MERGE Activities relating to 1.5ºC · BECC-MERGE Seminar, 26 Nov 2018 ... •LPJ-GUESS developments will be available to RCA-GUESS and EC-Earth Ongoing model improvements . Including

BECC-MERGE Seminar, 26 Nov 2018

MERGE Activities relating to 1.5ºC

Paul Miller1,2

(With thanks to many MERGE & BECC colleagues)

1 Dept. of Physical Geography and Ecosystem Science (INES)

2 Centre for Environmental & Climate Research (CEC)

• MERGE – overview

• LPJ-GUESS - continuously evolving for better impact studies

• Capturing often ignored (biophysical) feedbacks in low

emission scenarios

• MERGE & the EC-Earth Earth System Model – making unique

climate projections in CMIP6

• Towards better estimates of the remaining carbon budget in

line with a 1.5ºC target

Outline

MERGE for GAC 19-20 May 2010

• To advance the state-of-the-art for

representing biosphere-atmosphere

forcing and feedbacks in global and

regional Earth system (climate)

models

• To contribute to national and

international efforts to describe and

attribute climate change,

underpinning policy responses

• To educate a new generation of

young ESM model experts

• To support the ClimBEco graduate

school - www.climbeco.lu.se

• To improve the societal relevance of

climate models and their results

MERGE, a SRA that aims…

www.merge.lu.se

Target Group

PhD students engaged in research related to

biodiversity, ecosystem services and climate

Vision and Mission • To provide perspectives from natural science, social science and economics

• To improve student awareness of the discipline plurality needed in society to

drive sustainable (local, regional and global) development in the field of

environment and climate.

• To develop interdisciplinary training and give societal context, as well as to

promote improved communication skills, personal- and career development,

and in-depth intra-disciplinary understanding.

• To create and promote a network of peers and mentors. Contact: Anna Ekberg [email protected]

RA1 – ESM

development

climate-biosphere

interactions RA2 – learning from

the past

palaeoclimate & land

cover

RA4 – statistics

new methods for ESM

analysis and simulation

RA3 – particles

aerosol-cloud

parameterisations for

ESMs

Addressing the Research Challenges

The Four Linked Research Areas (RA) in MERGE

LPJ-GUESS DGVM

RCA-GUESS RCM

EC-Earth ESM

WRF RCM

LPJ-GUESS

Vegetation and land-use

reconstruction using REVEALS

Solar activity reconstructions

COSMO-wiso

LPJ-GUESS

Laboratory and field studies

EC-Earth (BVOC & SOA, TM5)

EMEP

ADCHEM

Landscape reconstruction using

Gaussian Markov Random Fields

Inverse modelling of CO2

Coupled Model Intercomparison Project (CMIP)

&

Representative Concentration Pathways (RCP)

IPCC AR5 Summary for

Policymakers (SPM)

RCP2.6

consistent with

2 ºC

Impact studies and biophysical feedbacks

• link structure to function, accounting for

feedbacks between them

• link ’fast’ (physiology, biogeochemistry)

and ’slow’ (demography, composition)

ecosystem processes

• account for transient ecosystem

dynamics when driving conditions

(climate, CO2) are changing rapidly

StructuresFunctions

space

time

Simulated natural

vegetation of Lithuania

LPJ-GUESS – global DGVM & ecosystem model*

Soil organic matter

Soil organic

matter & N

soil bio-

geochemistry

population dynamics

& disturbance

plant

biogeography

primary

production

& growth

Vegetation

PFTs / species

& Crops

Climate (historical & scenario)

land use/management

CO2

N-deposition

*e.g. Smith et al. 2001, 2014

Biome shifts are expected in the 21st century (RCP 8.5)

*Wårlind et al. 2014

Cover crop

Irrigation

Yearly cutting

Succession of abandoned farmland

PASTURE CROPLAND

NATURAL MGD. FOREST

Crop rotations

N limitation Land cover/land use

change (gross vs. net)

