The In˜uence of Climate Change on Human Migration: AI ......The In˜uence of Climate Change on...

The In�uence of Climate Change on Human Migration: A Case Study of South Africa

Rachel Licker1 ([email protected]), Marina Mastrorillo1 ([email protected]) , Michael Oppenheimer1 ([email protected]), Lyndon Estes1 ([email protected]), and Ruohong Cai1 ([email protected])

1Woodrow Wilson School of Public & International A�airs, Princeton University, Princeton, NJ 08544

Abstract Both gradual changes in climate as well as extreme events have been shown to in�uence human migration patterns in many regions of the world. Given the implications of migration for the well-being and security of migrants, as well as that of the communities sending and receiving them, it is important for policymakers from both the federal and international level to be prepared for the extent and manner in which climate change will alter migration �ows. Here we provide an overview of a new line of research aimed at identifying the relative in-�uence of climate on recent internal migration decisions in South Africa. We focus on South Africa as it is a country that has high-volume internal migration �ows, is projected to experience dramatic shifts in climate, and has a large population base that is vulnerable to climate change through, for example, its reliance on rainfed agriculture. We utilize the results of the nationally representative National Income Dynamics Survey (NIDS) to construct statistical models that predict the probability of an individual decision to migrate based on individual characteris-tics (e.g. age, gender, race, and education-level), household-level characteristics (e.g. assets and migration-network connections), and local climate (e.g. minimum and maximum temperature, as well as precipitation over monthly, seasonal, and annual time periods). We also incorporate local agricultural production data into the models as a proxy for the indirect e�ects of climate on local conditions. With this study, and together with a macro-level study being carried out in parallel, we consider the potential for projected changes in climate to in�uence South African migration patterns, as well as identify the characteristics of individuals whose de-cisions to stay or move will most likely be in�uenced by changing climate.

Poster Number CC-14

Climate and Migration

• Climate change – potential to dramatically alter human migration �ows.8

• For example, rural residents may seek additional sources of income in other rural or urban locations as crop growing conditions change.7,9

• Individuals may also move in increased numbers as a result of natural hazards.8

• Increased movement as a result of climate change is projected to have adverse consequences for human security.1

• To begin to understand how climate change will a�ect human migration: - Many studies look at the relationship between climate variability and mobility.2,6,10

References(1) Adger, W.N. et al., 2014: Human security. In: Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Field, C.B. et al. (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. XXX-YYY.(2) Barrios, S. et al., 2006. Climatic change and rural–urban migration: The case of sub-Saharan Africa. J. Urban Economics, 60(3), pp.357–371. (3) Chaney, N.W. et al., 2014. Spatial analysis of trends in climatic extremes with a high resolution gridded daily meteorological data set over sub-saharan africa. Journal of Climate, (in review)(4) Clarke, R. & K. Eyal, 2013. Microeconomic determinants of spatial mobility in post-Apartheid South Africa: Longitudinal evidence from the National Income Dynamics Study. Development South Africa. 31(1), pp.168-194.(5) Collinson, M. et al., 2006. Migration and changing settlement patterns: Multilevel data for policy. Report 03-04-01, Pretoria: Stats South Africa.(6) Feng, S. et al., 2010. Linkages among climate change, crop yields and Mexico–US cross-border migration. Proceedings of the National Academyof Sciences, 107(32), pp.14257–14262. (7) Feng, S. et al., 2012. Climate Change, Crop Yields, and Internal Migration in the United States. National Bureau of Economic Research. Working Paper 17734, p.43.(8) Foresight, 2011. Migration and Global Environmental Change: Future Challenges and Opportunities, Gov. O�ce for Science, London, UK.(9) Gray, C.L. & Mueller, V., 2012. Natural disasters and population mobility in Bangladesh. Proceedings of the National Academy of Sciences of the United States of America, 109(16), pp.6000–5. (10) Kniveton, D.R. et al., 2012. Emerging migration �ows in a changing climate in Dryland Africa. Nature Climate Change, 2, pp.444-447.(11) Kruger, A.C & Sekele, S.S., 2013. Trends in extreme temperature indices in South Africa: 1962-2009. Int. Journal of Climatology. 33, pp.661-676.(12) Leibbrandt, M. et al., 2010, Trends in South African Income Distribution and Poverty since the Fall of Apartheid, OECD Social, Employment and Migration Working Papers, No. 101, OECD Publishing, © OECD. doi:10.1787/5kmms0t7p1ms-en(13) National Planning Commission, 2013. Wave 3 Overview: NIDS. The Presidency Department, Republic of South Africa, pp.40.(14) Sen Roy, S. & M. Rouault, 2013. Spatial patterns of seasonal scale trends in extreme hourly precipitation in South Africa. Applied Geography., 39, pp.151-157.(15) Statistics South Africa, 2014. Poverty. [Online] Beta2.statssa.gov.za

