Proposal for a LandUse Model Intercomparison Project (LUMIP) for CMIP6 Summary Chairs: George C. HurD 1 and David M. Lawrence 2 SSG: Victor Brovkin, Nathalie de Noblet Ducoudre, Julia Pongratz, Kate Calvin, Elena Shevliakova, Chris Jones with input from many from Earth System Modeling, Integrated Assessment Modeling, and historical land use communiUes 1 Department of Geographical Sciences, University of Maryland 2 Climate and Global Dynamics Division, NCAR hDps://www2.cgd.ucar.edu/research/mips/lumip AGCI MeeUng Aspen August 4, 2014
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Proposal for a Land-‐Use Model Inter-‐comparison Project (LUMIP) for CMIP6-‐ Summary Chairs: George C. HurD1 and David M. Lawrence2
SSG: Victor Brovkin, Nathalie de Noblet Ducoudre, Julia Pongratz, Kate Calvin, Elena Shevliakova, Chris Jones
with input from many from Earth System Modeling, Integrated Assessment Modeling, and historical land use communiUes 1Department of Geographical Sciences, University of Maryland 2Climate and Global Dynamics Division, NCAR hDps://www2.cgd.ucar.edu/research/mips/lumip AGCI MeeUng Aspen August 4, 2014
LUMIP Timeline
• 2013 Summer: Concept • 2013 Fall: CMIP Proposal, WGCM Briefing • 2014 Spring: GLP MeeUng, Workshop 1 • 2014 July 17-‐18: GEWEX – Biogeophysics • 2014 July 21-‐22: Hamburg – Biogeochemistry • 2014 July 28-‐Aug 1: EMF Snowmass MeeUng • 2014 August 5-‐9: AGCI Aspen Joint-‐MIP Workshop • 2014 September 15: LUMIP proposal due • 2014-‐2017: DiagnosUcs, new scenarios, new data sets,
experimental design • 2015 GMD paper • 2018-‐2019: Model results and synthesis • 2020: WG1 AR6 Report published
Adapted from Meehl et al., EOS, 2014
Collection of coordinated activities to assess land role in climate and climate change
• Land Only simulations forced with obs historical climate (joint GSWP3, TRENDY, ISI-MIP protocol)
• Land Use = LUMIP land use forcing on climate, biogeophysics and biogeochemistry with policy relevance (LUCID)
• Carbon Cycle = C4MIP land biogeochemical feedbacks on climate change
• Land = LSMIP land systematic biases and biogeophys feedbacks including soil moisture and snow feedbacks
Land Terrestrial Processes in CMIP6
DECK incl atm only, land only, and ocean/sea-ice only runs
LUMIP Major Science QuesUons
• What are the effects of land use and land-‐use change on climate (past-‐future)?
• What are the effects of climate change on land-‐use and land-‐use change?
• Are there regional land management strategies with promise to help miUgate and adapt to climate change?
*AddiUonal detailed science quesUons to get at process level aDribuUon, uncertainty, data requirements, etc. *ParUcular focus on uncertainty, and separaUng effects of: fossil fuel vs. land use, biogeochemical vs biophysical, land cover vs land management.
LUMIP Major AcUviUes • Model metrics and diagnos5cs
– Develop set of metrics to assess/quanUfy model performance with respect to land use impacts on climate
– A diagnosUc protocol developed to quanUfy related model sensiUviUes
– Development of land use benchmarking data products for evaluaUon
• Data standardiza5on – Repeat and mature land use harmonizaUon processà enhanced land-‐use data set for CMIP6, passing maximum amount of common informaUon between relevant communiUes (Historical, IAMs, ESMs)
– Provide addiUonal required land management datasets – Data output standardizaUon, new variables
• Model experiments – Development of efficient model experiments designed to isolate and quanUfy land use and land management effects
Model Metrics and DiagnosUcs (Dral)
• Primary variables: net radiaUon, evapotranspiraUon, temperature, precipitaUon, and land carbon stocks, land nitrogen stocks, and fluxes, including runoff-‐ sufficient to close energy, carbon, nitrogen, water budgets
• Protocol: paired simulaUons w/wo factor, online and/or offline, range of spaUal and temporal scales and domains, ensemble members
• Leverage exisUng datasets for evaluaUon from mulUple ongoing landmips, supplement as needed
• Development of global benchmark maps for all forcing case
• Development of paired-‐sites data sets for land-‐use factor experiments
• ILAMB+LU extension
Data StandardizaUon (Dral) • Updated land-‐use history
– Pasture anomaly correcUon, new enhanced historical reconstrucUon, Landsat constraint
• New future scenarios – Idealized, Realis,c
• New land-‐use AND land-‐cover harmonizaUons with Mgt – Land-‐use transiUons, – F/NF gross transiUons, PFT land cover transi,ons – Harvest, Fer,lizer, Irriga,on, Crop type, Biofuel
• StandardizaUon of data usage – more informaUon, clear arUculaUon of best pracUces, straUfied comparisons
Overall Approach: Two track design: 1) idealized; 2) realisUc simulaUons Tiered prioriUzaUon of experiments
Track 1 (Start now) Idealized model experiments designed to: • Improve process understanding/assessment of how models represent impact of
changes in land state on climate; • QuanUfy model sensiUvity to potenUal land cover and land management
changes. Land cover/land management factors manipulated in simple standard fashion.
