation on
growing trend.36 As climate change raises the intensity and frequency of environ- livelihoods and social cohesion or the availability of services and infrastructure.
mental hazards such as droughts, floods and storms, more and more people are
The Climate Vulnerability Assessment aims to address this evidence gap, providing a
expected to leave their homes.37 Additionally, unpredictable temperatures and
more comprehensive and thematic assessment of locations of origin (that is, locations
rainfall patterns affect people practising land- and water-based livelihoods, with
that have already recorded displacement as a result of climate change and environ-
important consequences for food security and living conditions.38
mental degradation). Through this pilot, the assessment aims to build an evidence
Iraq is already witnessing the effects of climate change and environmental degra- base on the conditions in areas that have already recorded displacement, and to
dation.39 Temperatures are rising,40 rainfall is decreasing,41 droughts are more support identification of trends which indicate that locations and communities may
severe, water scarcity is increasing, sand and dust storms and flooding are
42 43
be vulnerable to displacement in the future.
more frequent.44, 45 Strikingly, temperatures in Iraq are soaring up to seven times
The first section of this report presents an overview of climate-induced displace-
faster than the global average, while annual rainfall is predicted to decrease by
ment in central and southern Iraq, analysing displacement flows, depopulation
9 per cent by 2050.46 Adding to these challenges, damming and other water
figures, time trends and fully abandoned locations. The next section discusses the
management policies in neighbouring countries have reduced Iraq’s water supply.47
top drivers of climate-induced depopulation,53 such as the presence of multiple
At the same time, Iraq’s demand for water continues to rise due to a growing
water issues, food insecurity and difficulties accessing services and infrastructure.
population, the expansion of cities and inefficient agricultural and industrial
The prevalence of different slow onset and extreme weather events, including
practices.48
drought, sand or dust storms and increased water salinity, is considered in the
As the processes of climate change and environmental degradation intensify, the following section, alongside factors contributing to a reduction in the irrigation
number of environmental migrants is likely to increase significantly. The drivers
49
water supply. The report then describes challenges accessing basic services and
of environmental migration – and their impact on vulnerable communities – are infrastructure, especially water for drinking and domestic use. The impact of
a critical area of research needed to inform targeted programming and advocacy. climate change and environmental degradation on land- and water-based live-
lihoods is assessed in the subsequent section, together with the adoption of
The International Organization for Migration’s (IOM) Displacement Tracking Matrix
mitigation measures to cope with these changes. Thereafter, the report explores
(DTM) in Iraq has conducted emergency tracking (ET) of families displaced by drought
the relationship between natural resources and tension and conflict, highlighting
conditions since June 2018.50 Enumerators continually collect data on locations of
the actors and resources involved, consequences and changes over time. Lastly,
displacement, the number of families displaced, the type of shelter in which they reside
remote sensing data on changes in vegetation health are evaluated in locations
and the year of their displacement.51 These figures are then published in quarterly
reporting high levels of climate-induced depopulation.
reports.52 However, not enough is known about what drove these families to displace,
35 The United Nations Office for Disaster Risk Reduction defines natural hazards as ‘natural processes or phenomena that may cause loss of life, injury or other health impacts,
property damage, loss of livelihoods and services, social and economic disruption, or environmental damage.’ United Nations Office for Outer Space Affairs, ‘Risks and disasters’
(n.p., n.d.).
36 IOM, IOM Perspectives on Migration, Environment and Climate Change (Geneva, n.d.).
37 Intergovernmental Panel on Climate Change, Global Warming of 1.5°C. An IPCC Special Report on the Impacts of Global Warming of 1.5°C Above Pre-Industrial Levels and Related Global
Greenhouse Gas Emission Pathways, in the Context of Strengthening the Global Response to the Threat of Climate Change, Sustainable Development, and Efforts to Eradicate Poverty,
Cambridge University Press (Cambridge and New York, 2018).
38 IOM, ‘Environmental migration’ (Geneva, n.d.).
40 Between 1970 and 2004, Iraq’s yearly average temperature has risen by 1-2 degrees Celsius. World Bank Group, Iraq: Systematic Country Diagnosis (n.p., 2017).
41 Rainfall is expected to drop 25 per cent by 2050. Theodore Karasik and Jacopo Spezia Depretto, ‘Climate change is exacerbating Iraq’s complicated water politics,’ Climate
Diplomacy (Berlin, 2019).
42 Based on the World Resources Institute’s Water Stress Index, Iraq is expected to reach 4.6 out of 5 points, with 5 reflecting the greatest degree of water scarcity. Andrew
Maddocks, Robert Samuel Young & Paul Reig, ‘Ranking the world’s most water-stressed countries in 2040,’ World Resources Institute (n.p., 2015).
43 Ibid.
44 United Nations Country Team in Iraq, Joint Analysis and Policy Unit, Sand and Dust Storms Fact Sheet (Baghdad, 2013).
45 REACH, United Nations Children’s Fund and United Nations Institute for Training and Research (UNITAR), Flood Hotspots in Iraq (October 2018–March 2019) (Baghdad, 2020).
46 Saleem A. Salman, Shamsuddin Shahid, Tarmizi Ismail, Eun-Sung Chung and Alaa M. Al-Abadi, Long-term trends in daily temperature extremes in Iraq, Atmospheric Research 198(6)
(August 2017).
47 Iraq’s drinking, sanitation and irrigation water comes almost entirely from the Tigris and Euphrates rivers. By 2030, flows in these rivers are anticipated to fall by 50 per cent,
largely as a result of water management practices in neighbouring countries such as Türkiye and the Islamic Republic of Iran. Pieter-Jan Dockx, Water scarcity in Iraq: From inter-
tribal conflict to international disputes, Institute of Peace and Conflict Studies (n.p., 2019).
48 Tobias Von Lossow, More than infrastructures: water challenges in Iraq, Planetary Security Initiative and Clingendael (n.p., 2018).
49 Ibid.
50 IOM, Iraq – Displacement Tracking Matrix (DTM) – Climate Methodology (Baghdad, n.d.).
51 Ibid.
52 Ibid.
53 Climate-induced depopulation refers to the share of the original population displaced by climate change or environment degradation. Climate-induced displacement, in contrast,
refers to the absolute numbers of people displacing.
7
METHODOLOGY
The assessment takes the form of key informant interviews conducted by IOM’s to leave their habitual place of residence, or choose to do so, either temporarily or
Rapid Assessment and Response Teams (RARTs), who are deployed across Iraq permanently, within a State or across an international border (emphasis added).’55
(20% of enumerators are female). IOM RARTs collect data utilizing a large and
This assessment focuses on families ‘displaced… due to drought, water scarcity
well-established network of KIs, including community leaders, mukhtars, local
or other climate-related factors since 2016.’56 In other words, it targets a subset
authorities and security forces.
of migrants who were forced to move due to climactic conditions. Accordingly,
Locations targeted for Climate Vulnerability Assessment have been identified the report will refer to this phenomenon as climate-induced displacement, in
by DTM Iraq’s RARTs from the ET caseload as the locations of origin of fami- keeping with the assessment’s focus, and the broader umbrella term of envi-
lies displaced due to environmental factors, including droughts, increased water ronmental migration.
salinity, biodiversity loss, soil degradation, changing rainfall patterns, increased
temperature, sand/dust storms and landslides. ENVIRONMENTAL DEPOPULATION RATE
Families displaced by environmental factors are grouped separately from those District where 0–10% of the original popula-
Low
displaced by the 2014–2017 conflict. In the rare instances where families were tion has left due to environmental issues
displaced by both the conflict and environmental conditions, these families are District where 11–40% of the original popula-
Medium
tion has left due to environmental issues
included in both the climate displacement caseload and the conflict displace-
ment caseload. Such is the case in Diyala Governorate, where some families District where 41–100% of the original popu-
High
lation has left due to environmental issues
were displaced by the conflict with the Islamic State in Iraq and the Levant then
returned to their area of origin, only to be re-displaced by drought.
ENVIRONMENTAL ISSUES
This pilot round collected data between August and October 2022 from 262
Issues include the presence of: 1) droughts, 2) increased water salinity, 3) biodi-
locations of origin containing 9,215 families in 29 districts across nine governor-
versity loss, 4) soil degradation, 5) changing rain fall patterns, 6) increased
ates of Iraq, who represent 88 per cent of the climate migrant caseload according
temperature, 7) sand or dust storms and 8) landslides.
to September 2022 ET data. This monitoring exercise focuses on governorates
in the central and southern regions of Iraq, along with Diyala, as climate-induced Low Presence of 0–2 environmental issues
displacement has primarily been observed in these governorates.
Medium Presence of 3–5 environmental issues
Location-level data have then been aggregated to identify trends or areas of concern High Presence of 6–8 environmental issues
at the district level. IOM key informants’ data have also been triangulated with geospa-
tial data, namely normalized difference vegetation index (NDVI) anomaly data, which
WATER ISSUES
measure changes in the health of vegetation compared to the long-term average.
These data were analysed in partnership with the World Food Programme (WFP). Issues include: 1) decrease in water supply, 2) damming or river diversion,
3) reduced rainfall patterns, 4) broken or inefficient water infrastructure,
5) reduced water allocation, 6) reduced water quality (e.g. salinity, pollution),
DEFINITIONS AND INDICATORS
7) population growth and/or agricultural intensification and 8) rising costs of
water trucking.
ENVIRONMENTAL AND CLIMATE MIGRATION
Low Presence of 0–2 water issues
There is currently no legal or internationally accepted definition for people
moving in response to environmental factors. Nonetheless, in 2007, IOM devel- Medium Presence of 3–5 water issues
oped a working definition of ‘environmental migrants,’ which highlights the High Presence of 6–8 water issues
diversity of environmental triggers for movement, the varying lengths of time
for such movement, the degree to which these movements are voluntary and LACK OF ACCESS TO SERVICES AND INFRASTRUCTURE
the potential destinations for these individuals.
Issues include: 1) less than 76 per cent of households have enough electricity
‘Environmental migrants are persons or groups of persons who, predominantly for for their needs; 2) less than 76 per cent of households have enough water for
reasons of sudden or progressive change in the environment that adversely affects their drinking and domestic needs; 3) reliance on water trucking (sometimes or
their lives or living conditions, are obliged to leave their habitual homes, or choose to always); the following services and infrastructure not being functional or acces-
do so, either temporarily or permanently, and who move either within their country sible within five kilometres: 4) primary school, 5) secondary school, 6) health
or abroad.’54 services or 7) markets; and 8) have not received aid/assistance.
Climate migration forms a subset of environmental migration, with individuals Presence of 0–2 issues related to infrastructure or services
Low
leaving their habitual place due to changes in the environment linked to climate
Medium Presence of 3–5 issues related to infrastructure or services
change. In particular, climate migration refers to:
High Presence of 6–8 issues related to infrastructure or services
‘The movement of a person or groups of persons who, predominantly for reasons of
sudden or progressive change in the environment due to climate change, are obliged
54 IOM, Discussion Note: Migration and the Environment (MC/INF/288 of 1 November 2007), prepared for the Ninety-Fourth Session of the IOM Council (Geneva, 2007).
55 IOM, International Migration Law No. 34, Glossary on Migration (Geneva, 2019).
56 IOM, Iraq – Displacement Tracking Matrix – Location-Based Climate Vulnerability Assessment Questionnaire [publication forthcoming].
8 IOM IRAQ
LACK OF ACCESS TO LIVELIHOODS CHALLENGES AND LIMITATIONS

