Miao Yu1,2, Guiling Wang1*, Dana Parr1, Kazi Farzan Ahmed1

Future changes of the terrestrial ecosystem based on a dynamic vegetation model driven with RCP8.5climate projections from 19 GCMs

Supplementary Materials

Miao Yu1,2, Guiling Wang1*, Dana Parr1, Kazi Farzan Ahmed1

1. Department of Civil & Environmental Engineering, University of Connecticut, Storrs, CT 06269

2. KeyLaboratory of Meteorological Disaster, Ministry of Education, Nanjing University of Information Science and Technology, Nanjing 210044, China

*Corresponding contact:

860 486-5648

Impact ofnew modifications on model results:

To illustrate the impact of the modifications on water-controlled phenology scheme, two 20-year sensitivity experiments were conducted: one using the default version of CLM4-CN model and the other including the refinement of the water-controlled phenology scheme for tropical broadleaf drought-deciduous trees. In both simulations, vegetation distribution is prescribed, so20 years of simulation is sufficient. Using one grid point as an example where vegetation is dominated by tropical broadleaf drought-deciduous trees, Fig. S1 shows the simulated monthly ELAI from these two experiments and the monthly precipitation from the meteorological forcing. The default version of CLM4-CN model produces a spurious increase of ELAI in the middle of the dry season. This error is corrected by our refinement of the water-controlled phenology scheme. The drought-incurred senescenceis captured reasonable well.

To illustrate the impact of all the modifications on the simulated vegetation distribution, four sensitivity experiments were conducted, each using a different version of the CLM4-CN-DV model. Results from these experiments are presented in Figs. S2 and S3using tropical broadleaf evergreen and drought-deciduous trees as examples. The default version of the model produces too much coverage of broadleaf evergreen trees in the southeastern part of Amazon and in Central Africa (Figs. S2a and S3a), due to the well-documented overestimation of GPP (reference – should be Bonan’s paper on the new GPP parameterization);The GPP parameterization from CLM4.5 substantially reduces the model-simulated tree coverage in these regions (Figs. S2b and S3b).The refinement of the water-controlled phenology scheme caused slight expansion of the tree coverage, especially for drought deciduous trees (Figs. S2c and S3c); adding a threshold of soil drought season length above which evergreen trees cannot survive increased areaswhere the simulated vegetation is dominated by drought deciduous trees, especially along the northernand southern borders of the forest in Africa (Figs. S2d and S3d).

Simulated Present Day Vegetation Distributions:

Needleleaf evergreen trees are mainly located in the Northern Hemisphere with the maximum coverage between 40ºN and 70ºN (Fig. S4a).This is captured very well by the CLM-CN-DV control run (Fig. S4b). However, the model overestimates the density of needleleaf evergreen trees in Southeast Asia. The model also simulates some coverage of needleleaf evergreen trees in the Southern Hemisphere (e.g., in western and southeastern South America and in Central Africa) that are not observed from the MODIS data. Gotangco Castillo et al. (2012) suggested that the simulated distribution of needleleaf evergreen trees is more reasonable than MODIS because MODIS dataset confines this type of vegetation in Northern Hemisphere. Most of the GCM-driven present-day simulations produce similar results with the control run except for IPSL-CM5A-MR and MIROC5. A spurious presence of needleleaf evergreen trees are produced in Central Africa by the simulations driven with output from BNU-ESM, CCSM, FGOALS-g2, the three GFDL models (GFDL-CM3, GFDL-ESM2G and GFDL-ESM2M) and INM-CM4, and in the Amazon by the simulation driven with the INM-CM4 climate (Fig. S4c-u), probably as a result of cold bias in these GCMs for the regions involved.

Broadleaf evergreen trees are mainly distributed throughoutthe Tropics, including the Amazon, Central Africa and the maritime subcontinents (between South Asia and Australia) (Fig. S5a). This pattern can be reproduced by the control run, but the simulated fractional coverage is smaller, with the values less than 50% (Fig. S5b). This is also the case for simulations forced by GCMs output, which results from a deficiency of CLM-CN-DV. Specifically, the model simulates a half-half split of vegetation between evergreen and drought-deciduous trees in the wet Tropics, where MODIS indicates dominance by evergreen trees. While all the GCM-driven runs show coverage of broadleaf evergreen trees in the maritime subcontinents, not all the experiments produce coverage inboth the Amazon and Central Africa. There is much less coverage of broadleaf evergreen trees in Africa under the climate of CMCC-CM, HadGEM2-ES and MRI-CGCM3, and much less in the Amazon under the GFDL-family models’ climates, the BCC-CSM1.1m, IPSL-CM5A-LR and MIROC5 climates. Simulations driven by CNRM-CM5 and IPSL-CM5A-MR climates produce little or no coverage of broadleaf evergreen trees both in Amazon and in Central Africa.

