Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
The State of CLM LMWG Activities
David Lawrence NCAR Earth System Laboratory
with input from members of LMWG and BGCWG
NCAR is sponsored by the National Science Foundation
Degrees in Boulder-based CLM group
0 2 4 6 8
Ecology
Hydrology
Electrical Engineering
Climatology
Math
Atmospheric Science
Aerospace Engineering
Finance
Geology
Chemical Engineering
Geography
Computer Science
Biology
Environmental Science
Physics
B.S.
Ph.D.
CLM at AGU
0
10
20
30
40
50
60
70
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
20
10
20
11
Presentations with CLM in abstract or title
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
20
04
20
05
20
06
20
07
20
08
20
09
20
10
20
11
% of AGU presentations that included CLM
LMWG Activities
www.ilamb.org
Jim Randerson, Forrest Hoffman, Pierre Friedlingstein, Chris Jones
Water and Energy: David Lawrence, Eleanor Blyth,
Martin Best
Example: Metric based on global RMSE
2ijt ijt ijt
i j t
obs,ijt
1RMSE = w (Model Obs )W
RMSEM = 0 1cσ
−
< −
∑∑∑
wijt are the area and month weights, W is the total of the weights
i,j,t are the lon,lat, and month
Model and Obs are climatological data (20yr avg)
Amplitude of σijt controls average position of M (0-1),
influence of σijt can be adjusted with c
What is best/most meaningful normalization ?
RMSE Metric tests spatial pattern and annual cycle
Not a direct test of a land model process
Global RMSE (LH, centered); FLUXNET-MTE (Jung et al. 2010)
ILAMB – Water and energy (and surface climate) benchmarking (very preliminary)
0
0.5
1
1.5
2
2.5
3
3.5
4
PREC_BASIN
RUNOFF_BASIN
RUNOFF_RATIO
albedo
LHF
PREC
TSA
CLM-CN parameter sensitivity analysis (DOE CSSEF)
• Ensemble capability developed for CLM-CN point version
• Up to 1000 simultaneous point ensembles varying pft-physiology and surface data
• Nearly 50% of runs do not produce any vegetation (parameters incompatible with growth)
• Will eventually be used for model parameter calibration
1000 member ensemble run at Niwot
Ridge flux site, full spinup and transient simulations for each simulation
• 9 parameters varied over ranges determined by literature survey
• Comparing observations against full model uncertainty range useful for validation
• Observations suggest model has too much cold-season GPP
N = 520
Point CLM Ensemble (1000)
members
Parameter samples mp leafcn flnr froot_leaf leaf_long rc_npool br_mr q10_mr r_mort
month
Post-processing (2M files)
Slide courtesy Dan Ricciuto
Model Intercomparison Projects
• TRENDY
– Historic carbon cycle simulations with revised CRUNCEP data
• GLACE-CMIP5
– How do projected changes in soil moisture feedback onto future climate change and carbon cycle
• Permafrost-carbon (Permafrost RCN)
– Historic and future permafrost loss and permafrost-carbon feedback
• ALMIP2
– AMMA Land surface and hydrological Model Inter-comparison Project
• GSWP3
– 20th century water (and carbon cycles)
• Land use / land cover change, LUCID extension
• PILDAS - Project for the Intercomparison of Land Data Assimilation Systems
Data assimilation
• See Tim Hoar’s talk
CLM Development Activities
Progress with CLM4.5
• Original goal was to release CLM4.5 with next CESM release in November 2012, but …
Branch tag chaos!