Issues include: 1) reliance on land including commercial farming, smallholder
LOCATIONS OF ORIGIN FOR CLIMATE-
farming, livestock rearing or transhumance; 2) reliance on water including fishing
INDUCED DISPLACEMENT
or fish farming; 3) loss of crops; 4) abandoned agriculture/livestock/fishing activ-
ities in classes (more than 50% of households); 5) unable to meet food needs Many of the locations of origin targeted for this assessment have not previously
(more than 50% of households); 6) struggle to feed livestock; 7) no access to been assessed by DTM Iraq. As a result, the first round of data collection should
grazing land or access to land of poor quality or rising price for fodder and 8) be considered a pilot round, with RARTs identifying as many locations of origin
adopted mitigation measures, including migration. for affected families as possible and identifying suitable KIs in each location to
report on movements and living conditions. Given the operational complexity of
Low Presence of 0–2 issues related to livelihoods
identifying locations for this assessment and the remoteness of some locations
Medium Presence of 3–5 issues related to livelihoods
of origin, not all locations of origin were assessed in the pilot round.
High Presence of 6–8 issues related to livelihoods
Furthermore, due to ongoing environmental programming in Al-Shatra district,
TENSION OR CONFLICT locations in other districts were prioritized for assessment, in coordination with
the Ministry of Migration and Displacement. As a result, Al-Shatra was not
Issues include: 1) presence of tension/conflict between groups, 2) tensions/conflict
assessed, although it is as a district of origin and displacement for climate migrants
are related to natural resources (one point for one resource and two points
according to the Climate ET data.
for more than one), 3) the number of groups involved in tension/conflict (with
one point for one group involved and two points for two or more groups), and
LOCATION-LEVEL DATA
4) tensions/conflict are increasing.
All data collected in the Climate Vulnerability Assessment were collected at
Low Presence of 0–2 issues related to tension or conflict
location level. This means that KIs were interviewed about the conditions faced
Medium Presence of 3–4 issues related to tension or conflict in locations of origin for climate-displaced families. These locations correspond
High Presence of 5–6 issues related to tension or conflict to a village for rural areas or, less commonly, a neighbourhood for urban areas
(that is, the fourth administrative division). While this approach allows exten-
RANDOM FOREST METHODOLOGY sive coverage over a short period of time, it often relies on one informant per
location. These KIs are mainly mukhtars and community or local council repre-
To understand the leading drivers of climate-induced displacement, a Random
sentatives who report on the views and experiences of a potentially large and
Forest classification algorithm was used. Random Forest is based on the Decision
diverse population, which might lead to limited representation for smaller groups
Tree model, a type of machine learning algorithm that ranks different predic-
with distinct characteristics, or discrepancies caused by social desirability bias.
tors based on their level of association with the dependent variable and sorts
Additionally, key household characteristics including sociodemographic indicators
observations into groups based on shared traits. Decision Tree analysis can be
(i.e. number of family members) and vulnerability factors (i.e. gender of the head
used to classify locations based on the most relevant independent variables and
of household or number of members living with a disability) are not accounted
predict which category locations might fall into based on those characteristics.
for in the datasets.
The Random Forest builds a set of uncorrelated Decision Trees (a forest) and
ranks the predictors used in each tree based on the mean decrease in the Gini
A specific challenge of location-level data with regards to climate indicators is
coefficient. This approach addresses issues of multicollinearity and overfitting
the complexity of the variables (e.g. the concept of land degradation) and the
associated with a single Decision Tree.
subjectivity of perceptions with regard to changes in the climate (e.g. unseasonal
The dependent variable for the Decision Trees composing the Random Forest variations in the temperature). As noted earlier in the section, the assessment
was ‘rate of depopulation’ (that is, the portion of the original population addresses this point by triangulating key informants’ data with available geospatial
displaced), grouped as follows: ‘low’ depopulation (0–10%), ‘medium’ depop- data, namely NDVI Anomaly data,57 in collaboration with WFP’s Vulnerability
ulation (11–30%) and ‘high’ depopulation (31–100%). In total, 28 independent and Analysis Mapping unit.58
variables were tested through the Random Forest analysis.
GROUPS NOT CONSIDERED WITHIN ASSESSMENT
Although the Random Forest addresses issues of multicollinearity in terms of
prediction, more caution is needed with respect to the score of factors. If two The Climate Vulnerability Assessment targets locations from which families have
or more variables are correlated and one is selected for the model, the other already displaced, identifying the vulnerability factors present in each location.
variables will most likely not be selected. This is because the correlated variable(s) However, without including locations from which there has been no climate-in-
do not offer additional explanatory power for the model. As a result, the impor- duced displacement to date, it is difficult to determine how target locations differ
tance of the two or more correlated variables is weakened. The top predictors from others in the area or to identify what differentiates locations with displace-
of climate-induced depopulation should therefore be viewed holistically, without ment from those without it. On the other hand, this assessment did not assess
placing significant emphasis on the magnitude of difference between each score. locations that had been fully abandoned. Future versions of this assessment will
seek to include such locations, as they may constitute the areas most impacted
A complete description of the model development and methodology can be
by climate change and environmental degradation.
found in the Annex.
57 The NDVI measures the ‘greenness’ of ground cover and is used as a proxy to indicate the density and health of vegetation. The NDVI anomaly indicates the variation of the
current period to the long-term average, where a positive value would signify enhanced vegetation conditions compared to the average, while a negative value would indicate
comparatively poor vegetation conditions.
58 The VAM team conduct telephone interviews with around 3,000 households to assess food security and market access across Iraq.
9 IOM IRAQ
FINDINGS
DISPLACEMENT DYNAMICS and time trends. Most people are leaving rural areas in favour of urban centres,
often within the same governorate. However, some people relocate to other
This section discusses the scope and magnitude of climate-induced displacement
rural areas in order to continue practising land- and water-based livelihoods.
across central and southern Iraq, highlighting districts with high rates of depop-
Additionally, levels of displacement increased significantly in 2022 compared to
ulation (Map 2). A small set of locations, concentrated in Thi-Qar Governorate,
prior years, reflecting displacement from new areas as well as elevated displace-
have been fully abandoned, representing the worst-case scenario. Other trends
ment from previously identified locations.
related to climate-induced displacement are considered through analysis of flows
Map 2. Portion of original population displaced due to environmental factors

Tikrit
Ra'ua Kifri
Samarra Al-Daur
Portion of original population displaced due to
environmental factors
Al-Khalis
Al-Thethar
Balad
Al-Muqdadiya 1 - 10%
Heet
Ana Tarmia
Ba'quba 21 - 30%
Baladrooz
Kadhimia 31 - 40%
Adhamia
w 2
Tha awra
1
ra
Th
Ramadi Abu Ghraib

Karkh
Al Resafa 41 - 50%
BaghdadMada'in >50%
Anbar Falluja
Mahmoudiya
Al-Azezia Badra
Al-Suwaira
Wassit
Kerbala Al-Musayab
Al-Mahawil
Ain Al-Tamur
Kerbala Al-Hindiya Babylon Al-Na'maniya
Kut
Ali Al-Gharbi
Hashimiya
Hilla
Al-Hai
Al-Rutba
Kufa
ISLAMIC REPUBLIC
Diwaniya Amara
OF IRAN
Afaq
Al-Manathera Qadissiya
Missan
Al-Kahla
Hamza
Al-Rumaitha
Al-Maimouna
Al-Shatra
Al-Kabir
Najaf Al-Samawa
Thi-Qar
Al-Khidhir
Nassriya
Najaf Al-Midaina
Suq
Al-Shoyokh Al-Qurna
Al-Chibayish Shatt
Al-Arab
Muthanna Basrah
Al-Salman Basrah Abu

SAUDI ARABIA Fao
KUWAIT Persian Gulf
SCALE OF DISPLACEMENT
Overall, 9,215 families (55,290 individuals) have displaced due to drought, Figure 3. Districts with the highest number of families displaced due to environmental
water scarcity or other climate-related factors between January 2016 and conditions
September 2022, representing 15 per cent of the total population who used
1,728
to live in these locations. In other words, more than 1 in 10 people originally
residing in these locations have displaced. The top governorates of origin for 1,321 1,257
climate migrants are Thi-Qar (2,812 families) and Missan (2,719), followed by
743
Muthanna (975) and Qadissiya (950). At the district level, the largest number of 545
climate migrants left from Qal’at Saleh in Missan Governorate (1,728 families),
along with Al-Rifa’i (1,321 families) and Nassriya (1,257 families) in Thi-Qar
Governorate (Figure 3). Qal'at Saleh Al-Rifa'i Nassriya Afaq Amara
Missan Thi-Qar Thi-Qar Qadissiya Missan
Number of families displaced
10 IOM IRAQ
In around two thirds of locations, over one in five people had already left due to water salinization, disputes over water allocation and restrictions on the use of
environmental issues at the time of the assessment. The governorates with the water. In Basra, Kerbala, Missan and Muthanna, by contrast, most displacement
highest rates of depopulation are Qadissiya (44%), followed by Kerbala (31%), took place before 2019.
Thi-Qar (27%) and Missan (25%). Among districts, depopulation is highest in
Al-Rifa’i, Thi-Qar Governorate (61%); Qal’at Saleh, Missan Governorate (51%); Figure 5. Number of climate migrants by year of displacement, as of September 2022
Afaq (44%) and Hamza (42%), both in Qadissiya Governorate, and Amara,
Missan Governorate (32%) (Figure 4). Ten locations, most of which in Thi-Qar 3,011
Governorate, are completely abandoned due to climate-related issues (Table 1).
1,850
Figure 4. Districts with highest portion of population displaced
1,475
1,251
1,048
61%
51% 251 329

44%
42%
2016 2017 2018 2019 2020 2021 2022
32%
Environmental issues may also hinder returns of people displaced due to the
conflict with the Islamic State of Iraq and the Levant. Although an uncommon
Al-Rifa'i Qal'at Saleh Afaq Hamza Amara occurrence, 2 per cent of internally displaced households, most of whom orig-
Thi-Qar Missan Qadissiya Qadissiya Missan
inate from Al-Khalis district in Diyala Governorate, re-displaced upon return
to their district origin due to drought, water scarcity or other climate factors.
Table 1. Number of locations completely abandoned by district
District, Governorate Number of locations DRIVERS OF CLIMATE-INDUCED DEPOPULATION