MODIS shows broadleaf deciduous trees mainly in eastern North America, Europe, Southeast Asia, mid-latitude Eurasia, eastern Amazon, Sahel and areas south of the Congo Basin (Fig. S6a). The model control run captures the spatial distribution well in the extratropics although with some overestimation, but strong biases exist in South America and Central Africa (Fig. S6b). This is partly related to the deficiency of CLM-CN-DV in distinguishing evergreen trees and deciduous trees in the Tropics. This deficiency is also reflected by simulations driven by the GCM climates. Broadleaf deciduous trees in Southeast Asia can be simulated under most of the GCM climates, but those in Europe and eastern North America are mostly underestimated. Compared with the control run, the HadGEM2-ES-driven present-day simulation produces the most similar deciduous tree coverage in theextratropics.

MODIS data indicates that shrubs dominate the northern high latitudes and regions south of 30ºS (Fig. S7a). CLM-CN-DV simulations including both the control run and those driven by GCM climates can produce shrubs in the Southern Hemisphere but fail to capture the existence of shrubs in the northern high latitudes (Fig. S7b-u).

The model control run underestimates grass coverage in central U.S., and central and eastern Asia. The spatial pattern of grass coverage is simulated well by the control run in eastern South America and Africa, although the coverage fraction is higher than in MODIS (Fig. S8a, b). Most of the GCM-driven simulations produce a similar grass distribution to the control run. Some of them even outperform the control run in capturing the grass coverage in central U.S., Asia and Africa. However, the present-day simulations corresponding to climate of several GCMs (e.g., BCC-CSM1.1m, CNRM-CM5, the three GFDL models, IPSL-CM5A-LR and IPSL-CM5A-MR) significantly overestimate grass coverage in the Amazon region, which is related to their underestimation of tree coverage in that region.

Table Captions

Table S1 Models specification

Figure Captions

Fig.S1Last five years of monthly precipitation (grey bars, in mm/day) and the simulated monthly ELAI from experiments using the original version of CLM4-CN model (dash line) and the modified version including the refinement of the water-controlled phenology scheme (solid line) for one grid point which is occupied primarily by tropical broadleaf deciduous trees

Fig.S2Last 20-year average of fractionalcoverage (%) of tropical broadleaf evergreen trees simulated by the CLM4-CN-DV model(a) in its default version, (b) with the incorporation of GPP parameterization from CLM4.5, (c) with both the incorporation of CLM4.5 GPP parameterization and the refinement of water-controlled phenology scheme, and (d) with all modifications in (c) and the addition of a threshold for the length of soil drought above which tropical broadleaf evergreen trees cannot survive (d). The model version used in producing results shown in (d) is used in this study

Fig.S3 Same as Fig. S2. But for broadleaf deciduous trees-tropical

Fig.S4 Fractional coverage (%) of needleleaf evergreen trees derived from MODIS data (a), and from the last 20 years average in present-day simulations driven with Qian et al.’s forcing (b) and with the 19 GCMs output (c-u)

