An integrated framework to assess plausible future

livelihood and poverty changes in deltas:

an application to coastal Bangladesh

AGU Fall meeting 2015

GC43G: Sustainable Deltas:

Multidisciplinary Analyses of Complex Systems II

17 December 2015, San Francisco

A.N. Lázár1, A. Payo1, R. J. Nicholls1, C. Hutton1, H. Adams2, M. Salehin3, A. Haque3, D. Clarke1, L. Bricheno4, J.A. Fernandes5, Mofizur Rahman6, Ali Ahmed6, P.K. Streatfield6

1 University of Southampton, United Kingdom, a.lazar@soton.ac.uk2 University of Exeter, United Kingdom3 Bangladesh University of Engineering & Technology, Bangladesh4 National Oceanography Centre, Liverpool, United Kingdom5 Plymouth Marine Laboratory, United Kingdom6 International Centre for Diarrhoeal Disease Research, Bangladesh

ESPA Deltas: Overarching aim

To provide policy makers with the knowledge and tools to enable them to evaluate the effects of policy decisions on ecosystem services and

people's livelihoods

Vision: Link science to policy at the landscape scale

What are the key drivers?

How will these change with time and how do they interact?

What are the consequences of these changes for ecosystem services?

How will these affect the people, particularly the poor?

How can policy processes use this science?

The Consortium21 partners and about 100 members from a range of disciplines

UK (7 partners)

• University of Southampton- Lead Robert Nicholls PI

• University of Oxford

• Exeter University

• Dundee University

• Hadley Centre MET office

• Plymouth Marine Laboratories

• National Oceanography Centre Liverpool

Bangladesh (12 partners)

• Institute of Water and Flood Management, Bangladesh University of Engineering and Technology (BUET) – Prof Rahman Lead PI

• Bangladesh Institute of Development Studies (BIDS)

• Institute of Livelihood Studies (ILS)

• Ashroy Foundation

• Institute of International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B)

• Center for Environmental and Geographic Information Services (CEGIS)

• Bangladesh Agricultural University

• Bangladesh Agricultural Research Institute (BARI)

• Technological Assistance for Rural Advancement (TARA)

• International Union for Conservation of Nature (IUCN)

• University of Dhaka

• Water Resources Planning Organization (WARPO)

India (2 partners)

• Jadavpur University: Indian Lead

• IIT Kanpur 3

Study Area: Land UseComposite ESPA Deltas Project Analysis

Science Questions from National Policy MakersGeneral Economic Division, Planning Commission, Bangladesh

Sealevel rise?

(… cm/100y)

More /extreme

storms?More/intense

rainfall?

Spatial

developments

Subsidence?

More summer

Drought?

Salt

Intrusion?

Decreased

river

Discharge?

Increased

river

Discharge?

Increased

Erosion?

5

Impacts on ecosystem services and rural livelihoods, infrastructure, poverty disaster preparedness of the above.

Model Inputs

Climate-precipitation-temperature-evaporation

Bay of Bengal-mean sea level-(subsidence)

Economy-market price-cost of farm inputs-wages

Levees/Polders-location-height-drainage rate

Demography-life expectancy-fertility rate-migration rate

Hydrology-discharge-sediment

Ecosystem Services-agriculture-aquaculture-fisheries-mangroves

Governance-subsidies-land use planning-infrastructure planning

-cyclone-storm surge

Hazards

Outputs

• fish catches• net earnings from

- farming, - aquaculture & - fishing

Livelihoods

• river salinity• groundwater salinity• union-wise soil salinity • crop productivity

Salinisation

I. Household outputs: a) Bayesian statistical module:

• asset-based relative poverty indicator

b) Process-based module:• economics (income, costs/expenses, savings/assets)

• relative wealth-level• calories / protein intake / BMI• monetary poverty indicators

Household Wellbeing, Poverty & Health

• water elevation • inundated area

Coastal hydrology

II. Regional economic outputs • sectoral output (tons, BDT)• GINI• GDP/capita• income tax revenue• household debt level

Bio-physical environment emulation is based on high fidelity models

• Climate (Met Office Hadley Centre)

• Hydrology (INCA, Delft-3D, FVCOM, ModFlow-SeaWat)

• Bay of Bengal (POLCOMS-GCOMS, fisheries species model)

• Mangrove (SLAMM, Markov chain & cellular automata model)

Own development

• regional soil salinity model

• extended FAO CROPWAT model(with salinity, temperature, CO2, aquaculture)

Verification / Validation

Soil salinity conceptual model

Who are the participants in ESPA Deltas and ΔDIEM development?