Nassriya, Thi-Qar 5 This section explores the top five drivers of climate-induced depopulation (that is,
Suq Al-Shoyokh, Thi-Qar 3 the share of the original population who are displaced) through Random Forest
analysis60 (Figure 6). These drivers, which include the presence of multiple water
Al-Mejar Al-Kabir, Missan 1
issues, food insecurity, difficult access to services and infrastructure, reliance on
Badra, Wassit 1 land for livelihoods and adoption of mitigation measures, reflect the relation-
ship between environmental and economic challenges in central and southern
Total 10
Iraq. Reduced water quantity and quality, triggered by a wide range of factors
at the international, national, governorate and community levels, are making
TRENDS IN CLIMATE-INDUCED DISPLACEMENT climate-sensitive livelihoods increasingly difficult to sustain and preventing people
Climate migrants in Iraq often move from rural to urban locations. Nearly all from meeting their basic needs. Families engaged in farming, livestock rearing and
(98%) originate from rural areas, while more than half (57%) resettled in urban fishing depend on water resources and conducive environmental conditions to
settings, which may offer more livelihood opportunities. Moreover, individuals make a living. A reduction in income from these livelihoods means families are
displaced by environmental factors tend to stay close to their location of origin. struggling to afford essential items like food and are adopting mitigation meas-
According to September 2022 ET data, around two thirds (67%) of people ures to cope with these changes. At the community level, locations of origin are
displaced by environmental factors remain within their governorate of origin and failing to provide residents with the resources needed to adapt to these changes
a further third (37%) stay within their district of origin. Many families move to through the provision of services and infrastructure. This analysis suggests that
nearby urban centres, including the southern governorates’ capital cities, such as at the moment of displacement, families are in precarious economic positions
Amara (Missan), Kut (Wassit) and Nassriya (Thi-Qar), as well as other important and see few alternatives other than relocation.
urban areas. The only significant exception are rural areas in Najaf, which have The strongest predictor of climate-induced depopulation was the presence
attracted a significant number of families from the governorates of Qadissiya of multiple water issues in a location, such as reduced rainfall, lower water
and Thi-Qar. Moving to another rural area enables families to maintain their allocation and broken or inefficient water infrastructure. As some water issues
previous livelihoods. For example, some water buffalo herders from the marshes were common to nearly all locations, especially reduction in irrigation water
of Thi-Qar relocate to riverine areas of Najaf to continue these practices.59 sources, decrease in rainfall and reduced water quality, these factors were less
As shown in Figure 5, among the climate migrants identified by DTM since useful for predicting depopulation. In contrast, issues related to the cost of
2016, one third (33%) were displaced between January and September 2022, water trucking and damming or river diversions appear to have a greater
the highest of any year recorded. Considering these figures do not cover the influence on depopulation.
entirety of 2022, end-of-year figures are likely to be even higher. Displacement These drivers highlight the multifaceted nature of Iraq’s water crisis, which is
in 2022 was especially high in the governorates of Diyala (100%), Babylon (84%), impacted by climactic factors but also local, national and international water
Wassit (77%) and Thi-Qar (62%). Key drivers of movement in these gover- management policies. Additionally, issues related to the cost of water trucking
norates over the past year are low rainfall and low water levels in rivers and suggests more remote rural locations are affected, as long-distance transportation
tributaries. Additional aggravating factors include low levels of groundwater, can increase the price of these services. Moreover, without alternative sources
59 Based on notes from DTM RARTs.

60 For more information on this analysis, please see the Methodology and Annex.
11 IOM IRAQ
of water, families practising agriculture and livestock rearing will depend more The fifth strongest predictor of depopulation is the adoption of mitigation
heavily on water trucking. Greater reliance on water trucking translates into measures by families, reinforcing the links between environmental and economic
higher costs for families. Finally, families facing challenges related to the cost of challenges. The adoption of mitigation measures implies that families have been
water trucking are likely those who are struggling to afford it, thus underscoring negatively impacted by changes in the environment and are taking steps to
the relationship between environmental and economic challenges. overcome these difficulties. The most common mitigation measure is sending
household members to another location to make money. This reflects the lack
Food insecurity, or, more precisely, half or more households in a given location
of alternative livelihood opportunities in areas of origin and the reduced income
struggling to meet their basic food needs, is the second strongest predictor of
due to decreased yields and herds. Additionally, in many locations, households
climate-induced depopulation. Climate change and environmental degradation
are lowering their expenditure to cope. This strategy also implies that families
can contribute to food insecurity by decreasing crop yields and herds. This
are less able to meet their needs. Moreover, selling assets such as land and live-
reduced supply, along with greater reliance on imports, can raise the price of
stock is another commonly indicated mitigation measure. While this approach
food.61 The relationship may also be indirect, with climate change resulting in
generates income in the short term, it reduces the family’s wealth and poten-
reduced agricultural production and thus fewer livelihood opportunities in a tradi-
tial earnings from agriculture and livestock rearing in the medium to long term.
tionally labour-intensive industry. Unemployment and lower household income,
in turn, increases the risk of food insecurity.62 Other non-environmental factors Figure 6. Top five predictors of climate-induced depopulation
observed in recent months, such as depreciation of the Iraqi dinar and consumer
price inflation, have also driven up the price of food.63 The connection between
food insecurity and climate-induced depopulation emphasizes the economic 1 Water issues 10.09
issues driving this trend. If food insecurity is contributing to displacement, this
indicates that families in the locations of origin are not earning enough to meet
2 Families struggling to meet food needs 8.64
their most essential needs.
The third strongest predictor of depopulation is difficult access to services or 3 Service and infrastructure issues 8.28
infrastructure. This flags the challenges faced by remote rural locations, where
access to education, health care and markets is constrained. Given the health-
and energy-related impacts of climate change,64, 65 difficult access to services
4 Reliance on land for livelihoods 7.73
impedes the ability of families to cope with worsening environmental conditions.
Additionally, challenges accessing secondary education can limit job prospects, 5 Adoption of mitigation measures 6.48
thus reinforcing communities’ reliance on climate-sensitive livelihoods, such as
agriculture, livestock rearing, pastoralism and fishing.
Note: Based on mean decrease in Gini coefficient from Random Forest analysis
Reliance on land for livelihoods is the fourth strongest predictor of the depopu-
lation rate. For farmers, climate change and environmental degradation contribute
to reduced crop yields through increased temperatures, decreased water supply,
ENVIRONMENTAL HAZARDS AND WATER ACCESS
reduced water quality and pests, among other factors.66 For livestock rearers,
negative environmental changes are eroding the quality of grazing land and leading Central and southern Iraq are witness to a wide range of slow onset and extreme
to higher costs for fodder. This makes it more difficult to feed livestock, resulting weather events. Most locations assessed face at least six types of environmental
in more deaths. Reduced yields and herds translate into less income for farmers hazards (Map 3), with droughts, sand/dust storms and increased water salinity
and livestock rearers, causing growing numbers to abandon these practices alto- being the most common. As a result, the quantity and quality of water in Iraq
gether. In the absence of alternative livelihood options in their locations of origin, have diminished significantly. In many locations, irrigation water is sourced from
families must relocate in search of new job opportunities. In some cases, families rivers, canals and wells. However, this water supply has decreased nearly every-
sell their farms or livestock to fund the journey, making it more difficult to return where as a result of climactic factors and issues related to water governance
and resume these activities if environmental conditions improve.67 and management.
61 World Food Programme and Social Inquiry, Improving Prospects for Peace and Stability in Vulnerable Communities in Southern Iraq. Thi-Qar Governorate Conflict Analysis (Baghdad,
2022).
62 World Bank Group, Iraq Economic Monitor – Reemerging Pressures: Iraq’s Recovery at Risk (Spring/Summer 2023, Washington D.C.)
63 Ibid.
64 World Health Organization and United Nations Framework Convention on Climate Change, Iraq – Health and Climate Change Country Profile 2021 (Geneva, 2021).
65 Mounir Mahmalat and Ali Ahmad, ‘In Iraq, mitigating fragility means addressing climate change’ [blog post], World Bank Group, 18 April 2023.
66 IOM and Social Inquiry, A Climate of Fragility – Household Profiling in the South of Iraq: Basra, Thi-Qar and Missan (Baghdad, 2022).
67 IOM and Social Inquiry, Migration into a Fragile Setting: Responding to Climate-Induced Informalization and Inequality in Basra, Iraq (Baghdad, 2022).
12 IOM IRAQ
ENVIRONMENTAL HAZARDS
Map 3. Percentage of locations experiencing six or more environmental hazards in the last three years by district
Tikrit
Ra'ua Kifri
Samarra Al-Daur
Percentage of locations experiencing six
or more environmental issues
Al-Khalis
Al-Thethar
Balad 0%
Al-Muqdadiya
Heet
Ana Tarmia
Ba'quba 41 - 69%
Baladrooz
Kadhimia 70 - 86%
Adhamia
ra2
1
aw
ra
Th
aw
87 - 100%
Th
Ramadi Abu Ghraib Al Resafa Mada'in
Karkh
Baghdad
Anbar Falluja
Mahmoudiya
Badra
Al-Azezia
Al-Suwaira
ISLAMIC REPUBLIC
OF IRAN
Kerbala Al-Musayab
Al-Mahawil Wassit
Ain Al-Tamur
Kerbala Al-Hindiya
Babylon Kut
Al-Na'maniya Ali Al-Gharbi
Hashimiya
Hilla
Al-Hai
Al-Rutba
Kufa Diwaniya Amara
Qadissiya Afaq
Al-Manathera
Missan
Al-Kahla
Hamza
Al-Rumaitha
Al-Maimouna Qal'at Saleh
Al-Mejar
Al-Shatra
Al-Kabir
Najaf Al-Samawa
Thi-Qar
Al-Khidhir
Najaf Nassriya Al-Midaina

Suq
Al-Shoyokh Al-Qurna
Al-Chibayish Shatt
Al-Arab
Basrah
Muthanna
Basrah Abu
Al-Salman
SAUDI ARABIA Fao
KUWAIT Persian Gulf
The most common environmental hazards are droughts and sand/dust storms, WATER SOURCES FOR IRRIGATION
which were reported in nearly all affected locations (99%). Increased water
In many locations, families rely on rivers or streams (64%), irrigation canals
salinity and soil degradation were also commonly reported (90% and 87%,
(61%) or bore holes or wells (50%) to meet their water needs for irrigation.
respectively), especially in the six governorates of Diyala, Missan, Muthanna,
However, water trucking – from private suppliers (50%) or the government
Qadissiya, Thi-Qar and Wassit. Additionally, many locations experienced changing
(28%) – is still necessary in most locations of Thi-Qar, Qadissiya, Missan, Diyala
rainfall patterns (86%) and increased temperature (80%), particularly in Basra,
and Wassit governorates (Figure 8).
Diyala, Missan, Muthanna, Qadissiya and Thi-Qar. Biodiversity loss (67%) mostly
affects Kerbala, Missan, Muthanna, Qadissiya and Thi-Qar, whereas landslides
Figure 8. Most common water sources for irrigation
were reported only in Thi-Qar (16%) and Qadissiya (8%) (Figure 7).
Figure 7. Most prevalent environmental hazards by percentage of locations

River/stream 64%
Irrigation canal 61%
Water truck/tanker
Sand or dust Increased Soil 50%
Droughts from private suppliers
storms water salinity degradation
99% 99% 90% 87% Bore hole/well 50%
Water truck/tanker from

28%
government suppliers
Tap 21%
Changing rain Increased Biodiversity Landslides
fall patterns temperature loss
Other 1%
86% 80% 67% 4%
13 IOM IRAQ
CHANGES IN WATER SUPPLY
Map 4. Percentage of locations with six or more water issues by district