Second meeting to straighten things out
Progress with CLM4.5
• Original goal was to release CLM4.5 with next CESM release in November 2012
• Too many loose ends, so now targeting May 2013 release (code freeze on Feb 1)
• For November 2013 release
• High resolution surface datasets
• High resolution RTM
• WRF-CLM
Progress with CLM4 development (for November CESM1.1 release)
– Unstructured grids (non lat/lon grids)
– High resolution input datasets (PFTs, lake area, glacier area, etc)
– High resolution River Transport Model (0.1o) – ocean mapping, inland depressions
– Multi-ensemble mode for CLM
– Update to MEGAN model for BVOC’s, replaces the 5 hardcoded VOC's to a general solution allowing up to 150 compounds
– Better snow accumulation/loss diagnostics
– WRF(3.2) / CLM
• this is a goal, not a promise
• currently working with WRF software engineers and scientists validating a western US simulation and refining scripts
CESM1.1: High resolution input datasets (-hires option in mksrfdat.pl)
Input dataset CLM4 resolution Updated resolution Status
PFT distribution 0.5o (MODIS) 3’ (MODIS) x
LAI / SAI 0.5o (MODIS) 0.5o (MODIS) x (?)
% Glacier 0.5o (IGBP DISCover)
1km (Gardner) [Bill] ?
% Lake, Lake depth 0.5o (Cogley, 1991) 3’ (GLWD) x
% Wetland 0.5o (Cogley, 1991) Prognostic x
% Urban 0.5o (?) 3’ Jackson et al. 2010 x
Soil texture (%sand, %clay)
5’ (IGBP) 5’ (IGBP for now; ISRIC-WISE for multiple soil classes) [Johann]
x
Soil organic matter 1.0o (IGBP) 5’ (ISRIC-WISE/NCSCD) [Dave] x (meta)
Soil color 0.5o (MODIS) 0.5o (MODIS) x
Fmax 0.5o 0.125o calc from HydroSHEDS x
RTM Directional Map 0.5o 0.1o x (coupling?)
Irrigation/Crop types 5’ 5’ (Navin) [Sam] X
Topography (GLCMEC) 10’ (USGS) 1km ?? (USGS) x
Progress with CLM4.5 development (for May 2013 release)
• Completed and on “science branch”
– Revised photosynthesis model, multilayer canopy option
– Cold region hydrology and snow model updates
– CENTURY-like vertically resolved soil biogeochemistry (option)
– New lake model
– CH4 emissions (coupling to CAM working)
– Switch Arctic C3 grass and boreal shrub from stress to seasonal decid.
– Soil history fields for vegetated landunit only, high vertical soil resolution namelist option, hydrology/soil temperature resolution namelist option
– Faster CN spinup
– C13/C14 enabled
• Close to being completed
– 2-way CLM / RTM interactions (flooding), variable flow rates
– Prognostic wetland distribution model (eliminate wetland landunit, what about coastlines???)
Grid cell averaged TLAI – CLM4
Problem with high-latitude Arctic C3 grass and boreal shrub phenology
stress deciduous 6 broadleaf_evergreen_temperate_tree 10 broadleaf_deciduous_temperate_shrub 11, broadleaf_deciduous_boreal_shrub 12 c3_arctic_grass 13 c3_non-arctic_grass 14 c4_grass 15 c3_crop 16 c3_irrigated
seasonal deciduous 3 needleleaf deciduous_boreal_tree 7 broadleaf_deciduous_temperate_tree 8 broadleaf_deciduous_boreal_tree 11 broadleaf_deciduous_boreal_shrub 12 c3_arctic_grass
Correlation of simulated TLAI vs MODIS TLAI: boreal shrub
CLM4
Phenology fix
Grid cell averaged TLAI – CLM4
Problem with high-latitude Arctic C3 grass and boreal shrub phenology
Grid cell averaged TLAI - Revised
Progress with CLM4.5 development (for May 2013 release)
• Completed and on “science branch”
– Revised photosynthesis model, multilayer canopy option
– Cold region hydrology and snow model updates
– CENTURY-like vertically resolved soil biogeochemistry (option)
– New lake model
– CH4 emissions (coupling to CAM working)
– Switch Arctic C3 grass and boreal shrub from stress to seasonal decid.