Track 2 RealisUc model experiments designed to: • Isolate the role of land cover/use change on climate relaUve to other forcings
Dral experimental design (Track 1)
Process understanding
Idealized experiments designed to assess biogeophysical role of land cover change on climate
CPL_1%DF Idealized 1% or 2% per year deforestation, once global deforest, continue run for 50 to 100 years (Tier 1)
1850-????
LND_DF, ATM_DF, CPL_DF
Land, atm, cpl simulations with some set of tropical, boreal, or temperate deforestation (defined by LUC4C/LUCID?) (Tier 3)
1980-2010
!
Dral experimental design (Track 1)
Land cover versus land management change (Tier 2)
Assess relative impact of land cover and incrementally more comprehensive land management change on land to atmosphere fluxes of water, energy, and carbon; forced with historical observed climate and projected climate anomalies
LND_control No land use change: Offline LND with transient CO2, N-dep, aerosol dep but land cover held at 1850 distribution
1850 – 2010 or 1850 – 2100
LND_grasscrop LND_control but w/ land use change with ‘grassland’ crop/pasture
LND_woodharv LND_control with wood harvest turned on
LND_pasture LND_grasscrop but with grazing ???
LND_crop Land use change with crop area utilizing prognostic crop model
LND_crop-irrig Land use change with crop model and realistic transient irrigated area
LND_crop-irrig-fert
Land use change with crop model + irrigation and realistic transient fertilization
!
Dral Experimental Design (Track 2)
Land use change impact on land to atmosphere fluxes of water, energy, carbon (Tier 1)
LND_allforce
Offline LND with crop, irrigation, fertilization schemes active with transient land cover and land management and CO2, N-dep, and aerosol dep forced with historical observed climate (LMIP)
1850-2012, 2013-2100?
LND_noLULCC Same as LND_allforce except with land cover held constant at 1850, no human impact
Land use change impact on past and future climate (Tier 1)
CPL_allforce All forcing simulation (DECK, ScenarioMIP) 1850-2100
CPL_noLULCC_hist
Same as ESM_allforce except with land cover/use held constant at 1850, concentration (for DA) and emission driven, no human impact
1850-2010 (3 ens for conc runs)
CPL_landpolicy Additional land mitigation policy scenarios for a particular RF scenario, keep all GHG the same, only change land use; emissions driven runs if possible
2014-2100 (# ens?)
CPL_noLULCC_fut
Future simulation with land cover/use held constant at 2014 levels; emissions driven runs if possible
2014-2100 (# ens?)
!
/2100?
Dral Experimental Design (Track 2)
Effects of climate change on land use and land use change
iESM ???
!
Topics for Discussion • What are the most important scenarios to study in LUMIP? (e.g. High/
low climate x High/low land-‐use?) • What are the largest policy relevant land-‐use changes contemplated? • What are the most important connecUons to capture between land-‐use
and atm chemistry, and how? • What is the most important informaUon for historical record and IAMs
to pass to ESMs in support scenarios? (e.g. Land cover change, Biofuels/CCS, Ag. Mgt?)
• To what extent is regional climate emphasized in CMIP6? • When is iniUal year, and is there an aDempt at harmonizaUon in that
year, what variables, what resoluUon? • Can we design and execute an effecUve land-‐use coupling experiment?
Is there criUcal mass for doing? • How can we improve workflow/informaUon flow between History/Obs,
ESM, IAM? • Details, details, details…. • ….
PARKING LOT
IAM-‐LUH-‐ESM INFO EXCHG
CMIP5 • Crop area • Pasture area • Wood harvest carbon • Urban area* • Biofuel area*
CMIP6? • Crop area • Pasture area • Wood harvest carbon • Urban area* • Biofuel area* • Land cover F/NF • Land cover PFT • Fer,lizer amt/t • Irriga,on amt/t • Transi,ons? • Narra,ve?
Land Experiments – PrioriUzaUon/CoordinaUon (DRAFT)
What we learned (CMIP5+)?