Tikrit
Ra'ua Kifri
Samarra Al-Daur
Percentage of locations experiencing six
or more water issues
Al-Khalis
Al-Thethar
Balad
Al-Muqdadiya 0%
Heet
Ana Tarmia
Ba'quba 21 - 50%
Baladrooz
Kadhimia 51 - 71%
Adhamia
Thaw awra2
ra1
Th
Ramadi Abu Ghraib Al Resafa
Mada'in 72 - 100%
Karkh
Baghdad
Anbar Falluja
Mahmoudiya
Badra
Al-Azezia
Al-Suwaira
ISLAMIC REPUBLIC
Kerbala OF IRAN
Al-Musayab
Al-Mahawil Wassit
Kerbala
Ain
Al-Tamur Babylon Kut
Al-Hindiya Al-Na'maniya Ali Al-Gharbi
Hashimiya
Hilla
Al-Hai
Al-Rutba
Kufa Diwaniya Amara
Qadissiya Afaq
Al-Manathera
Missan
Al-Shamiya Al-Rifa'i Al-Kahla
Hamza
Al-Rumaitha
Al-Maimouna
Al-Shatra
Al-Kabir
Najaf Al-Samawa
Al-Khidhir
Thi-Qar
Najaf Nassriya
Suq Al-Midaina
Al-Shoyokh Al-Qurna
Al-Chibayish Shatt
Al-Arab
Basrah
Muthanna
Basrah Abu
Al-Salman
SAUDI ARABIA Fao
KUWAIT Persian Gulf
Figure 9. Percentage of locations by water supply issues

A decrease in water supply was observed in nearly all locations assessed (99.6%).
Reduced rainfall patterns and degradation in water quality (increased water
salinity and pollution) were cited as the main drivers of this decrease (98% and
Reduced rainfall 98%
87%, respectively). Data also reveal the impact of water allocation practices
(60%) and damming/river diversions (50%),68 mainly stemming from a lack of Reduced water quality
(e.g. salinity, pollution) 87%
transboundary water sharing agreement between Iraq and its neighbours69, 70
Reduced water allocation 60%
and perceptions of inequitable water distribution within the country.71 A higher
portion of locations in Qadissiya (92%), Kerbala (92%), Thi-Qar (79%) and Cost of water trucks 50%
Missan (72%) attributed the decreased supply of water to reduced allocation.
Damming or river diversion 50%
Ineffective water management (that is, broken or inefficient water infrastructure,
Broken or inefficient
such as canals that suffer from evaporation or wasteful irrigation methods) was 36%
water infrastructure
reported in over one third of locations (36% overall), with peaks in Qadissiya Population growth/
10%
agricultural intensification
(96%) and Missan (81%). In half of locations (50%), KIs linked the decrease in
water supply to the costs of water trucking and in 10 per cent of locations to Other 1%
population growth or intensive agriculture techniques (largely in Diyala and

Kerbala governorates) (Figure 9).
68 As part of its Anatolia Project, Türkiye has constructed 22 dams on the Tigris and Euphrates rivers. The Islamic Republic of Iran also redirects tributaries from these rivers as part
of its water supply. Within Iraq, the Kurdistan Regional Government entered into an agreement with a private company to construct four dams. Nussaibah Younis, Early warning:
How Iraq can adapt to climate change, European Council on Foreign Relations (Berlin, 2022).
69 Bilateral water sharing agreements between Iraq and its neighbours exist but have been ignored or violated. Mervyn Piesse, Water governance in the Tigris-Euphrates Basin, Future
Directions International (Dalkeith, 2016).
70 Roger Guiu, When Canals Run Dry: Displacement Triggered by Water Stress in the South of Iraq, Internal Displacement Monitoring Centre, Social Inquiry and the Norwegian Refugee
Council (Baghdad, 2020).
71 Missan and Kut have experienced tensions over water sharing due to the perception by downstream governorates that upstream governorates were exceeding their allocation
of water. Additionally, disputes have arisen between Muthanna, Qadissiya and Thi-Qar governorates for similar reasons. Nussaibah Younis, Early warning: How Iraq can adapt to
climate change, European Council on Foreign Relations (Berlin, 2022).
14 IOM IRAQ
Reduced water levels have important implications for water quality. In Basra, ACCESS TO SERVICES AND INFRASTRUCTURE
for example, decreased water flows in the Shatt al-Arab waterway, along with
This section evaluates access to eight services, including electricity, water for
the intrusion of sea water from the Persian Gulf, have contributed to a three-
drinking and domestic use, primary and secondary school, markets and health
fold increase in the waterway’s salinity in the last 50 years. Increased salinity
services (Map 5). It also assesses the extent to which families rely on water
and pollution of water, in turn, led to the hospitalization of 118,000 people
trucking and whether families in the location have received any assistance in the
between August and November 2018. 72
past year. As described further below, difficulties accessing services and infra-
structure are widespread but are especially pronounced in Thi-Qar. In line with
the findings of the previous section, water is the most difficult service to access.
Map 5. Percentage of locations with six or more service issues by district

Tikrit
Ra'ua Kifri
Samarra
Al-Daur
Salah Al-Din Khanaqin Percentage of locations experiencing six
Haditha
Al-Khalis or more service issues
Al-Thethar
Balad
Al-Muqdadiya 0%
Heet
Ana Tarmia
Ba'quba 15 - 40%
Baladrooz
Kadhimia 41 - 60%
Adhamia
ra2
1
aw
ra
Th
aw
Mada'in
Th
Ramadi Abu Ghraib

Karkh
Al Resafa 61 - 86%
Baghdad
Anbar Falluja
Mahmoudiya
Badra
Al-Azezia
Al-Suwaira ISLAMIC REPUBLIC

OF IRAN
Kerbala
Al-Musayab
Al-Mahawil Wassit
Ain
Al-Tamur Kerbala Babylon Kut
Al-Hindiya Al-Na'maniya Ali Al-Gharbi
Hashimiya
Hilla
Al-Hai
Al-Rutba
Kufa Diwaniya Amara
Qadissiya Afaq
Al-Manathera
Missan
Al-Shamiya Al-Rifa'i Al-Kahla
Hamza
Al-Rumaitha
Al-Maimouna Qal'at Saleh
Al-Mejar
Al-Shatra
Al-Kabir
Najaf Al-Samawa
Al-Khidhir
Thi-Qar
Najaf Nassriya Al-Midaina
Suq
Al-Shoyokh Al-Qurna
Al-Chibayish Shatt
Al-Arab
Basrah
Muthanna
Basrah Abu
Al-Salman
SAUDI ARABIA Fao
KUWAIT Persian Gulf
Access to services and infrastructure appears to be challenging in nearly all (51%) or sometimes (43%). At the governorate level, 100 per cent of locations
assessed locations. KIs reported that no locations have adequate access to all in Missan and 97 per cent of locations in Thi-Qar always rely on water trucking.
eight screened indicators. Additionally, nearly 9 out of 10 locations faced at least Additionally, in 63 per cent of locations, less than 76 per cent of households
three service-related issues (63% experiencing 3–5 service issues and 25% expe- had enough water for drinking and domestic needs. With respect to aid, 95 per
riencing 6–8 issues). In Thi-Qar, 45 per cent of locations have inadequate access
73
cent of locations have not received any assistance in the past 12 months from
to all or nearly all screened indicators. At the district level, Al-Rifa’i in Thi-Qar, the government, humanitarian organizations, local charities or relatives or friends.
Baladrooz in Diyala, Amara in Missan, Nassriya in Thi-Qar and Al-Samawa in
Beyond water and humanitarian aid, around half of locations do not have access
Muthanna reported the highest portions of locations lacking access to all or
to functional secondary schools, markets and health centres within five kilome-
nearly all services.
tres (46%, 44% and 49%, respectively). More positively, electricity sufficiency and
The most difficult domains appear to be water and aid/assistance. In 94 per access to primary school are generally within most households’ reach (76% and
cent of locations, households have to rely on water trucking, either always 87%, respectively) (Figure 10).
72 Human Rights Watch, Basra is Thirsty: Iraq’s Failure to Manage the Water Crisis (n.p., 2019).
73 Data were synthetized in a composite index to better understand access to infrastructure and services. For a full description of indicators selected and related inadequacy
thresholds, please refer to the Methodology.
15 IOM IRAQ
Figure 10. Top service-related issues by percentage of locations Access to livelihoods appears to be relatively challenging in nearly all assessed
locations. Across the eight indicators assessed, more than three quarters (77%)
Have not received 95% of locations faced 6–8 livelihood-related issues. Only 1 per cent of locations
humanitarian aid/assistance
experienced two or fewer issues.75 In Wassit, Qadissiya, Missan and Basra, most
Sometimes or always
reliant on water trucking 94%
locations have inadequate access to all or nearly all screened indicators (100%,
Less than 76% have enough water 63% 96%, 94% and 91%, respectively).
for drinking and domestic needs
Always reliant on water trucking 51% Among these issues, most locations reported that households experienced loss
No health clinic present of crop production, livestock deaths or reduced fishing yields due to environ-
and functional within 5 km 49%
mental factors (99%), struggled to feed livestock (98%) and/or resorted to
No secondary school present
and functional within 5 km 46% mitigation measures (75%). As a result of these challenges, in 71 per cent of
No market present locations, over half of households abandoned agriculture, livestock or fishing
and functional within 5 km 44%
activities. Notably, more than 75 per cent of families stopped practising these
Less than 76% have enough
electricity for their needs 24% trades in all locations in Fao district, Basra Governorate and Hamza district,
No primary school present Qadissiya Governorate (Figure 12). Additionally, in 51 per cent of locations,
13%
and functional within 5 km
more than half of households struggled to meet food needs, especially in
Al-Na’maniya and Badra district in Wassit Governorate and Qal’at Saleh in
Missan Governorate.
LIVELIHOODS AND MITIGATION MEASURES
This section describes the economic impacts of climate change and environmental Figure 12. Portion of locations abandoning agriculture, livestock or fishing activities
degradation in assessed locations. In many locations, families rely on land- and
water-based livelihoods but have experienced decreased yields and herds in the 48%
last year. As a result, notable portions cannot afford enough food and must
adopt mitigation measures to cope with environmental changes and reduced
income. The most common mitigation measure is sending a household member
to another location for work, which reflects the role of migration as an adapta- 26%
23%
tion strategy to climate change. Other frequently reported mitigation measures,
such as reducing household expenditure, emphasize families’ precarious financial
position. In some cases, people are abandoning agricultural, livestock rearing and
fishing livelihoods altogether. However, in the absence of alternative job oppor- 2%
0%
tunities, these families will have no choice but to relocate in search of work.
None Some families Around half More than half Most
(1% - 24%) (25% - 49%) (50% - 75%) (76% - 100%)
Across assessed locations, the top livelihood activities include smallholder farming
(83% of locations), livestock rearing (79%) and informal or sporadic daily labour
(68%). Additionally, construction (38%), fishing (32%) and transhumance (21%) are
MITIGATION MEASURES
also commonly practised (Figure 11). Thus, many families are engaged in livelihoods
that depend on land and water resources, making them vulnerable to changes in the In locations where households adopted mitigation measures, sending one or
environment. To a limited extent, the public sector has absorbed some individuals more household members to another location for work was the most common
who are unable to maintain agricultural, livestock rearing or fishing activities. In this 74 approach, as reported in 95 per cent of locations. Families also decreased
assessment, only 9 per cent of locations report public service as a primary livelihood. consumption by conserving water (as recorded in 63% of locations) and reducing
household expenditures (55%). Particularly alarming mechanisms include pulling
Figure 11. Top livelihoods in districts of origin
children from school (11%) or sending them to work (35%), as lower educational
attainment can have long-term consequences such as diminished employment
Smallholder farming 83% prospects (Figure 13). Sending children to work was most frequently recorded
Livestock rearing in the governorates of Wassit (90%), Basra (67%) and Qaddisiya (52%). Basra
79%
also had the highest portion of locations in which families pulled children from
Informal or sporadic daily labour 68% school (67%).
Construction and related jobs 38%
Fishing 32%
Transhumance 21%
Other 19%
Public service 9%
74 Roger Guiu, When Canals Run Dry: Displacement Triggered by Water Stress in the South of Iraq, Internal Displacement Monitoring Centre, Social Inquiry and the Norwegian Refugee
Council (Baghdad, 2020).
75 Data were synthetized in a composite index to better understand access to livelihoods. For a full description of indicators selected and related inadequacy thresholds, please
refer to the Methodology.
16
Figure 13. Mitigation measures by portion of locations In just under a quarter of locations, tensions (18%) or open conflict (4%) between
the different population groups were reported, mainly between groups of the
same tribe or ethnic descent sharing the same livelihoods.83 Natural resources
Sent HH members to another 95%
location to make money were a driver of tension or conflict in only seven districts, especially Nassriya,
Kut and Suq Al-Shoyokh. In these areas, tensions primarily revolved around
Conserved water 63%
water (100%), grazing land (86%), livestock (50%) and arable land (41%) (Figure
14). A common source of tension was disagreements over water sharing, espe-
Sold assets, including land or livestock 55%
cially towards farmers who were perceived to have exceeded their water quota
for agriculture.84 Over the last 12 months, competition for natural resources
Reduced household expenditure 55%
has increased in 82 per cent of these locations. The most frequently reported
Diversified livelihood activities 39% consequences of these tensions were verbal confrontation (88%), although
procurement of weapons (20%), displacement (14%) and open conflict (13%)
Sent children to work 35% were indicated in a minority of locations.
Figure 14. Natural resources contributing to tensions among locations affected by