– Soil history fields for vegetated landunit only, high vertical soil resolution namelist option, hydrology/soil temperature resolution namelist option
– Faster CN spinup
– C13/C14 enabled ???
• Close to being completed
– 2-way CLM / RTM interactions (flooding), variable flow rates
– Prognostic wetland distribution model (eliminate wetland landunit, what about coastlines???)
Progress with CLM4.5 development
• Next up for “science branch
– Crop model improvements: Interactive fertilization, biological soy N fixation, new C:N ratios, N retranslocation, separate organs pool, connect crops and irrigation, crops always on separate landunit
– Separate above and below-ground litter
– VIC hydrology (optional)
– Urban developments: multiple urban classes, improved spatially-explicit representation of present-day and future urban characteristics, improved anthropogenic heat fluxes
– Revised fire model (Li or Kloster version, TBD)
– Revised bare ice glacier albedo
– Add fields to mksrfdat: lake depth, slope, multiple urban classes, crop/irrigation, new fire fields (pop. density, lightning)
Progress with CLM4.5 development
• Development work and/or testing still required
– CLM4.5 as an option, maintain ability to revert to CLM4
– Soil evaporation param and rooting depth changes to solve hydrological response to land cover change problem
– Ecosystem Demography (option)
– Dynamic landunits (e.g., glacier to vegetated transitions)
– Riverine transport of nutrients
– Bare soil roughness length and friction velocity parameter change (testing)
NCAR is sponsored by the National Science Foundation
Improving Throughput of Science in CLM – via Software Testing (Wed. SEWG 9:35am)
Erik Kluzek, Software Engineer Land Model Working Group Liason National Center for Atmospheric Research (NCAR)
Beyond CLM4.5
Other ongoing CLM development activities
– Sub-surface hydrological processes – lateral redistribution of water
– Sub-grid PFT distribution (elevation dependence)
– Sub-grid soil moisture and snow heterogeneity
– IAM-ESM coupling (iESM) (SDWG)
– Soil microbial dynamics, multi-phase transport, multiple tracers in soil (CLM4-BeTR)
– Spatially explicit soil depth
– Water isotopes
– Peatlands
– Excess soil ice in permafrost / thermokarst
– Other nutrient cycles (e.g., phosphorous)
– Ozone poisoning of vegetation
– Data assimilation
– 3-D canopy radiation
– Prognostic canopy air space
http://www.cesm.ucar.edu/models/cesm1.0/clm/index.shtml
252 references
Control Model
Modified Model
Control Model
Modified Model
Revised surface flux calculations
Effects of Anthropogenic Heat and Urban Density
31
JJA Surface Temperature (°F) 1990-2009 Urban – Rural MIN Air Temp
Urban Fraction
CLM Bibliography: http://www.cesm.ucar.edu/models/cesm1.0/clm/clm_bibliography.htm
• Paolino, D. A., J. L. Kinter, et al. (2012). "The Impact of Land Surface and Atmospheric Initialization on Seasonal Forecasts with CCSM." Journal of Climate 25(3): 1007-1021.
• Zubler, E. M., U. Lohmann, et al. (2011). "Intercomparison of aerosol climatologies for use in a regional climate model over Europe." Geophysical Research Letters 38.
• Yuan, X. and X.-Z. Liang (2011). "Evaluation of a Conjunctive Surface-Subsurface Process Model (CSSP) over the Contiguous United States at Regional- Local Scales." Journal of Hydrometeorology 12(4): 579-599.
• Xu, L. and P. Dirmeyer (2011). "Snow-atmosphere coupling strength in a global atmospheric model." Geophys. Res. Lett. 38(13): L13401.
• Wang, G. L. (2011). "Assessing the potential hydrological impacts of hydraulic redistribution in Amazonia using a numerical modeling approach." Water Resources Research 47.
• Wang, A. H. and X. B. Zeng (2011). "Sensitivities of terrestrial water cycle simulations to the variations of precipitation and air temperature in China." Journal of Geophysical Research-Atmospheres 116.