(16) van Oldenborgh, G.J. et al., 2014. Annex I: Atlas of Global and Regional Climate Projections. In: Climate Change 2014: The Physical Science Basis. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F. et al. (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 1311-1393.

South Africa

Changes in climate already observed • General warming, warm extremes, in cold extremes11 • Changes in precipitation extremes, + trends in summer, - & + trends in winter14

Significant and consequential changes in climate projected

Year

0

10000

20000

30000

40000

50000

60000

70000

80000

1993 1995 2000 2008

Ann

ual P

er C

apita

Inco

me

(200

0 Ra

nds)

Year

White Mixed Ethnic Origin Asian African

Ann

ual p

er c

apita

inco

me

(200

0 Ra

nds)

(data from Leibbrandt, 2010)

Key Questions

• How has climate in�uenced the internal migration �ows in South Africa in recent history?

• Given historical relationships, how might future changes in climate alter human migration �ows within South Africa?1364

Annex I Atlas of Global and Regional Climate Projections

AI

Figure AI.50 | (Top left) Time series of relative change relative to 1986–2005 in precipitation averaged over land grid points in Southern Africa (35°S to 11.4°S, 10°W to 52°E) in October to March. (Top right) Same for sea grid points in the West Indian Ocean (25°S to 5°N, 52°E to 75°E). Thin lines denote one ensemble member per model, thick lines the CMIP5 multi-model mean. On the right-hand side the 5th, 25th, 50th (median), 75th and 95th percentiles of the distribution of 20-year mean changes are given for 2081–2100 in the four RCP scenarios.

(Below) Maps of precipitation changes in 2016–2035, 2046–2065 and 2081–2100 with respect to 1986–2005 in the RCP4.5 scenario. For each point, the 25th, 50th and 75th percentiles of the distribution of the CMIP5 ensemble are shown; this includes both natural variability and inter-model spread. Hatching denotes areas where the 20-year mean differences of the percentiles are less than the standard deviation of model-estimated present-day natural variability of 20-year mean differences.

Sections 9.4.1.1, 9.6.1.1, Box 11.2, 12.4.5.2, 14.8.7 contain relevant information regarding the evaluation of models in this region, the model spread in the context of other methods of projecting changes and the role of modes of variability and other climate phenomena.

-80

-60

-40

-20

0

20

40

60

80

1900 1950 2000 2050 2100-80

-60

-40

-20

0

20

40

60

80

(%)

Precipitation change Southern Africa October-March

RCP8.5RCP6.0RCP4.5RCP2.6historical

2081-2100 mean

-80

-60

-40

-20

0

20

40

60

80

1900 1950 2000 2050 2100-80

-60

-40

-20

0

20

40

60

80

(%)

Precipitation change West Indian Ocean October-March


2081-2100 mean

1364


AI




-80

-60

-40

-20

0

20

40

60

80

1900 1950 2000 2050 2100-80

-60

-40

-20

0

20

40

60

80

(%)



2081-2100 mean

-80

-60

-40

-20

0

20

40

60

80

1900 1950 2000 2050 2100-80

-60

-40

-20

0

20

40

60

80

(%)



2081-2100 mean

1364


AI




-80

-60

-40

-20

0

20

40

60

80

1900 1950 2000 2050 2100-80

-60

-40

-20

0

20

40

60

80

(%)



2081-2100 mean

-80

-60

-40

-20

0

20

40

60

80

1900 1950 2000 2050 2100-80

-60

-40

-20

0

20

40

60

80

(%)



2081-2100 mean

1365

Atlas of Global and Regional Climate Projections Annex I

AI

Figure AI.51 | (Top left) Time series of relative change relative to 1986–2005 in precipitation averaged over land grid points in Southern Africa (35°S to 11.4°S, 10°W to 52°E) in April to September. (Top right) Same for sea grid points in the West Indian Ocean (25°S to 5°N, 52°E to 75°E). Thin lines denote one ensemble member per model, thick lines the CMIP5 multi-model mean. On the right-hand side the 5th, 25th, 50th (median), 75th and 95th percentiles of the distribution of 20-year mean changes are given for 2081–2100 in the four RCP scenarios.