this issue
TENSION AND CONFLICT

Climate change and environmental degradation have the potential to under- Water 100%
mine both vertical social cohesion76 (that is, citizens’ trust in the government
and authorities) and horizontal social cohesion77 (that is, trust between citizens
Grazing land 86%
from different groups or backgrounds).78 With respect to vertical social cohesion,
poor water quality has triggered protests against authorities in Basra and Thi-Qar
Livestock 50%
governorates and constituted an important grievance among citizens during
protests in 2019.79, 80, 81 In terms of horizontal social cohesion, disputes over
Arable land 41%
access to arable land and water resources have contributed to clashes between
farmers, fishers and livestock herders, as well as between tribes.82
Fisheries 9%
Based on this assessment’s findings, tension and conflict over natural resources
appear to be a localized issue, reported in only seven districts. Individuals prac-
tising the same livelihoods, often within the same tribe or ethnic group, are
experiencing tensions related to water, land and livestock. In most cases, these The most problematic district appears to be Al-Qurna in Basra, with KIs reporting
disputes result in verbal confrontation but in select locations, confrontations have 5–6 issues in around one third of locations.85 Results also indicate an environmen-
triggered displacement. Increased competition over natural resources in the past tal-stressed situation in the three districts of Al-Midaina, Basra Governorate and
year suggest the risk of such tensions growing in the near-term. Nassriya and Suq Al-Shoyokh in Thi-Qar Governorate.
76 The United Nations Development Programme (UNDP) defines vertical social cohesion as ‘the degree of trust in national, subnational, or local governments and institutional
processes, e.g. elections, access to justice and public services. It is reflected in the ability of governments to ensure effective service delivery, promote inclusive political processes
and public policies, and the trust, legitimacy and confidence that citizens ascribe to governments, institutions and political processes.’ UNDP, Social Cohesion Hub, UNDP Conceptual
Framework of Social Cohesion (n.p., n.d.)
77 UNDP defines horizontal social cohesion as ‘the sense of trust, relationships and interactions among citizens across different identities or other social constructs, and in the way
that social organizations, civil society and social institutions exhibit a sense of interdependency and ‘common destiny’. Ibid.
78 Laura Birkman, Dorith Kool and Eva Struyken, Water Challenges and Conflict Dynamics in Southern Iraq: An In-Depth Analysis of an Under-Researched Crisis, Water, Peace and Security
(n.p., 2022).
79 Human Rights Watch, Basra is Thirsty: Iraq’s Failure to Manage the Water Crisis (n.p., 2019).
80 IOM, Conflict and Environmental Degradation in Iraq’s Dhi Qar Governorate (Baghdad, 2023) [internal publication].
81 Laura Birkman, Dorith Kool and Eva Struyken, Water Challenges and Conflict Dynamics in Southern Iraq: An In-Depth Analysis of an Under-Researched Crisis, Water, Peace and Security
(n.p., 2022).
82 Ibid.
83 Data were synthetized in a composite index to better understand the incidence of tensions and conflict. For a full description of indicators selected and related thresholds, please
refer to the Methodology.
84 According to RARTs from the central and southern governorates.
85 In northern Basra, tensions over water have spurred decades-long tribal disputes, causing death and injuries to dozens of people. An estimated 10 per cent of current tribal
disputes stem from water scarcity. Nussaibah Younis, Early warning: How Iraq can adapt to climate change, European Council on Foreign Relations (Berlin, 2022).
17 IOM IRAQ
CHANGES IN VEGETATION HEALTH IN HOTSPOT LOCATIONS
Map 6. 2014 Normalized Difference Vegetation Index (NDVI) in districts Map Badra
8. 2014–2022 NDVI anomaly data for locations reporting
reporting
2014 NDVI high ratesreporting
in districts of depopulation
high rates of depopulation highest
2014 rates
- 2022 of depopulation
NDVI anomaly data for locations reporting highest rates of depopulation
Water
Kut No Vegetation Cover High Vegetation Cover Loss
Ali
Al-Gharbi
Wa ssit
Al-Hai Amara High Vegetation
Cover
High Vegetation Cover Gain

Kut
Missa n
Qadissiy a
Al-Rifai Ali
Al-Na'maniya
Wa ssit Al-Gharbi
Al-Kahla
Al-Maimouna Al-Mejar Qal'at Saleh

M
Al Teeb area
Al-Shatra Al-Kabir
Al-Hai
Thi- Qa r
Al-Khidhir B asra h Amara
Al-Midaina Afaq
Nassriya
Suq
Al-Shoyokh
Al-Qurna
Shatt
M issa n
Mutha nna Al-Chibayish
Al-Arab Al-Rifai
Qadissiya Al-Kahla
Map NDVI
2022 7. 2022 NDVIreporting
in districts in districts
highreporting high rates
rates of depopulation of depopulation
Kut
Al-Rumaitha
Water
No Vegetation Cover M
Al Naeumh
Ali
Al-Maimouna Al-Mejar
Al-Gharbi Al-Kabir Qal'at Saleh
Al-Na'maniya Wa ssit Al-Shatra
Al-Hai High Vegetation
Cover
M
Amara Al Sakhara
Afaq Thi- Qa r M village
Al Hbish
Qadissiy a
Missa n Al-Khidhir
Al-Rifai Nassriya
Al-Kahla M
Al Rafie
Al-Midaina
tha
Al-Maimouna Al-Mejar
Qal'at Saleh Al-Qurna
Al-Kabir Suq
Al-Shatra Shatt
Al-Shoyokh
Thi- Qa r
Al-Chibayish
B a sra h Al-Arab
Al-Khidhir
Nassriya Al-Midaina M ut ha nna Basrah