Continuous forestry

Daily grazing

Detailed forest management

Land use and management now treated in LPJ-GUESS*

*Lindeskog et al. 2013. Earth System Dynamics 4: 385-407

Olin et al. 2015. Biogeosciences 12: 2489-2515

*Olin et al. 2015. Earth System

Dynamics 6: 745-768

Projections of agricultural crop yields*

But! Offline studies miss feedbacks to climate via

changed land-atmosphere energy balance

Albedo (reflected sunlight) differs between forest and open land, especially during period of snow lie

LH

SH LH

SH

incoming

shortwave

radiation

outgoing

heat

Greater evaporative surface area of tall, dense vegetation enhances latent heat flux, reducing surface warming (Often deeper roots and greater roughness lengths too)

Slide: Ben Smith

Zhu et al. (2016), Nature Climate Change

Change in land surface greenness

1982-2009

Vegetation cover and productivity have increased

in concert with recent decades’ climate warming (< 1.5 ºC) and

CO2 increases

Recent greening trends have mitigated

background climate warming, mainly due

to increased evapotransipration

Trend attribution from models:

• CO2 fertilization (increased NPP)

• Climate change

• Nitrogen deposition

• Land cover change

High-latitude land cover patterns are already changing

due to impacts of climate (< 1.5 ºC) on vegetation

1912 2009

Treeline advance, Mt Nuolja, Sweden (Van Bogaert et al. 2011. J. Biogeog.) Shrubification of Alaska

Tape et al. 2006 Global Change Biology 12: 686

Soil organic

matter

Soil organic

matter

soil C dynamics

population

dynamics

& disturbance

migrationphenology

& growth

Vegetation

climateCO2CO2

Rossby Centre

Atmospheric Model

RCA3

temperature

radiation

soil water

RCA-GUESS: a regional Earth system model*

leaf area index annual

daily

fractional cover

- broadleaved forest

- needleleaved forest

- open land vegetation

daily

LPJ-GUESS v3 dynamic vegetation model

*Smith et al. 2011

Tellus 63A: 87-106

CO2

CH4

climate at

domain

boundaries

CO2

CH4

open loop for

biogeochemical

feedback

+ Multiple published studies for CORDEX Europe, Africa, Arctic, S. America

Projected vegetation change in the Arctic* coupled simulation (veg biophysical feedbacks included)

grass – tundra

Birch

Larch

Pine

Spruce

Forest

PFTs

100 yrs →

rela

tive c

over

(% L

AI)

LPJ-GUESS

RCA4

RCP

forcing

EC-EARTH

*W. Zhang et al.

GRL, 2018

*W. Zhang et al. GRL 2018

Projected feedbacks to surface air temperature*

LPJ-GUESS

RCA4

RCP

forcing

EC-EARTH

Additional temperature change due to vegetation feedback

(2071-2100)(1961-1990)

• Seasonality shift – longer growing season, earlier temperature peak

• Evaporative cooling evens out growing season temperature profile

Arctic vegetation feedbacks show the potential for

dynamic vegetation change to alter regional climate

• Seasonality shift – longer growing season, earlier temperature peak

• Evaporative cooling evens out growing season temperature profile

→ favours further shrub encroachment and treeline advance

→ Enhances the terrestrial C sink (Zhang et al. 2014)

-3

-2

-1

0

1

2

3

J F M A M J J A S O N D

(b)Tem

pera

ture

( C

)∆T ( C

) albedo feedback

evapotranspiration feedback

Additional temperature change

(2071-2100)(1961-1990)

*W. Zhang et al.

