AGWA: Risk Management Framework for Water Resources Climate Adaptation

Rolf Olsen,1 PhDEugene Stakhiv,1,2 PhD

1Institute for Water ResourcesU.S. Army Corps of EngineersAlexandria, Virginia, USA

2Johns Hopkins UniversityBaltimore , Maryland, USA

Outline

• Background on Alliance for Global Water Adaptation (AGWA)

• Risk Management Framework– Breakout sessions– Next steps – stress tests

• Examples– Flood risk– Reservoir regulation

• Application of framework to United States-Canada Great Lakes Study– Example: ecosystems

AGWA: A Brief Overview

• The Alliance for Global Water Adaptation is a group of regional and global development banks, aid agencies and governments, a diverse set of non-governmental organizations (NGOs), and the private sector focused on how to manage water resources in way that is sustainable even as climate change alters the global hydrological cycle.

• Focused on how to help practitioners, investors, and water planners and managers make systematic, consistent, and resilient decisions

AGWA network • alliance4water.orgDevelopment banks and capacity-building groups.The World Bank, the Asian Development Bank, European Investment Bank, KfW, the Inter-American Development Bank, GiZ, the Cooperative Programme on Water and Climate.

Non-governmental OrganizationsConservation International, the Delta Alliance, International Water Association, the Swedish Environmental Institute (IVL), the Global Water Partnership, Deltares, Environmental Law Institute (ELI), Stockholm Environmental Institute (SEI), Organization for European Cooperation and Development (OECD), Stockholm International Water Institute, Wetlands International, IUCN, The Nature Conservancy, ICIMOD, WWF. GovernmentalUS Army Corps of Engineers, US State Department, NOAA, UN Water, UN Habitat, UNECE, Water Utilities Climate Alliance, WMO, CONAGUA, Seattle Public Utilities,

The Private SectorCeres, UNEP FI, World Business Council for Sustainable Development

Key partners

Water & Climate Coalition, the Adaptation Partnership, the Global Environment Facility, Nairobi Work Programme

Uncertainty

• Models not developed for adaptation purposes but for testing hypotheses about greenhouse gas mitigation.

• Low confidence, especially for quantitative purposes

• Little agreement across models, scenarios

• Often result in a series of “no regret” options

• Stakeholders often feel disempowered by process, which is often experienced as deterministic

Source: Wilby & Dessai, 2010, Weather

Traditional approaches amplify or hide uncertainty

Source: AGWA, “Caveat Adaptor,” 2013

UZH R. watershed [Dneister R. Basin](Zhelezhniak, et al. 2013)

0 200 400 600 800Discharge , cm s

0

20

40

60

80

100R

etu

rn P

eri

od

, yr

s

TO P K AP I (1961-1990)

D H S VM (1961-1990)

TO P K AP I (2011-2050)

D H S VM (2011-2050)

observed

Top-down vs. bottom-up approaches

top-down approaches to risk

assessmentdecision-scaling risk assessment

1. Define your system’s breaking points

2. Assemble multiple climate data sources and link to breaking points

3. Assess plausibility and test vulnerability

1. Downscale climate model projections

2. Estimate shifts in water supply

3. Determine system responses to changes in these variables

Weaver et al., 2012, WIREs Climate Change

Purpose

• Purpose of these sessions: Develop a risk assessment of the performance of water resources management under the threat of future climate changes and variability using a ‘bottom-up’ approach.

• A bottom-up approach is a stakeholder driven process to assess vulnerability rather than a reliance on predictive models of the future.

Establish Decision ContextEstablish Decision Context

Identify RisksIdentify Risks

Analyze RisksAnalyze Risks

Evaluate RisksEvaluate Risks

Risk MitigationRisk Mitigation Mon

itor,

Eval

uate

, Mod

ifyM

onito

r, Ev

alua

te, M

odify

Cons

ult,

Com

mun

icat

e an

d Co

llabo

rate

Cons

ult,

Com

mun

icat

e an

d Co

llabo

rate

Risk

Ass

essm

ent

Adapted from ISO 31000- Risk Management—Principles and Guidelines

Risk-Informed Decision Making

Background

• Risk management has two basic parts: assessing risks and developing solutions.

• Risks can be assessed either qualitatively or quantitatively

• Vulnerabilities such as flood inundation and flood damages can be quantified.

• Other vulnerabilities (ecosystems) can be categorized qualitatively by stakeholders in terms of ‘coping zones’ and relative degrees of ‘risk tolerance.’

• Risk management options (solutions) need to take into account both types of information.

Defining System Objectives

• For each sector (flood risk, ecosystems, and agriculture), what specific objectives are you trying to achieve?– Flood risk examples: reduce long-term flood damages;

reduce vulnerability of infrastructure to disruption; reduce human fatalities from flooding

– Ecosystem examples: improve biodiversity; preserve wetlands; increase fish stocks

– Agriculture examples: increase agricultural production; increase farm income

Measuring System Performance

• What metrics would you use to define success or failure?– Flood risk examples: reduction in flood damages– Ecosystem examples: biodiversity indicators; fish

biodiversity and catch amounts; health of indicator species– Agriculture examples: area of land irrigated; crop yields

• A metric is a measurable quantity that can be used to measure the performance of a system.

Identify Problems• Identify climate concerns, hazards and thresholds.

What river flow and climate conditions are associated with these hazards?– Flood risk: What are the current flooding problems in the Dniester

river basin? At what flood water levels and flood flow values are populations affected? At what flood water levels and flood flow values does major infrastructure become unusable?

– Ecosystems: What are the current major ecosystem problems? What is the major source/cause of ecosystem disruption (infrastructure, floods, droughts, or pollution)? What river flow values support these ecosystems? How has drought affected ecosystems? Have changes in flow patterns caused by reservoir regulation altered ecosystems?

