Integrating Biogeochemical and Price

Forecasting Models to Predict Bioenergy

Crop Supply and Environmental Impacts

S.M. Swinton, A. Egbendewe-Mondzozo, S. Kang, W.M. Post, W. Thompson & J.C. Binfield

Michigan State U., U. Lomé, Oak Ridge Nat. Lab & U. Missouri

In a nutshell

Research questions

What biomass crops will be grown in Great Lakes region?

What environmental consequences?

How do market feedbacks affect future trajectories?

Model elements: Biochemical & Multi-market

Model comparisons (biomass supply & environmt)

Fixed prices

Market price feedbacks

A. Egbendewe-Mondzozo et al.

2015, App. Econ Persp & Policy

How would bioenergy crops change economic &

environmental performance on Great Lakes cropland?

Research Questions

What conditions to supply energy crops?

What crops grown and which ones displaced?

Environmental consequences?

Effect of market feedbacks on predicted trajectories?

Sub-teams by model

Great Lakes Bioenergy Research Center (GLBRC)

Biophysical models (EPIC) at Oak Ridge & Pacific Northwest Labs (ORNL, PNNL)

Market price & quantity models at U. Missouri Food & Ag Policy Research Inst (FAPRI)

Land use bioeconomic decision model at Michigan State U.

www.glbrc.org

3

www.glbrc.org

4 Regional Intensive Modeling Areas in

Michigan & Wisconsin represent

agricultural land in southern Great Lakes

Spatial bioeconomic model:

Predicted biomass supply, land use, management choices & profit-environment tradeoffs

EPIC biophysical simulations

Crop yields & environmental

outcomes

Crop prices, production &

transport costs

FAPRI market & policy simulations;

Transport cost GIS

Spatial bioeconomic model of crop production:

Biomass supply & environmental trade-offs

PNNL & ORNL U. Missouri

Regional Intensive Modeling Areas in Michigan & Wisconsin (MSU)

EPIC simulates biophysical and biogeochemical processes as affected

by climate, soil, and management interactions

Developed by USDA and maintained and Texas A&M University

Key processes simulated

Weather: generated, historical, climate projections

Plant growth and yield

Crops, grasses, trees Complex rotations, intercropping, land

use change Radiation use efficiency Plant stresses

Water balance; irrigation, drainage

Heat balance; soil temperature

Carbon cycling, including eroded carbon

Nitrogen cycling

Erosion by wind and water

Plant environment control: tillage, fertilizers, irrigation, pesticides

Carbon emissions coefficients

Erosion

Operations

Pesticide fate

EPIC Model (Williams, 1995)

Precipitation

C, N, & P cycling

Plant

growth

Soil

layers

Solar radiation

Runoff

Wind

Residue CResidue C

Metabolic LitterMetabolic Litter Biomass C Passive C

Slow C Leached CStructural LitterStructural Litter

Residue CResidue C

Metabolic LitterMetabolic Litter Biomass CBiomass C Passive CPassive C

Slow CSlow C Leached CLeached CStructural LitterStructural Litter

Carbon Model in EPIC (Izaurralde et al., 2006)

(Slide: Izaurralde, 2009)

EPIC simulation of crop yield and

environmental outcomes in SW Mich & SC Wisc

82 cropping systems (70 land units in Mich.; 80 in Wisc.):

Crops

6 annuals: Corn, corn silage, soy, wheat, canola, & alfalfa

7 perennial: Switchgrass, miscanthus, poplar, 4 grass & prairie mixes

Tillage: no-till or chisel

Fertilization: high or medium

Residue removal: No or 50% www.glbrc.org

7

EPIC simulates crop yield & 5 environ. outcomes

Soil erosion

Soil carbon loss

Nitrate leaching to water

Phosphorus runoff to water

Greenhouse gas (GHG) flux (CO2 + N2O + methane)

www.glbrc.org

8

Simulated land use near central,

imagined biorefinery in each region

FAPRI multi-market supply-demand model

predicts U.S. ag prices & quantities

www.glbrc.org

9

Meyers et al. 2010, J Intl Ag Trade & Devt

FAPRI predicts rising relative price of

biomass compared to corn grain to 2022

www.glbrc.org

10

Profit maximization model simulates

choice of crop systems

Choose crops to maximize expected profit given parameters for:

Crop yield & selected ecosystem outcomes

Land quality by watershed

Prices & policy Math programming

GAMS modeling language

Calibrated to validly represent baseline 2007-9

www.glbrc.org

11

Egbendewe-Mondzozo et al,

Biomass & Bioenergy (2011)

Exogenous rise in biomass price: Annual biomass crops supplied first, then perennials

www.glbrc.org

12

- Corn@$3.5/bu=$164/t

- Biomass@$54/t Biomass@$20/t

Michigan RIMA

Environmental disservices rise with

annual crop biomass, fall with perennials

-3

-2

-1

0

1

2

3

4

5

0

0.0

2

0.0

4

0.0

6

0.0

8

0.1

0.1

2

0.1

3

0.1

5

0.1

7

0.1

9

0.2

1

0.2

3

0.2

4

0.2

6

0.2

8

0.3

0.3

2

0.3

4

0.3

5

0.3

7

0.3

9

0.4

1

0.4

3

0.4

5

0.4

6

0.4

8

0.5

0.5

2

0.5

4

0.5

6

0.5

7

0.5

9

0.6

1

Envi

ron

me

nta

l Dis

serv

ice

(pro

p’l

Δ)

Biomass:corn grain price ratio

Soil Erosion

Nitrate Loss

N2O Emissions

Phosphorus Loss

Soil Carbon Loss

Michigan RIMA

www.glbrc.org

13

Corn

Residue

supply starts

Switchgrass

Production

starts

Market-mediated biomass price rise compared to exogenous rise: Now food prices go up too,

so much less biomass supply

www.glbrc.org

14

Michigan RIMA

Corn stover still supplied,

but much less perennial

switchgrass

Markets dampen price effects, so predict worse

environmental impacts from bioenergy production

Exogenous biomass price rise

Faster bioenergy crop supply (food prices stay constant)

Intensified corn prod. gives way to perennial grasses

Environmental benefits rise with perennial grasses

Both states grow bioenergy perennial crops

Market-mediated price rise

Slower bioenergy crop supply (food prices adjust up)

Intensified corn remains, even if some perennials too

Environmental harm rises with intensive corn prod.

Michigan RIMA grows some perennial grasses (due to low corn yields) Wisconsin stays with corn & food crops

www.glbrc.org

15

“Hub & spokes” approach to collaborative

modeling

Land use decision model at MSU integrated parameters from:

EPIC multi-year runs provided both commercial & environmental outputs

FAPRI provided price forecasts

Evaluation

Good results: More realistic forecasts

But specialized roles, so …

• Full modeling team never assembled

• Lessons diffused from “hub” modelers

www.glbrc.org

16

Looking for future

modeling collaborators. . .