Suq
B asra h Shatt
Al-Qurna
Al-Shoyokh Al-Chibayish Al-Arab
Mutha nna
Abu
To better understand the relationship between physical changes in the environment The NDVI anomaly analysis focuses on five hotspot locations with the highest
and climate-induced displacement, this section compares DTM displacement figures rates of climate-induced depopulation.89 The hotspot locations all fall within
with data on vegetation health, as measured through the Normalized Difference Thi-Qar and Missan governorates, near the end of the Euphrates and Tigris
Vegetation Index (NDVI) (Map 6, Map 7, Map 8). rivers, respectively. In addition to receiving reduced flows due to damming and
river diversions in neighbouring countries, these downstream governorates are
NDVI is a measure of the ‘greenness’ of vegetation86 that can be used to under-
affected by the water management, consumption and treatment practices of
stand the density of vegetation cover as well as the health of vegetation.87 Changes
governorates located further upstream.
in vegetation health over time can be detected using NDVI anomaly data. This indi-
cator compares the average NDVI in a given month to the long-term average for Despite having high levels of climate-induced depopulation, these locations had
the same month. 88
varying levels of vegetation loss. This suggests climate-induced depopulation
has both environmental and economic dimensions. Data on changes in vegeta-
NDVI anomaly data were extracted from the United States Geological Survey/NASA
tion health must be contextualized using relevant socioeconomic information,
Landsat remote sensing data and analysed in partnership with WFP’s Vulnerability
such as reliance on climate-sensitive livelihoods, access to basic services, food
and Analysis Mapping team.
insecurity and adoption of mitigation measures. Integrating environmental and
The maps above present the NDVI from August 2014 to August 2022 and the socioeconomic data will improve understanding of the consequences of climate
NDVI anomaly in August 2022 compared to August 2014–August 2021. The year change and environmental degradation for communities and thus the precise
2014 is the oldest on record for Landsat-8. August was selected as the month for drivers of climate-induced displacement.
comparison to coincide with the data collection period (August–October 2022).
86 Landsat Missions, ‘Landsat Normalized Difference Vegetation Index,’ United States Geological Survey (Reston, n.d.).
87 Ibid.
88 Cindy Schmidt and Amber McCullum, Creating and Using Normalized Difference Vegetation Index (NDVI) from Satellite Imagery, NASA Applied Remote Sensing Training Program
(ARSET) (Washington, D.C., 2016).
89 Fully abandoned locations were not selected for this analysis, given the limited assessment of these areas.
18 IOM IRAQ
NASSRIYA DISTRICT Additionally, dense vegetation increased in Amara from 0 per cent (0.16 km2)
to 0.10 per cent (0.99 km2), in contrast to other districts assessed. This increase
Nassriya district in Thi-Qar Governorate shows moderate to significant vegetation
in vegetation is likely attributable to reforestation initiatives that sought to plant
loss throughout the district. Between 2014 and 2022, dense vegetation decreased
thousands of trees in Amara to reduce the severity of dust storms. Moreover,
by 69 per cent, while sparse vegetation fell by 49 per cent. On the other hand,
as the district hosts the capital of the governorate (Amara city), it may have
barren land and built-up areas now comprise 99 per cent of the district. This loss
benefitted from increased attention from authorities and programmatic actors.
is especially pronounced in the areas directly north of Suq Al-Shoyokh district and
in the easternmost portion of the district. Additionally, since 2014, Nassriya expe- The decline in sparse vegetation is also more modest than that in other districts
rienced the largest percentage decline in water sources of the districts examined in (-23%). Some vegetation loss is visible in the northern and eastern regions of
this section (-95%). As indicated in NDVI maps of 2014 and 2022, Lake Sulaibiyat, the district, although more significant loss can be detected in the western half
located in the southwest of the district, has shrunk significantly. of the district and directly north of Al-Kahla district.
Two locations within Nassriya district are labelled within the NDVI anomaly map: Al Teeb, located in the north of the district, has experienced some vegetation
Al Rafie and Al Hbish. Nearly 9 in 10 families (89%) have left Al Rafie, while 95 loss, although less than observed in other locations. This is a likely a reflection of
per cent of the original population have been displaced from Al Hbish. the limited vegetation cover observed in this area in 2014. Nearly 8 in 10 families
originally residing in the location have been displaced (78%).
Al Rafie shows moderate to significant vegetation loss. The impact in Al Hbish
appears more tempered, although significant vegetation loss can be observed to
TAKEAWAYS
the west of the district.
The NDVI anomaly data indicates moderate to high vegetation loss. Qal’at Saleh in
AL-RIFA’I DISTRICT particular has suffered significant vegetation loss in its marshes. Additionally, water
bodies in Qal’at Saleh and Nassriya have visibly shrunk between 2014 and 2022.
Al-Rifa’i district in Thi-Qar Governorate has experienced moderate vegetation
loss throughout the district. Already in 2014, Al-Rifa’i had low levels of dense However, the locations with the highest rates of depopulation were not neces-
vegetation (covering 0.04% of the district) and a high portion of barren land and sarily those that experienced the highest degree of vegetation loss. This suggests
built-up areas (91%). Since 2014, sparse vegetation declined by 49 per cent, while that changes in vegetation cover are contributing to displacement but are not
water sources diminished by 36 per cent. its sole determinant. As highlighted in the Random Forest analysis, dependence
on land-based livelihoods and difficulties accessing basic services influence the
The vegetation loss observed in Al Naeumh, located in the south west, is consis-
degree to which climate change and environmental degradation impact families
tent with the patterns across the district. Around 7 in 10 families (71%) have been
and thus their vulnerability to displacement.
displaced from Al Naeumh, compared to 61 per cent within Al-Rifa’i district overall.
Smallholder farming and livestock rearing are among the top livelihoods in all
QAL’AT SALEH five locations assessed. Additionally, fishing is a key livelihood in Al Hbish and Al
Sakhara Village, while in Al Naeumh, transhumance is one of the most common
Qal’at Saleh district in Missan Governorate has experienced dramatic vegetation
livelihoods. In line with the Random Forest analysis, families in four locations
loss since 2014. Sparse vegetation cover fell from 36 per cent (414 km2) to 2 per
struggle to meet their basic food needs and have adopted mitigation measures
cent of the district area (19 km2) between 2014 and 2022. Additionally, in 2022,
to overcome these challenges.
Qal’at Saleh had no dense vegetation remaining (0.00% of the district). Vegetation
loss is particularly significant in the eastern half of the district, corresponding to Regarding access to services, around half of families do not have enough water for
the Hawizeh Marshes. In parallel, the portion of barren land and built-up areas drinking or domestic purposes in Al Hbish, Al Naeumh and Al Rafie. Moreover,
has jumped from 57 per cent (653 km2) to 95 per cent of the district area several locations did not have a secondary school, health clinic or market within
(1,094 km ). Water sources have also shrunk to half their size between 2014
2 five kilometres. Despite these difficulties, families in these locations have not
and 2022. Jaraiyah Lake, visible on the NDVI maps in the southeastern corner received any aid or assistance.
of the district, has also shrunk dramatically over this period.
Future analysis comparing NDVI anomaly and displacement data should consider
Water scarcity in the district appears to be driven by damming and reduced river selecting districts with similar living conditions in terms of livelihoods and services.
flows. Located on the border with the Islamic Republic of Iran, Qal’at Saleh and Doing so will help isolate the effect of changes in vegetation health on displace-
its marshes have been impacted by the buildup of earthern dams in the neigh- ment patterns. Additionally, selecting the peak month for vegetation in the
bouring country. Additionally, reduced flows from the Tigris River have caused locations under assessment will help illustrate the full impact of climate change
water levels in tributaries to drop. In some cases, secondary rivers have dried up and environmental degradation on vegetation health.
entirely. This has increased the distance to water bodies, making it more chal-
This analysis highlights the importance of bringing together environmental and
lenging for families to meet their basic water needs.
key informant data to understand climate-induced displacement. On the one
Al Sakhara Village, located in the south of the district, has experienced moderate hand, NDVI data can be used to examine the physical consequences of climate
to significant vegetation loss, particularly in its eastern half. Around three quarters change and environmental degradation on land and water sources. On the
of the original population (73%) have left this location. other hand, key informant data shed light on the human consequences of these
environmental changes. As climate-induced displacement reflects an interaction
AMARA DISTRICT between environmental and societal factors, making sense of this trend requires
integrating these two sets of data.
Amara district in Missan Governorate shows lower levels of vegetation loss
compared to the other districts under consideration here. This is likely related
to the high levels of barren land already present in 2014 (98% of the district).
19 IOM IRAQ
CONCLUSION
Locations reporting climate-induced displacement are confronting multifaceted Based on Random Forest analysis, the presence of multiple water-related issues,
challenges related to environmental hazards, water resources, access to services, especially high costs for water trucking and damming or river diversions, is the
sustainable livelihoods and intercommunal tensions. Many locations are experi- strongest predictor of climate-induced depopulation. Another key predictor is
encing a range of negative environment hazards and conditions, from droughts food insecurity, which underscores the difficult living conditions and economic
and sand/dust storms to increased water salinity and soil degradation. Moreover, precarity that families are facing prior to displacement. The third strongest
the supply of irrigation water appears to be shrinking in most locations, driven predictor is difficulties accessing basic services, which further compromise the
by reductions in the quantity and quality of water, high costs of water trucking, ability of communities to cope with environmental stressors. Dependence on
damming and diversions and inadequate infrastructure. Widespread difficulties climate-sensitive livelihoods like agriculture, livestock rearing and fishing is also
accessing basic services further undermine the resilience of communities in the a leading predictor of climate-induced depopulation. Negative environmental
face of environmental changes. These factors make it increasingly difficult to prac- changes can result in a loss of crop production, fishing yields and livestock
tice key livelihoods in agriculture, livestock rearing and fishing, forcing significant deaths, thereby reducing income levels and the ability to maintain these liveli-
portions to abandon these livelihoods and find alternative strategies to cope hoods. Finally, the adoption of mitigation measures is strongly associated with
with these worsening conditions. Furthermore, some groups practising these climate-induced depopulation. Reliance on mitigation measures indicates fami-
livelihoods are experiencing tensions over natural resources, especially water lies have been negatively impacted and are taking steps to overcome these
and grazing land. In short, climate change and environmental degradation are challenges. However, these efforts may not suffice in the face of significant envi-
contributing to a complex, interconnected web of issues in the locations assessed, ronmental challenges and may even undermine the wellbeing of households in
making it difficult for families to remain where they are. the medium- to long-term.
RECOMMENDATIONS
Climate-induced displacement in central and southern Iraq is primarily driven by The Government of Iraq, in partnership with inter- and non-governmental organi-
the country’s multifaceted water crisis. Addressing these challenges will require zations, should prioritize these areas for programmatic interventions and further
engagement and interventions at several political levels: assessment.
• International level: Greater cooperation over the distribution of water

resources between Iraq and neighbouring countries is needed.90 AREAS FOR FURTHER DATA
COLLECTION AND RESEARCH
• National level: The Government of Iraq, with the support of non-gov-
ernmental and intergovernmental organizations, should strengthen water Further data collection and research are needed to understand the relation-
management policies and develop monitoring systems related to water ship between slow-onset hazards and displacement, to identify high-risk areas
scarcity. of origin and displacement and sources of future vulnerability and to support
programmatic interventions. While this assessment collected data at the loca-
• Governorate level: Authorities must work to improve existing water infra- tion level, more granular assessments disaggregated by sex and age could shed
structure and acquire climate-smart infrastructure. Additionally, platforms light on the different consequences faced by different groups. Additionally,
for dialogue and mediation should be established to reduce the potential for to complete the quantitative data provided by this assessment, qualitative
tension and conflict over natural resources. research should seek to understand the multitude of factors feeding into the
• Community level: Communities require greater support to diversify liveli- decisions of families and individuals to leave their areas of origin. Moreover,
hoods and raise awareness of the importance of water conservation. monitoring activities should integrate the perspectives of impacted families with
remote sensing and satellite data on key environmental indicators.91 Given the
potential for greater climate-induced displacement in the future, researchers
AREA-BASED PROGRAMMING
and programmatic actors should further develop early warning or predictive
Many locations are experiencing wide-ranging challenges across the domains models which could signal shifting trends in the short to long term, building on
considered. However, 10 districts have shown particularly high levels of vulner- existing examples.92 Finally, local authorities and the central government should
ability including: be provided with technical assistance on data collection and analysis to build
their capacity to assess, monitor and address the consequences of climate
• Qadissiya Governorate: Afaq and Hamza districts;
change and environmental degradation based on their data.
• Missan Governorate: Al-Maimouna, Al-Mejar Al-Kabir, Amara, Qal’at Saleh
and Al-Kahla;
• Thi-Qar Governorate: Al-Rifa’i, Nassriya and Suq Al-Shoyokh districts.

91 Ibid.
92 IOM, Somalia – Drought Related Displacement Analysis (December 2021) (Mogadishu, 2022).
20 IOM IRAQ
ANNEX
ENVIRONMENTAL DEPOPULATION AND partitioning (or branching) examines all possible values of all available predictive
PREDICTIVE ANALYSIS: RANDOM FOREST MODEL variables and selects the grouping of cases that allows the maximum homoge-
neity in each group with respect to the value of the dependent variable. The
The main scope of this analysis is to identify patterns of environmental migra-
branches and internal nodes at the top represent the independent variables with
tion based on the response to environmental changes and predict depopulation
the strongest connection with the dependent variable.97
rate through the selection of specific characteristics related to environmental
and water issues, livelihoods and access to infrastructure and services. The unit Different tree types are distinguished by the manner of node partitioning. In this
of the analysis is the location and the analysis is based on the use of Random analysis, the Chi-Square Automatic Interaction Detection (CHAID) algorithm
Forest, a classification algorithm consisting of many Decision Trees that operate was applied, where CHAID denotes an automatic and iterative procedure of
as an ensemble.93, 94, 95 Random Forest uses bagging and feature randomness tree development based on Pearson’s chi-square statistic and corresponding
when building each individual tree to create an uncorrelated forest of trees p-value. At each step, CHAID chooses the independent (predictor) variable
whose prediction by committee is more accurate than that of any individual that has the strongest interaction with the dependent variable. Categories of
tree. Compared to a single Decision Tree, Random Forest works better with each predictor are merged if they are not significantly different with respect to
multicollinearity in terms of prediction as it selects a random subset of features the dependent variable.98
to grow each tree so that it is less probable that collinear terms will both be
selected, alleviating the issue of multicollinearity.
THE RANDOM FOREST MODEL:
CHARACTERISTICS AND PERFORMANCE
TREE-BASED METHODS
Two different Random Forest models were developed and tested, the only differ-
Tree-based methods are a simple and yet powerful approach to handle large ence being the definition of the dependent variable, ‘environmental depopulation
and complex datasets with many features interacting in non-linear ways, identify rate’. In the first model, the variable ranges from ‘low depopulation’ (0–10%)
significant ‘hidden’ patterns and interpret them in a single and comprehensive to ‘medium depopulation’ (11–30%) and ‘high depopulation’ (31–100%). In the
model. They partition the space into smaller regions, where interactions are second model, it ranges from ‘low depopulation’ (0–10%) to ‘medium depopu-
more manageable, and then fit a simple model in each region. They can be used lation’ (11–40%) and ‘high depopulation’ (41–100%). Both models employ the
for both classification and prediction.96 same 28 independent variables, which are listed in Table 1. All independent vari-
ables were defined as ordinal variables or binary categorical variables.
The tree structure comprises a hierarchically organized set of groups called
‘nodes.’ At the top of the tree (the root node) sits the full dataset of cases that As discussed in greater detail later in this section, only Model 1 was included
is recursively split into a number of ‘child’ nodes – each containing a subgroup in the main report due to its higher accuracy on the test dataset and for the
of cases – interconnected by branches, until no more partitioning is possible sake of simplicity.
and only the terminal nodes (the ‘leaves’ of the tree) are left. The criterion for
Table 1. Variables used for the analysis
VARIABLE STRUCTURE
Name Symbol Values (modalities) Cases Percentage
Low = 0–2 issues 1 1%
Environmental issues X1 Medium = 3–5 issues 69 27%
High = 6–8 issues 182 72%
Water issues X2 Medium = 3–5 issues 167 66%
Low = 0–2 issues 31 12%
Access to services and infrastructure X3 Medium = 3–5 issues 158 63%
Livelihood issues X4 Medium = 3–5 issues 55 22%
93 Leo Breiman, Random forests, Machine Learning, 45:5–32 (October 2001).