– Agriculture: What are the current problems for irrigated agriculture? Discuss problems during past droughts.

Non-climate Causes of System Stress

• Identify key drivers and stressors. – Drivers are forces that can have major influences on the

system of interest. Potential drivers could be of physical, biological or economic origin (i.e., climate, invasive species, population growth, etc.).

– Stressors are changes that occur that are brought about by the drivers.

– Examples:• Flood: population living in floodplain; important infrastructure in

flood plain • Ecosystems: water quality (toxic chemicals, dissolved oxygen, water

temperature); overfishing; invasive species• Agriculture: irrigation infrastructure not performing as designed; soil

fertility

Risk Tolerance

• Risk tolerance is the willingness to bear a known risk based on its severity and likelihood

• What range of conditions would have unfavorable though not irreversible agricultural impacts? How often can you tolerate such conditions?– Flood examples: disruption of transportation; damaged homes; reduced

economic output– Ecosystem examples: loss of wetlands; diminished fish stocks– Agricultural examples: reduced farm income; lower agricultural

productivity• What range of conditions would have severe, long-lasting or

permanent adverse impacts?– Flood example: population does not return and rebuild after a flood– Ecosystem example: extinct species– Agricultural examples: farmland is abandoned

Coping Range: Coping range represents the magnitude or rate of disturbance various systems like communities,

enterprises, or ecosystems can tolerate without significant

adverse impacts or the crossing of critical thresholds.

Resilience Range: Resilience range is the magnitude of

damage a system can tolerate, and still autonomously return to

its original state. Failure Range: Failure range

starts from the point where magnitude of damage is such that a system can no longer

tolerate it without significant adverse impacts.

Risk Matrix

Likelihood

• Likelihood is the chance of something happening, whether defined, measured or determined objectively or subjectively, qualitatively or quantitatively.

• Statistical models are generally based on assumption of stationarity, that is the past is representative of future.

• Estimating probabilities for Global Climate Models is problematic; uncertainty is not quantifiable.

Confidence

• Confidence is the validity of a finding based on the type, amount, quality, and consistency of evidence (e.g., mechanistic understanding, theory, data, models, expert judgment) and on the degree of agreement (Intergovernmental Panel on climate Change Fifth Assessment Report, 2013).

• Our confidence in the likelihood and potential consequences will influence our decision.

Robustness Tests

• How well does the system remain functioning under a range of circumstances?

• System is tested with an array of approaches– Observed hydrology– Stochastic hydrologic sequences – Global climate model projections– ‘Weather generator’ if available

• Test risk management solutions across a range of possible conditions.

Stress test: drought characteristics

Severity = Volume/Duration

Duration, Severity and Intensity

Stress test: climate events

0

500

1000

1500

2000

2500

3000

3500

4000

1 13 25 37 49

Month #

Mon

thly

ave

rage

inflo

w (

cfs)

610

615

620

625

630

635

640

Lak

e le

vel (

ft.)

Historical

Climate 1

Historical - lake level

Climate 1 - lake level

Dniester Basin Flooding

Uncertainty and Flood Damage CalculationF

lood

Sta

ge (

S)

Flo

od S

tage

(S

)

Flood Discharge (Q)

Flood Discharge (Q)

Fre

quen

cy

Fre

quen

cy

Flood Damage (D)

Flood Damage (D)

Q

S

P

Q

P

D

S

D

UEB - Upper Error Bound

LEB - Lower Error Bound

UEB

LEB

U.S. Army Corps of Engineers Procedures - HEC-FDA;1992

Quantifying Frequency

Description Flood ZoneCoping Range

Frequency Range Mid-point estimated frequency

Approximate numericalvalue events/year

Rank

Remote Flood Zone 1 <1 : 200 years 1 in 500 yr 0.002 1

Rare Flood Zone 2 1:50 yrs - 1: 200 years

1 in 100 yr 0.01 2

Infrequent Flood Zone 3 1:10 years to 1: 50 years

1 in 20 yr 0.05 3

Occasional Flood Zone 4 1:5 yrs to 1:10 yrs 1 in 7 yr 0.15 4

Frequent Flood Zone 5 1:2 yrs to 1:5 yrs 1 in 4 yr 0.25 5

Regular Flood Zone 6 1:1 yr to 1:2 yrs 1 in 1.5 yr 0.50 6

Common Flood Zone 7 0.2 yr to 1:1 yr 1 in 0.3 yr 0.70 7

Quantifying Severity/ConsequencesEconomic/Safety/Health

Description

Approximate Numerical value

Equivalent fatalities per event

Ranking

Minor Damages < $ 103/ 0.005 1 (minor)

More serious damages e.g. multiple minor injuries

2 (significant)

Major injuries/property damage

3 (moderate)

Multiple Major / single fatality 4 (Major)

Multiple fatalities (1-10) 5

Severe economic damagesMultiple fatalities (10 to 100)

6 (Severe)

Catastrophic damagesMultiple fatalities (>100)

> $109/1000 fatalities 7 (catastrophic)

Choosing Management Alternatives

Dniester Dams and Reservoirs

Reservoir Regulation

• Potential adaptation measures– Provide more naturalized flow patterns for ecosystems

while maintaining economic benefits– Change allocation of storage space to different uses– “Dynamic rule curves”: Shift reservoir storage allocation

based on current hydrological conditions in basin.– More use of forecasts in reservoir operations

Reservoir Rule Curves and Storage Allocation

Less Precipitation

More Precipitation

Shasta Reservoir, California, USA

New Bullards Bar Reservoir, California, USAAllocation of Reservoir Storage Space