50+ agencies

Stakeholders

Five main groups of participants

1586 surveyed households

Households Integrators UsersESPA Deltas team

~100 specialists

ClimateDemographyAgricultureOceanographyFisheriesMangrovesSocial scienceEconomicsWater resource management

Long iteration route that involves seeking advice from a broader team

v

10

Iterative learning with stakeholders

Stakeholder

Integrator

ESPA Deltas Team

Shorter iteration, running ΔDIEM with different inputs, SSPs,…

Users

Stakeholder EngagementWorkshop at General Economic Division, Dhaka16 September 2015

11

StakeholderScenarios for Bangladesh

IntegratedModel

Socio/EnvironmentalModels

Data

SimulationsS1, S2, …, Sn

AdaptationResponses

(e.g. coastal defence, irrigation projects, etc.)

Climate change, etc.

Qu

alit

ativ

e

Qu

anti

tati

ve

Sem

i-Q

ual

itat

ive

Qu

anti

tati

ve

ScenariosNarrative

Bas

elin

e

iterative learning loop

Participatory Modelling(an Iterative Learning Loop)

ExpertsStakeholder

Meeting

13

Stakeholder

User

ESPA Deltas Team

Planned Interventions & Governance

Bussiness as Usual

More Sustainable

Less Sustainable

Global Climate/Demographic/Economic system

Delta Hydrology

Bay of Bengal

Integrator

Fisheries&

AquacultureAgriculture

Mangrove

System startup

Human Wellbeing changes and responses

Household Health, Food & Income

Emulators

Governance Socio-EconomicPhysical & Ecological

nursery

spwaning areas

Floods protection

system

Historical dataHH Survey DataHousehold

DDIEMStructure Issues, scenarios,

interventions

14

ESPA Deltas: Components

Bay Bengal ModelGCOMS

GCM/ RCM

Catchment Models:GWAVA / INCA MODFLOW HydroTrend

Delta ModelFVCOM,Delft3D

Crop Model: CROPWAT

Coastal Fisheries ModelSize- & Species-based models

Temp, rainfall

Sea

leve

l, SL

P, S

ST, w

ind

s

Water, sediment, nutrients

Water flow, level, salinity, temp, sediment, nutrients

Primary productivity, T,S,O2, currents

Mangrove Model

Quantitative Physical/Ecological Models

Inland Fisheries Model

Morphology &

Land CoverAquaculture Model

Surg

e le

vel

Lan

d U

se

Laws, policies

Gaps, Conflicts, Implementation efficiencies

Key issues, Scenarios

Governance research

Know

ledge inte

gra

tion /

Scenarios (

ΔD

IEM

)

Populationprojections

Demographics, economics & poverty

Qualitative surveyrelationship b/t environment & social behaviour

Quantitative survey(consumption, assets, employment, migration, health, poverty, …)

Pro

cess

un

der

stan

din

g fo

r ea

ch s

oci

o-e

colo

gica

l gro

up

+ q

uan

tifi

ed b

ehav

iou

r

Spatial associative model b/t land use and poverty

Plausible futures

SRES

A1

B

(RC

P4

.5/6

-8

.5)

Development Scenarios

Less Sustainable

(LS)

Business As Usual

(BAU)

More Sustainable

(MS)

warmer (Q8)

moderately warmer &

wetter (Q0)

warmer & wetter(Q16)

by

20

50

16

Key factors associated with highest asset poverty (Census): Red is soil salinity, Green is waterlogging,

Yellow is access to market Amoaka Johnson et al. (in review) Sustainability Science.