94 Tin Kam Ho, Random decision forests, Proceedings of the Third International Conference on Document Analysis and Recognition, Institute of Electrical and Electronics Engineers,
Montreal, Quebec, Canada, 14–16 August 1995.
95 Tin Kam Ho, The random subspace method for constructing decision forests, Institute of Electrical and Electronics Engineers Transactions on Pattern Analysis and Machine Intelligence,
20(8): 832–844 (August 1998).
96 Trevor Hastie, Robert Tibshirani & Jerome Friedman, The Elements of Statistical Learning: Data Mining, Inference, and Prediction (Second Edition), Springer (Stanford California, 2001).
97 Leo Breiman, Classification and Regression Trees. First Edition. Routledge (New York, 2017).
98 International Business Machines Corporation (IBM), IBM SPSS Decision Trees 21 (2012).
21 IOM IRAQ
VARIABLE STRUCTURE
Low = 0–2 issues 228 90%
Tension issues X5 Medium = 3–4 issues 9 4%
< 60% of HHs left in 2021–2022 119 47%
Recent displacement X6
60% or > HHs left in 2021–2022 133 53%
Present 125 50%
Damming/river diversion X7
Not present 127 50%
Present 91 36%
Broken/inefficient water infrastructure X8
Not present 161 64%
Present 152 60%
Reduced water allocation X9
Not present 100 40%
Present 126 50%
Cost of water trucking X10
Not present 126 50%
Present 208 82%
Small-holder farming X11
Not present 44 18%
Present 198 79%
Livestock rearing X12
Not present 54 21%
Present 52 21%
Transhumance X13
Not present 200 79%
No activities 33 13%
Reliance on land X14 1-2 activities 166 66%
3-4 activities 53 21%
Half or more HHs 179 71%
Abandoned fishing activities X15
Less than half HHs 73 29%
Present 126 50%
Reliance on bore hole/wells X16
Not present 126 50%
Reliance on water trucking/ Present 70 28%
X17
tanker (government supply) Not present 182 72%
Present 170 67%
Biodiversity loss X18
Not present 82 33%
Present 219 87%
Soil degradation X19
Not present 33 13%
Present 201 80%
Increased temperature X20
Not present 51 20%
Less than 76% of HHs 159 63%
Water sufficiency X21
76% of HHs or more 93 37%
Present 35 14%
Reduced meal size as mitigation measure X22
Not present 217 86%
Present 18 7%
Spent savings as mitigation measure X23
Not present 234 93%
Present 53 21%
Borrowed money as mitigation measure X24
Not present 199 79%
Changed agricultural activi- Present 44 17%
X25
ties as mitigation measure Not present 208 83%
None 15 6%
Some = 1–24% 63 26%
Families in need of food X26 Less than half = 25–49% 46 18%
More than half = 50–75% 65 26%
Most or all = 76–100% 63 25%
Half or more HHs 128 51%
Unmet food need X27
Less than half HHs 124 49%
22 IOM IRAQ
VARIABLE STRUCTURE
Low = 0–2 mitigation measures 84 33%
Mitigation measures X28 Medium = 3–5 mitigation measures 115 46%
High = 6-8 mitigation measures 53 21%
Low = 0–10% 104 41%
Depopulation rate – Model 1
Y1 Medium = 11–30% 88 35%
High = 31–100% 60 24%
Low= 0–10% 104 41%
Depopulation rate – Model 2
Y2 Medium= 11–40% 108 43%
High= 41–100% 40 16%
Table 2 and Table 3 display information about the performance of the models in cent if the model is properly split into a training (80% of locations) and a test set
terms of their accuracy (that is, the proportion of correctly classified data using (20% of location) to deal with overestimation. Model 2 has an overall accuracy
the designed model) and predictive power. According to the classification matrix, of 92 per cent if all locations are used; however, accuracy drops to 47 per cent
the overall accuracy of Model 1 is 91 per cent if all locations are used and 61 per when the model is split into a training and a test set to deal with overestimation.
Table 2. Confusion Matrix
MODEL 1
Model 1 (all locations) Model 1 (train and test)
Observed
0–10% 11–30% 31–100% 0–10% 11–30% 31–100%
0–10% 104 12 5 12 4 2
11–30% 0 75 4 7 10 2
31–100% 0 1 51 1 3 8
MODEL 2
Observed
0–10% 11–30% 31–100% 0–10% 11–30% 31–100%
0–10% 103 12 4 12 10 2
11–30% 1 94 1 8 9 4
31–100% 0 2 35 0 2 2
Table 3. Classification Statistics
MODEL 1
Accuracy 0.9127 0.6122
95% Confidence Interval (0.8708, 0.9445) (0.4624, 0.748)
No Information Rate (NIR) 0.4127 0.4082
P-Value [Accuracy > NIR] < 2.2e-16 0.003128
Kappa 0.8643 0.4085
McNemar’s Test P-Value 0.0003007 0.716938
Class 1 Class 2 Class 3 Class 1 Class 2 Class 3
Sensitivity 1 0.8523 0.85 0.6 0.5882 0.6667
Specificity 0.8851 0.9756 0.9948 0.7931 0.7188 0.8919
Positive Predictive Value 0.8595 0.9494 0.9808 0.6667 0.5263 0.6667
Negative Predictive Value 1 0.9249 0.955 0.7419 0.7667 0.8919
Prevalence 0.4127 0.3492 0.2381 0.4082 0.3469 0.2449
Detection Rate 0.4127 0.2976 0.2024 0.2449 0.2041 0.1633
Detection Prevalence 0.4802 0.3135 0.2063 0.3673 0.3878 0.2449
Balanced Accuracy 0.9426 0.9139 0.9224 0.6966 0.6535 0.7793
23 IOM IRAQ
MODEL 2
Accuracy 0.9206 0.4694
95% Confidence interval (0.8801, 0.9508) (0.3253, 0.6173)
NIR 0.4286 0.4286
P-Value [Accuracy > NIR] < 2.2e-16 0.3307
Kappa 0.8717 0.1202
McNemar’s Test P-Value 0.003437 0.4091
Class 1 Class 2 Class 3 Class 1 Class 2 Class 3
Sensitivity 0.9904 0.8704 0.875 0.6 0.4286 0.25
Specificity 0.8919 0.9861 0.9906 0.5862 0.5714 0.95122
Positive Predictive Value 0.8655 0.9792 0.9459 0.5 0.4286 0.5
Negative Predictive Value 0.9925 0.9103 0.9767 0.68 0.5714 0.86667
Prevalence 0.4127 0.4286 0.1587 0.4082 0.4286 0.16327
Detection Rate 0.4087 0.373 0.1389 0.2449 0.1837 0.04082
Detection Prevalence 0.4722 0.381 0.1468 0.4898 0.4286 0.08163
Balanced Accuracy 0.9411 0.9282 0.9328 0.5931 0.5 0.60061
RANKING OF FACTORS99
Location-level data were used to investigate the relative importance of each measures – that is, the number of mitigation measures that families adopt at the
domain for environmental depopulation and identify the independent variables location – are respectively the fourth and fifth most important factors in Model
most associated with it. The ranking of factors, according to the mean decrease 1 and third and fourth in Model 2.
in the Gini coefficient,100 is depicted in Figure 1. Despite a slight difference in
Singularly, the cost of water trucking and damming or river diversions appear
ranking, both models selected the same variables as the top five factors.
to be of high relevance.101
The presence of multiple water issues is the single most important factor in
Other important risk factors are reliance on bore holes/wells, transhumance
Model 1 and in fifth position for Model 2. Food needs – that is ‘half or more
and fishing activities. Model 2 also outlines how environmental displacement
households or more are in need of food’ – is the second most important factor
that started earlier than 2021 and bio-diversity loss may be other red flags
in both models, whereas difficult access to services is the third most important
that may signal a path towards increasing depopulation. Tension issues – that is,
factor in the Model 1 and the most important in Model 2. This may point in the
a combined indicator taking into account tensions between groups, tension over
direction of particularly remote rural locations, where access to secondary
natural resources and whether there has been an increase in tension – ranked
education, health and markets may be more problematic. Reliance on land
within the first ten factors in both models. As a result, this indicator acquires a
(fourth most important factor in Model 1 and third in Model 2) and mitigation
relative importance.
99 Even though collinearity is mostly dealt with when it comes to predictive power, more caution is demanded when dealing with the scoring of factors. If two or more variables
are correlated, once one of the two is selected for splitting a node in a Tree, it is highly likely that the other variable will not be selected for further splitting since its effect has
already been taken into account by the selected variable. As a result, it does not provide further explanation of the output variation. Hence, the overall importance of the two
(or more) correlated variables in the Random Forest will be reduced.
100 The mean decrease in Gini coefficient is a measure of how each variable contributes to the homogeneity of the nodes and leaves in the resulting Random Forest. The higher
the value of mean decrease accuracy or mean decrease Gini score, the higher the importance of the variable in the model.
101 As noted in the Tree Model, this is often the case because a reduction in irrigation sources, a decrease in rainfall or in water quality were observed in nearly all locations, making
these factors unfit for classification purposes despite being the most significant risk factors.
24 IOM IRAQ
Figure 1. Ranking of factors based on mean decrease in Gini coefficient
10.09
Water issues
6.79
8.64
Families in need of food
7.87
8.28
Access to services and infrastructure
9.19
7.73
Reliance on land
7.46
6.48
Coping strategies
6.67
6.08
Cost of water trucking
5.74
5.76
Damming/river diversion
5.75
5.31
Tension issues
4.61
5.23
Reliance on bore hole/wells
5.02
5.10
Transhumance
4.34
5.03
Abandoned fishing activities
4.28
4.87
Broken/inefficient water infrastructure
4.70
4.78
Reduced water allocation
4.45
4.37
Reliance on water trucking/tanker (government supply)
3.93
4.25
Recent displacement
4.68
4.19
Biodiversity loss
4.74
4.06
Water sufficiency
4.27
3.55
Soil degradation
3.11
3.20
Small-holder farming
3.08
3.17
Environmental issues
2.19
3.08
Livelihood issues
2.98
3.08
Changed agricultural activities as coping strategy
3.27
2.97
Livestock rearing
2.74
2.96
Unmet food need
2.80
2.87
Borrowed money as coping strategy
3.90
2.77
Reduced meal size as coping strategy
3.25
1.97
Increased temperature
1.75
1.37
Spent savings as coping strategy
1.36
0.00 2.00 4.00 6.00 8.00 10.00 12.00
Model 1 Model 2
25 IOM IRAQ
Table 4. Ranking of factors
PREDICTORS MODEL 1 MODEL 2