-80

-60

-40

-20

0

20

40

60

80

1900 1950 2000 2050 2100-80

-60

-40

-20

0

20

40

60

80

(%)

Precipitation change Southern Africa April-September


2081-2100 mean

-80

-60

-40

-20

0

20

40

60

80

1900 1950 2000 2050 2100-80

-60

-40

-20

0

20

40

60

80

(%)

Precipitation change West Indian Ocean April-September


2081-2100 mean

1365


AI

Figure AI.51 | (Top left) Time series of relative change relative to 1986–2005 in precipitation averaged over land grid points in Southern Africa (35°S to 11.4°S, 10°W to 52°E) in April to September. (Top right) Same for sea grid points in the West Indian Ocean (25°S to 5°N, 52°E to 75°E). Thin lines denote one ensemble member per model, thick lines the CMIP5 multi-model mean. On the right-hand side the 5th, 25th, 50th (median), 75th and 95th percentiles of the distribution of 20-year mean changes are given for 2081–2100 in the four RCP scenarios.



-80

-60

-40

-20

0

20

40

60

80

1900 1950 2000 2050 2100-80

-60

-40

-20

0

20

40

60

80

(%)

Precipitation change Southern Africa April-September


2081-2100 mean

-80

-60

-40

-20

0

20

40

60

80

1900 1950 2000 2050 2100-80

-60

-40

-20

0

20

40

60

80

(%)

Precipitation change West Indian Ocean April-September


2081-2100 mean

Precipitation change RCP4.5 in 2081-2100: October-March

1363


AI

Figure AI.49 | (Top left) Time series of temperature change relative to 1986–2005 averaged over land grid points in Southern Africa (35°S to 11.4°S, 10°W to 52°E) in June to August. (Top right) Same for sea grid points in the West Indian Ocean (25°S to 5°N, 52°E to 75°E). Thin lines denote one ensemble member per model, thick lines the CMIP5 multi-model mean. On the right-hand side the 5th, 25th, 50th (median), 75th and 95th percentiles of the distribution of 20-year mean changes are given for 2081–2100 in the four RCP scenarios.

(Below) Maps of temperature changes in 2016–2035, 2046–2065 and 2081–2100 with respect to 1986–2005 in the RCP4.5 scenario. For each point, the 25th, 50th and 75th percentiles of the distribution of the CMIP5 ensemble are shown; this includes both natural variability and inter-model spread. Hatching denotes areas where the 20-year mean differences of the percentiles are less than the standard deviation of model-estimated present-day natural variability of 20-year mean differences.

Sections 9.4.1.1, 9.6.1.1, 10.3.1.1.4, Box 11.2, 14.8.7 contain relevant information regarding the evaluation of models in this region, the model spread in the context of other methods of projecting changes and the role of modes of variability and other climate phenomena.

-2

0

2

4

6

8

1900 1950 2000 2050 2100

-2

0

2

4

6

8

(°C

)

Temperature change Southern Africa June-August


2081-2100 mean

-2

0

2

4

6

8

1900 1950 2000 2050 2100

-2

0

2

4

6

8

(°C

)

Temperature change West Indian Ocean June-August


2081-2100 mean

1363


AI




-2

0

2

4

6

8

1900 1950 2000 2050 2100

-2

0

2

4

6

8

(°C

)



2081-2100 mean

-2

0

2

4

6

8

1900 1950 2000 2050 2100

-2

0

2

4

6

8

(°C

)



2081-2100 mean

1362


AI

Figure AI.48 | (Top left) Time series of temperature change relative to 1986–2005 averaged over land grid points in Southern Africa (35°S to 11.4°S, 10°W to 52°E) in December to February. (Top right) Same for sea grid points in the West Indian Ocean (25°S to 5°N, 52°E to 75°E). Thin lines denote one ensemble member per model, thick lines the CMIP5 multi-model mean. On the right-hand side the 5th, 25th, 50th (median), 75th and 95th percentiles of the distribution of 20-year mean changes are given for 2081–2100 in the four RCP scenarios.