R.J.Nicholls@soton.ac.ukR.J.Nicholls@soton.ac.uk

Q0BAUQ0LS Q0MS

Q8BAUQ8LS Q8MS

Q16BAUQ16LS Q16MS

2050Asset-poverty: Likelihood for being in the poorest quintal

i) Climate has minimal effect

ii) Changes in sustainability practices impacts around the Sundarbans fringe

iii) The stubborn poverty in the East is due to access and infrastructure

Less Sustainable Future 2050

More Sustainable Future 2050

Investment in Sustainability (water resource management, agricultural adaptations, coastal defence)

Requires investment in transport infrastructure

Sustainability levels identified by stakeholders

World Bank (national)Simulation (min/mean/max)

All 9 scenarios shows a decrease in the mean $1.9 poverty (consumption-based;

% population), but some unions remain very poor

The GINI coefficient measures the inequality. The HIES GINI coefficient for rural areas is captured well with the simulated mean value. The national average World Bank GINI coefficient is lower than our simulated value. However, we simulate remote rural areas, where we expect larger income differences. Spatially, the year 2014 is plotted. It seems that the Barisal division has higher GINI coefficients, thus the income-based inequality is larger there.

HIES (rural, 2000-2010), World Bank (national, 2013-14)Simulation (min/mean/max)

Q0 LS Q0 BAU Q0 MS

Q8 LS Q8 BAU Q8 MS

Q16 LS Q16 BAU Q16 MS

2050

Mean socio-economic inequality (GINI, % population) decreases, but some areas remain very inequal HIES (rural, 2000-2010),

World Bank (national, 2013-14)Simulation (min/mean/max)

ΔDIEM builds on simulators & observationsPrimary/secondary data and expert knowledge provide the basis of the Household module

ESPA Deltas seasonal household survey & HIES datasets:

• 21 household archetypes based on seasonality of livelihoods

• economic decisions

• poverty/health indicators

Archetypes Season 1 Season 2 Season 3

1 Small Business Small Business Small Business

2 Small scale Manufacturing Small scale Manufacturing Small scale Manufacturing

3 Farming Farming Farming

4 Fishing Fishing Fishing

5 Small scale Manufacturing Farm Labour Small scale Manufacturing

6 Small scale Manufacturing Small Business Small Business

7 Small Business Small Business Farming

8 Small Business Small Business Small scale Manufacturing

9 Farming no Job Farming

10 Farming Small scale Manufacturing Small scale Manufacturing

11 Farming Small Business Farming

12 Small scale Manufacturing Small Business Small scale Manufacturing

13 Farming Small Business Small Business

14 Small scale Manufacturing Fishing Small scale Manufacturing

15 Small Business Small scale Manufacturing Small Business

16 Farm Labour Farm Labour Small scale Manufacturing

17 Farm Labour Small scale Manufacturing Small scale Manufacturing

18 Small scale Manufacturing Small scale Manufacturing Small Business

19 Small Business Small scale Manufacturing Small scale Manufacturing

20 Forest Dependent in >1 season

21 All Other types

Observation (HIES)Simulation (min/mean/max)

Simulated mean household food expenditure (BDT/month) follows sparse observations, but shows large annual fluctuations pre-2020

Q0-BAU

Labsa (Satkhira)

Observations:IWM Annual Research Report (BARI 2009-2014)

Simulated soil salinity (dS/m)reproduces observed seasonality & magnitude

Q0 LS Q0 BAU Q0 MS

Q8 LS Q8 BAU Q8 MS

Q16 LS Q16 BAU Q16 MS

Inundated Area (% change 2050-2000) for non-protected land greatly increases under Q16 scenario, but not under other climates

Take home message

ΔDIEM:

o a transdisciplinary tool

o a holistic way for thinking and communication

Baseline projection results:

o Socio-economic scenarios are more important than climate scenarios at household level.

o Importance of Ecosystem Services is likely to decrease

o Poverty and Inequality are likely to decrease, but not in marginal areas