GRL 2018.

max

+1

mean among yrs

1

min

RCP 2.6

RCP 4.5

RCP 8.5

LPJ-GUESS

RCA4

RCP

forcing

EC-EARTH

MERGE & the EC-Earth Earth System Model –

making unique climate projections in CMIP6

No CMIP5 model included both dynamic vegetation and

carbon-nitrogen interactions

Ur IPCC 5:e Syntes “Summary for Policymakers”

RCP2.6

consistent with

2 ºC

LPJ-GUESS

Version 4

(vegetation dynamics) HTESSEL

(land surface)

NEMO

/LIM3

(ocean, sea ice)

Software

Coupler

LAI, high &

low vegetation

Vegetation cover

& types

Crops and pasture

temperature

radiation

Precipitation

Soil state IFS

(atmosphere)

Information

- Dynamic vegetation and terrestrial C & N

cycling in CMIP6

EC-Earth-Veg

CO2

CMIP6

External Forcing &

Boundary Conditions:

Land use

CO2 concentrations

N deposition

runoff Update vegetation and land

surface properties

CMIP6 Experimental Design

*Ciais et al. 2013, IPCC AR5

EC-Earth-Veg

MERGE commitments to CMIP6

EC-Earth DECK experiments to begin this month

fro

m E

yri

ng

et

al.

2015

C4MIP

LUMIP

& LS3MIP

PMIP4

ScenarioMIP

Riahi et al.,

Global LULCC scenarios Today (Mha)

• Forest: 4200

• Cropland: 1650

• Pasture: 3300

ScenarioMIP runs with EC-Earth-Veg

O’Neill et al, 2016

Expected to achieve 1.5 / 2 ºC

Towards better estimates of the remaining

carbon budgets in line with

ambitious temperature targets

*Ciais et al. 2013.

IPCC-AR5, Ben Smith

Glo

bal C

flu

x (

Gt C

yr

1)

Approx. 25% of C emissions Are taken up by terrestrial ecosystems due to surplus of photosynthesis relative to respiration and emissions from wildfires

The land and oceans currently take up about

50% of mankind’s annual CO2 emissions

Emissions: 400+/-20 GtC

Land use: 145+/-50 GtC

Atm: 230+/-5 GtC

Ocean: 155+/-20 GtC

Land: 160+/-60 GtC

Headline IPCC AR5 result: Temperature responds

nearly linearly to cumulative emissions (”TCRE”)

IPCC AR5 WG1 SPM

Policy relevant

uncertainty in

”allowable emissions”

2011-2100 CO2 emissions must be limited to meet

1.5 °C and 2 °C targets

Schellnhuber et al., 2016, IPCC 2018

Emitted by 2017: 2200 +/- 320 GtCO2

Annual emissions: 42 GtCO2

Remaining 1.5 ºC budgets:

50 % chance: 580 GtCO2

66 % chance: 420 GtCO2

CMIP6 Experimental Design

*Ciais et al. 2013, IPCC AR5

EC-Earth-Veg EC-Earth-CC

LPJ-GUESS

Version 4

(vegetation &

biogeochemistry)

HTESSEL

(land surface)

NEMO/PISCES

/LIM3

(ocean, sea ice,

BGC)

Software

Coupler

LAI, high &

low vegetation

Vegetation cover

& types

Crops and pasture

temperature

radiation

Precipitation

Soil state

NEE/CO2

[CO2]

IFS

(atmosphere)

TM5

(chemistry,

transport)

Information

Carbon Cycle

Global climate-carbon dynamics

and feedbacks

EC-Earth-CC

CO2

CMIP6

External Forcing &

Boundary Conditions:

Land use

CO2 emissions

N deposition

runoff

Carbon-climate

feedbacks

Update vegetation and land

surface properties

• Land Use Change (LUC) emissions were treated differently in the models

• Some AR5 models had dynamic vegetation

• Some AR5 models had nutrient (N) limitations on plant growth

• No AR5 model included dynamic vegetation AND C-N interactions

• More agreement as to the sign and approximate size of the ocean sink

Friedlingstein et al. 2014

J. Climate

LUC emissions prescribed

LUC emissions calculated

With N limitation on plant growth

An uncertain future for the terrestrial carbon sink 11 CMIP5 ESMs (RCP 8.5 CO2 emissions)

OCEAN

Land acts as a C sink

Realistic terrestrial C pools and fluxes EC-Earth-Veg Fully coupled simulation 1870-2015

LPJ-GUESS

Obs SST

& Sea ice

IFS

LUH2 &

N deposition D. Wårlind & L. Nieradzik (Lund Univ.)