-2

0

2

4

6

8

1900 1950 2000 2050 2100

-2

0

2

4

6

8

(°C

)

Temperature change Southern Africa December-February


2081-2100 mean

-2

0

2

4

6

8

1900 1950 2000 2050 2100

-2

0

2

4

6

8

(°C

)

Temperature change West Indian Ocean December-February


2081-2100 mean

1363


AI




-2

0

2

4

6

8

1900 1950 2000 2050 2100

-2

0

2

4

6

8

(°C

)



2081-2100 mean

-2

0

2

4

6

8

1900 1950 2000 2050 2100

-2

0

2

4

6

8

(°C

)



2081-2100 mean

1362


AI

Figure AI.48 | (Top left) Time series of temperature change relative to 1986–2005 averaged over land grid points in Southern Africa (35°S to 11.4°S, 10°W to 52°E) in December to February. (Top right) Same for sea grid points in the West Indian Ocean (25°S to 5°N, 52°E to 75°E). Thin lines denote one ensemble member per model, thick lines the CMIP5 multi-model mean. On the right-hand side the 5th, 25th, 50th (median), 75th and 95th percentiles of the distribution of 20-year mean changes are given for 2081–2100 in the four RCP scenarios.



-2

0

2

4

6

8

1900 1950 2000 2050 2100

-2

0

2

4

6

8

(°C

)

Temperature change Southern Africa December-February


2081-2100 mean

-2

0

2

4

6

8

1900 1950 2000 2050 2100

-2

0

2

4

6

8

(°C

)

Temperature change West Indian Ocean December-February


2081-2100 mean

• RCP 8.5, 2081-2100substantial consecutive droughtday index

• 10% Apr-Septprecipitation acrossmuch of South Africa

• Coincides with partof maize growing season (end of Sept-end of June)

• Extensive warmingthroughout the year,particularly inland.

IPCC AR5 WGI Annex 116

Data

Climate data

Variables: • Annual, Seasonal, Monthly -- Precipitation & Tmin, Tmax, Tave

• Derived from 0.25 degree daily reanalysis data3

Individual & Household Data

• NIDS Survey (University of Cape Town)13: 2008, 2010, 2012, 2014* (in progress)• Nationally representative panel study • Tracks individuals• Question themes include: Migration; Livelihoods; Health & Education; Household Composition & Structure; Labour Market Participation & Economic Activity

Between Wave 1-3: 3190 individuals moved • 1 out of 7 South Africans 13• 80% individuals, 20% whole household• 40% of Wave 1 individuals - shared household with a migrant• Attrition - high for migrants in Wave 2, Negative between Wave 2 & 3

Next Steps• Development of statistical models - test of approaches that account for attrition• Testing & selection of predictor variables• Comparison of results with parallel macro-level study

NIDS Survey

Households (7296)

Household Members (31163)

Adult (16871)

Child (9605)

Proxy (1750)

Wave 1, 2008

Non-Resident, Temporary Sample Members

(2916)

Continuing Sample Members (28247)

Refused/ Not Available

Not located

Not tracked

Moved Outside SA

Died

An important country to examine potential e�ects of climate change on migration

South Africa

Highly mobile population• 1996-2001: 12% of population moved, 0.3% international, 11.7% internal5• High rate of temporary labor migration - Legacy of Apartheid

Many highly vulnerable individuals - unevenly distributed across population

• 56.8%: 2009 Poverty level15 • 4 million people: Smallholder agriculture• Apartheid by-product: Income distribution inequalities

Past and Future Climate Change

Study Design

Statistical model - Probability of individual carrying out an internal migration given:

• Individual characteristics

• Household characteristics

• Local characteristics

{ e.g. age, gender, race, education, labor status }

{ e.g. income, assets, network connections }

{ e.g. monthly, seasonal & annual climate, annual maize & wheat yield }

The In˜uence of Climate Change on Human Migration: AI ......The In˜uence of Climate Change on...

Documents

Transcript of The In˜uence of Climate Change on Human Migration: AI ......The In˜uence of Climate Change on...