Fig.S5 Same as Fig. S4, but for broadleaf evergreen trees

Fig.S6Same as Fig. S4, but for broadleaf deciduous trees

Fig.S7 Same asFig. S4, but for shrubs

Fig.S8 Same as Fig. S4, but for grasses

Table S1 Models specification

Model name / Modeling center / Original resolution
(lonlat) / References
ACCESS1.0 / CSIRO (Commonwealth Scientific and Industrial Research Organisation, Australia), and BOM (Bureau of Meteorology, Australia) / 145192 / (Bi et al. 2013)
BCC-CSM1.1 / Beijing Climate Center, China Meteorological Administration / 64128 / (Wu et al. 2013)
BCC-CSM1.1(m) / Beijing Climate Center, China Meteorological Administration / 160320 / (Wu et al. 2014)
BNU-ESM / College of Global Change and Earth System Science, Beijing Normal University / 64128 / (Ji et al. 2014)
CCSM4 / National Center for Atmospheric Research / 192288 / (Gent et al. 2011)
CMCC-CM / Centro Euro-Mediterraneo per I CambiamentiClimatici / 240480 / (Bellucci et al. 2013)
CNRM-CM5 / Centre National de RecherchesMeteorologiques/Centre Europeen de Recherche et Formation Avancees en CalculScientifique / 128256 / (Voldoire et al. 2012)
FGOALS-g2 / LASG, Institute of Atmospheric Physics, Chinese Academy of Sciences / 60128 / (Li et al. 2013)
GFDL-CM3 / Geophysical Fluid Dynamics Laboratory / 90144 / (Griffies et al. 2011)
GFDL-ESM2G / Geophysical Fluid Dynamics Laboratory / 90144 / (Dunne et al. 2012)
GFDL-ESM2M / Geophysical Fluid Dynamics Laboratory / 90144 / (Dunne et al. 2012)
HadGEM2-ES / Met Office Hadley Centre, contributed by InstitutoNacional de PesquisasEspaciais / 144192 / (Collins et al. 2011)
INM-CM4 / Institute for Numerical Mathematics / 120180 / (Volodin et al. 2010)
IPSL-CM5A-LR / Institut Pierre-Simon Laplace / 9696 / (Dufresne et al. 2013)
IPSL-CM5A-MR / Institut Pierre-Simon Laplace / 143144 / (Dufresne et al. 2013)
MIROC5 / Atmosphere and Ocean Research Institute (The University of Tokyo), National Institute for Environmental Studies, and Japan Agency for Marine-Earth Science and Technology / 128256 / (Watanabe et al. 2010)
MIROC-ESM / Japan Agency for Marine-Earth Science and Technology, Atmosphere and Ocean Research Institute (The University of Tokyo), and National Institute for Environmental Studies / 64128 / (Watanabe et al. 2011)
MIROC-ESM-CHEM / Japan Agency for Marine-Earth Science and Technology, Atmosphere and Ocean Research Institute (The University of Tokyo), and National Institute for Environmental Studies / 64128 / (Watanabe et al. 2011)
MRI-CGCM3 / Meteorological Research Institute / 160320 / (Yukimoto et al. 2012)

Fig.S1 Last five years of monthly precipitation (grey bars, in mm/day) and the simulated monthly ELAI from experiments using the original version of CLM4-CN model (dash line) and the modified version including the refinement of the water-controlled phenology scheme (solid line) for one grid point which is occupied primarily by tropical broadleaf deciduous trees

Fig.S2 Last 20-year average of fractionalcoverage (%) of tropical broadleaf evergreen treessimulated by the CLM4-CN-DV model(a) in its default version, (b) with the incorporation of GPP parameterization from CLM4.5, (c) with both the incorporation of CLM4.5 GPP parameterization and the refinement of water-controlled phenology scheme, and (d) with all modifications in (c) and the addition of a threshold for the length of soil drought above which tropical broadleaf evergreen trees cannot survive (d). The model version used in producing results shown in (d) is used in this study

Fig.S3 Same as Fig.S2, but for tropical broadleaf deciduous trees.

Fig.S5 Same as Fig. S4, but for broadleaf evergreen trees

Fig.S6 Same as Fig. S4, but for broadleaf deciduous trees

Fig.S7 Same asFig. S4, but for shrubs

Fig.S8 Same as Fig. S4, but for grasses

References

Bellucci A, Gualdi S, Masina S, Storto A, Scoccimarro E, Cagnazzo C, Fogli P, Manzini E, Navarra A (2013) Decadal climate predictions with a coupled OAGCM initialized with oceanic reanalyses. Climate Dynamics 40 (5-6):1483-1497

Bi D, Dix M, Marsl S, O’Farrell S, Rashid H, Uotila P, Hirst A, Kowalczyk E, Golebiewski M, Sullivan A (2013) The ACCESS coupled model: Description, control climate and evaluation. Aust Met Oceanog J 63:9-32

Collins WJ, Bellouin N, Doutriaux-Boucher M, Gedney N, Halloran P, Hinton T, Hughes J, Jones CD, Joshi M, Liddicoat S et al (2011) Development and evaluation of an Earth-system model – HadGEM2. Geosci Model Dev Discuss 4 (2):997-1062. doi:10.5194/gmdd-4-997-2011