NPP (GtC / yr)

Vegetation C (GtC)

Soil & Litter C (GtC)

• Most scenarios that meet the 1.5 C target include substantial

negative emission technologies (NET, incl. BECCS) and

agriculture and forestry/land use measures (AFOLU).

• EC-Earth-CC can be used to explore alternative pathways

The difficult paths to 2 ºC, or 1.5 ºC

IPCC SPM 2018

• Necessary BECCS and/or forestry practices (AR) will require the

use of substantial land area

• And result in as-yet unknown local climate feedbacks and other

conflicts

IPCC (2018): BECCS 0-16 GtCO2 / yr. AFOLU: 1-5 GtCO2 / yr

Mitigation conflicts

Smith et al. 2016

IDAG (Mha)

• Skog: 4200

• Odlad mark: 1650

• Bete: 3300

Today

Ca 10 GtC / yr

Lowe & Bernie 2018

Phil Trans. R. Soc. A

Biosphere-controlled non-CO2 atmospheric trace gases and aerosols

may generate significant feedbacks – sign and magnitude are uncertain

Consideration of biosphere-controlled non-CO2 atmospheric

trace gases and aerosols could reduce the remaining C budget still further

Lowe & Bernie 2018

Phil Trans. R. Soc. A

IPCC 2018 SPM:

50% Reduced budget

due to extra feedbacks

• Field studies, observations and modelling highlight the importance of

including detailed biosphere-atmosphere interactions in coupled

modelling frameworks

• So a modeller’s work is never done! We will soon include:

• Permafrost-carbon interactions and wetland CH4 emissions (LPJG

v4.1)

• Phosphorous limitation in addition to N important in the tropics

• Improved wildfire parameterizations (with BLAZE) (LPJG v4.1)

• Plant functional types (PFTs) must be added that can better quantify

the potential for bioenergy carbon capture and storage (BECCS) to

mitigate climate change

• O3 damage, N2O, microbial processes, allocation improvements etc.

• LPJ-GUESS developments will be available to RCA-GUESS and

EC-Earth

Ongoing model improvements

Including permafrost

Observed

(Stores > 1600 GtC) LPJ-GUESS (1961-90)

LPJ-GUESS (1901-2014)

Figures:

Adrian Gustafson

Including methane

LPJ-GUESS (1901-2017)

g CH4 m-2 yr-1

Global CH4 emssions (2000-2017)

Figures:

Wenxin Zhang

MERGE for GAC 19-20 May 2010

Vegetation can also have a cooling effect on

climate through its influence on cloud

properties

Figure: Erik Swietlicki (Lund U)

Global temperature

Aerosol Optical Depth

Biota (Vegetation)

CCN

Biogenic VOC

Atmospheric particle

concentrations

Clouds

LPJ-GUESS BVOC emissions

LPJ-GUESS

Version 4.1

(vegetation &

BGC)

HTESSEL

(land surface)

NEMO/PISCES

/LIM3

(ocean, sea ice,

BGC)

Software

Coupler

LAI, high &

low vegetation

tile fraction

& types

Crops and pasture

temperature

radiation

Precipitation

Soil state

CH4, N2O, BVOC, NEE/CO2

[CO2]

IFS

(atmosphere)

TM5

(SOA, chemistry,

transport)

Implemented

ESM fluxes

ESM in H2020

CO2

External Forcing &

Boundary Conditions:

Land use

CO2 emissions

N deposition

runoff

Closed loop for

biogeochemical

feedbacks

Update albedo, roughness lengths,

root distribution

www.crescendoproject.eu

Goal – improved budget understanding

IPCC SPM 2018

MERGE for GAC 19-20 May 2010

Tack!