Water issues 10.093595 6.788625
Families in need of food 8.635013 7.873919
Access to services and infrastructure 8.280772 9.187541
Reliance on land 7.729013 7.461919
Mitigation measures 6.482925 6.66863
Cost of water trucking 6.079559 5.744366
Damming/river diversion 5.763959 5.754345
Tension issues 5.306974 4.609381
Reliance on bore hole/wells 5.234257 5.016876
Transhumance 5.097684 4.34187
Abandoned fishing activities 5.029385 4.277717
Broken/inefficient water infrastructure 4.870359 4.7037
Reduced water allocation 4.77612 4.44546
Reliance on water trucking/tanker (government supply) 4.367511 3.929374
Recent displacement 4.251637 4.683404
Biodiversity loss 4.189264 4.736337
Water sufficiency 4.063657 4.269101
Soil degradation 3.553309 3.107334
Small-holder farming 3.196347 3.081293
Environmental issues 3.174667 2.191187
Livelihood issues 3.081334 2.976742
Changed agricultural activities as mitigation measure 3.080598 3.271313
Livestock rearing 2.974343 2.735444
Unmet food need 2.958935 2.799299
Borrowed money as mitigation measure 2.873577 3.896902
Reduced meal size as mitigation measure 2.76915 3.25481
Increased temperature 1.973129 1.750646
Spent savings as mitigation measure 1.372394 1.3639
26 IOM IRAQ
REFERENCES
Binder, L., B. Šedová, L. Rüttinger, J. Tomalka and S. Gleixner
2022 Climate Risk Profile: Iraq. Potsdam Institute for Climate Impact Research & adelphi, Potsdam.
Birkman, L., D. Kool and E. Struyken
2022 Water Challenges and Conflict Dynamics in Southern Iraq: An In-Depth Analysis of an Under-Researched Crisis. Water, Peace and Security.
Breiman, L.
1984 Classification and Regression Trees. First Edition. Routledge, New York.
2001 Random Forests. Machine Learning, 45:5–32.
Dockx, P.-J.
2019 Water Scarcity in Iraq: From Inter-Tribal Conflict to International Disputes. Institute of Peace and Conflict Studies.
Guiu, R.
2020 When Canals Run Dry: Displacement Triggered by Water Stress in the South of Iraq. Internal Displacement Monitoring Centre, Social Inquiry and Norwegian
Refugee Council, Geneva.
Hastie, T., R. Tibshirani and J. Friedman
2001 The Elements of Statistical Learning: Data Mining, Inference, and Prediction. Second Edition. Springer, Stanford.
Ho, T. K.
1995 Random decision forests. Proceedings of the Third International Conference on Document Analysis and Recognition, Institute of Electrical and
Electronics Engineers, Montreal.
1998 The random subspace method for constructing decision forests. Institute of Electrical and Electronics Engineers Transactions on Pattern Analysis and
Machine Intelligence, 20(8):832–844.
Human Rights Watch
2019 Basra is Thirsty: Iraq’s Failure to Manage the Water Crisis.
Intergovernmental Panel on Climate Change
2018 Global Warming of 1.5°C. An IPCC Special Report on the Impacts of Global Warming of 1.5°C Above Pre-Industrial Levels and Related Global Greenhouse Gas
Emission Pathways, in the Context of Strengthening the Global Response to the Threat of Climate Change, Sustainable Development, and Efforts to Eradicate
Poverty. Cambridge University Press, Cambridge and New York.
27 IOM IRAQ
International Organization for Migration (IOM)
2007 Discussion Note: Migration and the Environment (MC/INF/288 of 1 November 2007), prepared for the Ninety-Fourth Session of the IOM Council.
Geneva.
2011 Intersessional Workshop on Climate Change, Environmental Degradation and Migration: Background Paper. International Dialogue on Migration, Geneva.
2014 Migration, Environment and Climate Change: Evidence for Policy Glossary. Geneva.
2019 International Migration Law No. 34, Glossary on Migration. Geneva.
2022a Iraq – Displacement Tracking Matrix (DTM) – Climate-Induced Displacement – Central and Southern Iraq (1-15 September 2022). Baghdad.
2022b Migration, Environment and Climate Change in Iraq. Baghdad.
2022c People on the Move in a Changing Climate – Linking Policy, Evidence and Action. Geneva.
2022d Somalia – Drought Related Displacement Analysis (December 2021). Mogadishu.
2023 Climate-Induced Displacement – Central and Southern Iraq. July, Baghdad.
n.d. Environmental Migration. Geneva.
n.d. Iraq – Displacement Tracking Matrix (DTM) – Climate Methodology. Baghdad.
n.d. IOM Perspectives on Migration, Environment and Climate Change. Geneva.
IOM and Social Inquiry
2022a A Climate of Fragility – Household Profiling in the South of Iraq: Basra, Thi-Qar and Missan. Baghdad.
2022b Migration into a Fragile Setting: Responding to Climate-Induced Informalization and Inequality in Basra, Iraq. Baghdad.
Landsat Missions
n.d. Landsat Normalized Difference Vegetation Index. United States Geological Survey, Reston.
Maddocks, A., R. Samuel Young and P. Reig
2015 Ranking the World’s Most Water-Stressed Countries in 2040. World Resources Institute.
Mahmalat, M. and A. Ahmad
2023 In Iraq, mitigating fragility means addressing climate change [blog post]. World Bank Group, 18 April.
Oxfam
2018 Measuring Household Stress: Introducing the multi-sector Coping Strategy Index for Afghanistan. Kabul.
Piesse, M.
2016 Water Governance in the Tigris-Euphrates Basin. Future Directions International, Dalkeith.
REACH, United Nations Children’s Fund (UNICEF) and United Nations Institute for Training and Research (UNITAR)
2020 Flood Hotspots in Iraq (October 2018 – March 2019). Baghdad.
28 IOM IRAQ
Salman, S. A., S. Shahid, T. Ismail, E.-S. Chung and A. M. Al-Abadi
2017 Long-term trends in daily temperature extremes in Iraq. Atmospheric Research, 198:97-107.
Schmidt, C. and A. McCullum
2016 Creating and Using Normalized Difference Vegetation Index (NDVI) from Satellite Imagery. NASA Applied Remote Sensing Training Program (ARSET),
Washington, D.C.
The International Business Machines Corporation (IBM)
2012 IBM SPSS Decision Trees 21.
United Nations Country Team in Iraq, Joint Analysis and Policy Unit
2013 Sand and Dust Storms Fact Sheet. Baghdad.
United Nations Development Programme (UNDP)
n.d. UNDP Conceptual Framework of Social Cohesion.
United Nations Office for Outer Space Affairs
n.d. Risks and Disasters.
Von Lossow, T.
2018 More than infrastructures: water challenges in Iraq. Planetary Security Initiative & Clingendael.
World Bank Group
2017 Iraq: Systematic Country Diagnosis.
2023 Iraq Economic Monitor – Reemerging Pressures: Iraq’s Recovery at Risk. Spring/Summer. Washington D.C.
World Food Programme and Social Inquiry
2022a Improving Prospects for Peace and Stability in Vulnerable Communities in Southern Iraq. Thi-Qar Governorate Conflict Analysis. Baghdad.
2022b Prospects for Resilience Amid Fragility: Conflict Analysis of Al-Qurna and Al-Dair Districts in Basra Governorate. Baghdad.
World Health Organization and United Nations Framework Convention on Climate Change
2021 Iraq – Health and Climate Change Country Profile 2021. Geneva.
Younis, N.
2022 Early warning: How Iraq can adapt to climate change. European Council on Foreign Relations, Berlin.
29 IOM IRAQ
IOM IRAQ
iraq.iom.int UNAMI Compound (Diwan 2),

International Zone,
[email protected] Baghdad/Iraq @IOMIraq
IOM Iraq thanks the U.S. Department of State, Bureau of Population, Refugees and Migration (PRM) for its continued support.
© 2023 International Organization for Migration (IOM)
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means,
electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher.

Politics of Climate-Inducted Displacement in Iraq

Uploaded by

Copyright:

Available Formats

Politics of Climate-Inducted Displacement in Iraq

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Politics of Climate-Inducted Displacement in Iraq

Uploaded by

Copyright:

Available Formats

DRIVERS OF

This publication was issued without formal editing by IOM.

International Organization for Migration

© 2023 International Organization for Migration (IOM)

DEFINITIONS AND INDICATORS 8

CHALLENGES AND LIMITATIONS 9

DRIVERS OF CLIMATE-INDUCED DEPOPULATION 11

ENVIRONMENTAL HAZARDS AND WATER ACCESS 12

ACCESS TO SERVICES AND INFRASTRUCTURE 15

LIVELIHOODS AND MITIGATION MEASURES 16

TENSION AND CONFLICT 17

CHANGES IN VEGETATION HEALTH IN HOTSPOT LOCATIONS 18

AREAS FOR FURTHER DATA COLLECTION AND RESEARCH 20

CHAID Chi-Square Automatic Interaction Detection

DTM Displacement Tracking Matrix

IOM International Organization for Migration

NASA National Aeronautics and Space Administration

NDVI Normalized Difference Vegetation Index

NIR No Information Rate

RARTs Rapid Assessment and Response Teams

WFP World Food Programme

To better understand the root causes of these movements, the International

OVERALL VULNERABILITY DISPLACEMENT DYNAMICS

Figure 1. Top five predictors of climate-induced depopulation

Al-Rifa'i Qal'at Saleh Afaq Hamza Amara

2 Families struggling to meet food needs 8.64

3 Service and infrastructure issues 8.28

4 Reliance on land for livelihoods 7.73

5 Adoption of mitigation measures 6.48

Map 1. Portion of original population displaced due to environmental factors

Al-Salman Basrah Abu

SAUDI ARABIA Fao

KUWAIT Persian Gulf

ENVIRONMENTAL HAZARDS AND WATER ACCESS SERVICES AND INFRASTRUCTURE

LIVELIHOODS climate-smart infrastructure and awareness-raising campaigns around

• Qadissiya Governorate: Afaq and Hamza districts;

Iraq, improvement of existing water infrastructure and acquisition of on their data.

LACK OF ACCESS TO LIVELIHOODS CHALLENGES AND LIMITATIONS

Map 2. Portion of original population displaced due to environmental factors

Ramadi Abu Ghraib

Al-Salman Basrah Abu

SAUDI ARABIA Fao

KUWAIT Persian Gulf

Number of families displaced

Governorate, are completely abandoned due to climate-related issues (Table 1).

51% 251 329

District, Governorate Number of locations DRIVERS OF CLIMATE-INDUCED DEPOPULATION

59 Based on notes from DTM RARTs.

Najaf Nassriya Al-Midaina

SAUDI ARABIA Fao

KUWAIT Persian Gulf

Figure 7. Most prevalent environmental hazards by percentage of locations

Irrigation canal 61%

Water truck/tanker from

CHANGES IN WATER SUPPLY

Map 4. Percentage of locations with six or more water issues by district

SAUDI ARABIA Fao

KUWAIT Persian Gulf

Figure 9. Percentage of locations by water supply issues