1 of 18

Ecosystem Service-Integrated Assessment Modeling (ES-IAM):

Challenges & research frontiers for cross-scale modeling & data integration

Becky Chaplin-Kramer, Lead Scientist, Natural Capital Project

University of Minnesota and Stanford University

@beckyck rchaplin@umn.edu bchaplin@stanford.edu

2 of 18

3 of 18

1

2

3

4

5

4 of 18

1

5 of 18

24%

39%

Most important sites (5% of NCP)

Importance to country-level prioritization (across 12 types of NCP)

Critical natural assets

Sites providing 90% of NCP

Natural assets

On land

In the ocean

 DOI: 10.1101/2020.11.08.361014

Nitrogen retention

Sediment retention

Crop pollination

Fodder for livestock

Timber production

Fuelwood production

Flood regulation

Riverine fish provision

Marine fish provision

Access to terrestrial nature

Marine recreation

Coastal risk reduction (terrestrial and marine)

How much (and where) nature is needed to support human well-being?

39% of natural assets on land provide 90% of value to people

6 of 18

Johnson et al. 2020 https://www.wwf.org.uk/globalfutures

Spatial Economic Allocation Landscape Simulator (SEALS)

How (through what policy instruments) can we reach desired futures?

How are global economic drivers’ expected impacts on regional land-use allocations likely to play out in local spatial patterns of land-use change?

  • Downscaling coarse changes to finer spatial scales necessary for ecosystem service modeling
  • Representing uncertainty of those changes, bounds of realistic futures

7 of 18

1

8 of 18

1

2

9 of 18

max

min

Number of people downstream

What we’ve got:

What we want:

Cost of dredging, reduced reservoir capacity

But where are all the dams, globally?

High resolution imagery and machine-learning to fill gaps in global databases

E.g., for sediment retention

10 of 18

1

2

11 of 18

1

2

3

12 of 18

max

min

What we’ve got:

Sediment retained

Retention set by LULC class

What we want:

Retention set by ecosystem function, reflecting long-term feedbacks in ecosystem condition

13 of 18

Percent difference from observed

Categorical (LULC)

Continuous (EVI)

Error: 11.5 t/ha

Error: 5.8 t/ha

Improvement in accuracy of sediment modeling using indices of ecosystem function instead of land-use

Percent difference from observed

14 of 18

1

2

3

15 of 18

1

2

4

3

16 of 18

0

+100%

+200% N export

-100%

200

Regional Rivalry

(SSP3)

Sustainability

(SSP1)

Fossil Fueled

(SSP5)

million people

100

Chaplin-Kramer et al. 2019 DOI: 10.1126/science.aaw3372

SSP5 (2050)

Oceania

North Asia

Africa

Eurasia

South America

North

America

Change in…

Water Pollution

Coastal Risk

Lost Crop Production

South Asia

1200

400

400

200

400

200

200

200

200

30

Millions of people negatively impacted:

6

6

11

21

4

22

Number of People Impacted (Millions)

940

2,491

538

306

337

301

460

2,238

1,413

401

1,176

586

22

27

22

657

1,470

1,015

124

280

118

13

11

15

189

484

410

17

108

29

27

25

29

14

263

165

30

84

33

18

21

18

68

430

322

151

301

246

151

136

154

48

86

397

4

4

3

Sustainability

Regional Rivalry

Fossil-Fueled Development

Scenarios in 2050

Billions in South Asia and Africa facing increased risk while conditions improve in North Asia and North America

17 of 18

Chi et al. In review. http://beta.povertymaps.net/

Within-country equity mapping: what are the distributions of providers and beneficiaries of ecosystem services?

18 of 18

1

2

3

4

5