Dufresne J-L, Foujols M-A, Denvil S, Caubel A, Marti O, Aumont O, Balkanski Y, Bekki S, Bellenger H, Benshila R (2013) Climate change projections using the IPSL-CM5 Earth System Model: from CMIP3 to CMIP5. Climate Dynamics:1-43

Dunne JP, John JG, Adcroft AJ, Griffies SM, Hallberg RW, Shevliakova E, Stouffer RJ, Cooke W, Dunne KA, Harrison MJ et al (2012) GFDL’s ESM2 Global Coupled Climate–Carbon Earth System Models. Part I: Physical formulation and baseline simulation characteristics. Journal of Climate 25 (19):6646-6665. doi:10.1175/jcli-d-11-00560.1

Gent PR, Danabasoglu G, Donner LJ, Holland MM, Hunke EC, Jayne SR, Lawrence DM, Neale RB, Rasch PJ, Vertenstein M (2011) The community climate system model version 4. Journal of Climate 24 (19):4973-4991

Gotangco Castillo CK, Levis S, Thornton P (2012) Evaluation of the new CNDV option of the Community Land Model: Effects of dynamic vegetation and interactive nitrogen on CLM4 means and variability. Journal of Climate 25 (11):3702-3714

Griffies SM, Winton M, Donner LJ, Horowitz LW, Downes SM, Farneti R, Gnanadesikan A, Hurlin WJ, Lee H-C, Liang Z et al (2011) The GFDL CM3 coupled climate model: Characteristics of the ocean and sea ice simulations. Journal of Climate 24 (13):3520-3544. doi:10.1175/2011jcli3964.1

Ji D, Wang L, Feng J, Wu Q, Cheng H, Zhang Q, Yang J, Dong W, Dai Y, Gong D (2014) Description and basic evaluation of BNU-ESM version 1. Geoscientific Model Development Discussions 7 (2):1601-1647

Li L, Lin P, Yu Y, Wang B, Zhou T, Liu L, Liu J, Bao Q, Xu S, Huang W (2013) The flexible global ocean-atmosphere-land system model, Grid-point Version 2: FGOALS-g2. Advances in Atmospheric Sciences 30:543-560

Voldoire A, Sanchez-Gomez E, y Mélia DS, Decharme B, Cassou C, Sénési S, Valcke S, Beau I, Alias A, Chevallier M (2012) The CNRM-CM5.1 global climate model: description and basic evaluation. Climate Dynamics:1-31

Volodin E, Dianskii N, Gusev A (2010) Simulating present-day climate with the INMCM4.0 coupled model of the atmospheric and oceanic general circulations. Izvestiya, Atmospheric and Oceanic Physics 46 (4):414-431

Watanabe M, Suzuki T, O'ishi R, Komuro Y, Watanabe S, Emori S, Takemura T, Chikira M, Ogura T, Sekiguchi M (2010) Improved climate simulation by MIROC5: mean states, variability, and climate sensitivity. Journal of Climate 23 (23):6312-6335

Watanabe S, Hajima T, Sudo K, Nagashima T, Takemura T, Okajima H, Nozawa T, Kawase H, Abe M, Yokohata T et al (2011) MIROC-ESM 2010: model description and basic results of CMIP5-20c3m experiments. Geosci Model Dev 4 (4):845-872. doi:10.5194/gmd-4-845-2011

Wu T, Li W, Ji J, Xin X, Li L, Wang Z, Zhang Y, Li J, Zhang F, Wei M (2013) Global carbon budgets simulated by the Beijing Climate Center Climate System Model for the last century. Journal of Geophysical Research: Atmospheres

Wu T, Song L, Li W, Wang Z, Zhang H, Xin X, Zhang Y, Zhang L, Li J, Wu F et al (2014) An overview of BCC climate system model development and application for climate change studies. Acta Meteorol Sin 28 (1):34-56. doi:10.1007/s13351-014-3041-7

Yukimoto S, Adachi Y, Hosaka M (2012) A new global climate model of the Meteorological Research Institute: MRI-CGCM3: model description and basic performance (special issue on recent development on climate models and future climate projections). Journal of the Meteorological Society of Japan 90:23-64