• Enabled first global model emission driven projecUons of both CO2 and climate including effects of spaUal land-‐use changes
• Land-‐use effects on global climate are generally modest relaUve to FF, but sUll important
• Land-‐use transiUons are needed for accurately tracking land cover change resulUng from land-‐use change
• Land-‐use effects are complex and challenging to diagnose • Different models implemented standardized land-‐use data sets
differently • PotenUally important impacts, management pracUces, biophysical
effects, policy opUons, uncertainUes, and feedbacks not adequately accounted for in current design
• SubstanUal opportuniUes exist to build on CMIP5 approach and improve data and models for CMIP6
17
PrioriUes for CMIP6 (Land Use)
1. Repeat and mature the LUH process (more data, more terms, increased resoluUon, longer period, beDer communicaUon)
2. Work to standardize products, and usage of products 3. Focus: links between LU change, LC change, C fluxes, Biophys. 4. New emphasis: LU management, policy relevance, uncertainty 5. New scenarios: Esp. SSPs and with added mulU-‐objecUve
consideraUons 6. Expand RCP-‐RF definiUon to include biophysical 7. Joint harmonizaUon of LU emissions and LU changes 8. Diagnose ESMs, IAMs, and IAVs to quanUfy effecUve data
Area of forested land in shiling culUvaUon fallow (2000) 4.42 x 106 km2 (FAO) 4.56-‐6.19 x 106 km2 3.7 x 106 km2
Rates of clearing land in shiling culUvaUon 0.6-‐0.09 x 106 km2/yr
(Rojstaczer et al. 2001) 0.48-‐0.65 x 106 km2/
yr 0.58 x 106 km2/yr
Percentage of US Forests that are secondary (2000) 94-‐99% 100%
Mean age of Eastern US Secondary Forests (2000) 38yrs * 71 yrs 63 yrs
Total gross transiUons (2000) 1.6 x 106 km2/yr 2.0 x 106 km2/yr
Total net transiUons (2000) 0.17 x 106 km2/yr 0.19 x 106 km2/yr
Global Cropland Area (1990) 12.1 x 106 km2 15.1x 106 km2
Global Pasture Area (1990) 25.8 x 106 km2 33.1 x 106 km2
Global Primary Land Area (1990) 57.7 x 106 km2 58.4 x 106 km2
HurD et al. (2009, 2011)
23
pasture
IMAGE
AIM
MESSAGE
crop
ΔF (IAM)
ΔF
(LU
H)
24
25
LUH-AIM (6)
LUH-MESSAGE (8.5) LUH-GCAM (4.5)
LUH-IMAGE (2.6)
HurF et al. 2011
HurD et al. (2011) 26
Wood harvest
Gross T. Net T.
Sec. Sec. Age
Total C Net C
Shiling CulUvaUon
Start Date Future Scenario
27 Brovkin et al (2013)
28 Brovkin et al (2013)
29 Brovkin et al (2013)
Simulated Atmospheric CO2
Courtesy R. Stouffer Courtesy P. Thornton
Courtesy L. Chini
31
Land Use = Land Cover • Land use is the human use of the land. • Land cover is the physical material at the surface of the earth.
• Important to keep these separate, to drive process-‐based models, and aDribute results.
uncertainty
Ume
With Remote Sensing
Hansen et al 2010
!" #" $" %" &"
!"
#"
$"
%"
&"
'" ('"
'"
('"
ARS ALUH
33
New Consistent Land-‐cover HarmonizaUon
LUH2 Land-‐cover Classes (Proposed)
• Based on widely used classificaUon (IGBP) • ENL, EBL, DNL, DBL, and mixed forests • Closed and open Shrublands, savanna • Grassland, pasture • Urban • Croplands
• Add important crop funcUonal types (CFTs)
• C4 • C3 perennial • C3 annual • N fixers • Rice
• Align with IAMs and ESMs • Advance implementaUon
Leff et al. (2004) Global maps of 18 major crops at 5-‐minute resoluUon (e.g., wheat→) based on cropland map of RamankuDy & Foley (1998) and mid-‐1990s agricultural census data.
Leff et al (2004)
What is other? FAO: 157 other ‘major’ crops
How might these be aggregated to simplify & generalize?
C3 annual C3 N-fixi n g C3 perenni a l C4 annual units
total biomass 5 0 0 0 3 8 0 0 8 4 0 0 16000 kg C/ha Grain % 3 9 2 9 4 1 2 7 % Shoot % 4 3 4 6 4 0 3 9 % Grain C:N 2 3 1 4 3 9 9 1 -- Shoot C:N 5 9 3 4 4 2 1 1 0 -- Root C: N 4 6 3 3 4 5 1 1 0 -- Max. height 0 . 7 0 . 5 3 . 2 1 . 7 m Max root dept h 1 . 0 1 . 5 1 . 3 1 . 0 m Max. LAI 4 . 0 3 . 0 4 . 2 4 . 7 m2/m2 Water requi r e d 2 5 0 2 2 0 2 6 0 1 6 0 kg H2O/kg C Heat requi r e d 1 8 0 0 2 3 0 0 - - 2 9 0 0 °C-days Max. psn rate 4 2 3 8 4 9 6 3 kg CO2/ha/h
1. We acquired crop physiological parameters from the DNDC model for major crops (thirteen C3-‐annuals, three C3 N-‐fixers, three C3-‐perennials, three C4-‐annuals). 2. We calculated area-‐weighted averages to get mean global crop parameters.
Basic global mean crop parameters
Developing global mean crop-‐type physiological parameterizaUons
37
Importance of Management Effects
Mueller et al. 2014
NDVI trends 1981-‐2010
Global irrigated area 1900-‐2000 (Freydank & Siebert 2008) and global N fer,lizer use 1900-‐2010 (Smil 2001; IFA 2014).