• Wang, A., D. P. Lettenmaier, et al. (2011). "Soil Moisture Drought in China, 1950-2006." Journal of Climate 24(13): 3257-3271.
• Tawfik, A. B. and A. L. Steiner (2011). "The role of soil ice in land-atmosphere coupling over the United States: A soil moisture-precipitation winter feedback mechanism." Journal of Geophysical Research-Atmospheres 116.
• Sun, S. S. and G. L. Wang (2011). "Diagnosing the equilibrium state of a coupled global biosphere-atmosphere model." Journal of Geophysical Research-Atmospheres 116.
• Shi, X. Y., J. F. Mao, et al. (2011). "The impact of climate, CO2, nitrogen deposition and land use change on simulated contemporary global river flow." Geophysical Research Letters 38.
• Shi, C. X., Z. H. Xie, et al. (2011). "China land soil moisture EnKF data assimilation based on satellite remote sensing data." Science China-Earth Sciences 54(9): 1430-1440.
• Schwinger, J., S. J. Kollet, et al. (2011). "Sensitivity of Latent Heat Fluxes to Initial Values and Parameters of a Land-Surface Model." Vadose Zone Journal 9(4): 984-1001.
• Sakaguchi, K., X. B. Zeng, et al. (2011). "Natural and drought scenarios in an east central Amazon forest: Fidelity of the Community Land Model 3.5 with three biogeochemical models." Journal of Geophysical Research-
CESM1.1: Unstructured Grids in CLM
• Capability to run with non lat/lon or logically rectangular grids
– Surface dataset generation tool for non lat/lon grids
– CLM code support to deal with non lat/lon surface datasets and generate appropriate history files
– Post-processing utility to map non lat/lon history files to 2d for visualization/analysis
• New ways to run CLM
– Cubed sphere and regionally refined grids
– Collection of tower sites in parallel
– Catchment grid
LAI July 2003 (~5km resolution)
CESM1.1: High resolution input datasets
Increasing emphasis and demand across CESM/CLM research communities for high resolution, but … CLM input datasets mostly at ~0.5o and RTM fixed at 0.5o
Adding Flooding Capability / 2-way CLM-RTM interactions
CLM
Ocean (POP)
RTM
Surface Runoff
Baseflow
New Capability: Flood Water Taken From RTM is Sent to CLM
• Crop model and irrigation
– Connect crops and irrigation
– Interactive fertilization based on N demand
– Separate organs/grains pool
– New planting date dataset and phenological heat units
– Biological N fixation for soy (legumes are N fixers)
– Crop C:N ratios
– N retranslocation
CLM4.5 (potential release with CESM update late 2012)
CLM4.5 (potential release with CESM update late 2012)
– Dynamic landunits
• Prescribed and prognostic land unit transitions: e.g., glacier to vegetated, vegetated to crop, vegetated to urban, etc.
– Faster CN spinup
• Speedup of 2x to 4x for CN spinup
Faster CLM carbon cycle spinup procedure C
LM
4
GPP abs(NEE)
CL
M4
.5
2x to 4x faster spinup depending on location (further quantification needed) Will allow more rapid model equilibration for model development parameter sensitivity and optimization
Exit spinup mode
CLM4.5 (potential release with CESM update late 2012)
– Improved fire algorithm
• Includes human triggers and suppression (Kloster et al., 2010)
Newer fire parameterization scheme (Li and Zeng, 2011; Li et al., 2012)
CLM4.5 (potential release with CESM update late 2012)
– Improved improved fire algorithm
• Includes human triggers and suppression (Kloster et al., 2010)
• Agricultural, deforestation fires, economic influence (Li and Zeng, 2011; Li et al, 2012)
Spatial correlation R = 0.52 Kloster R= 0.65 Li Carbon emissions GFED3 = 2.1 PgC/yr Kloster = 2.0 PgC/yr Li = 2.1 PgC/yr
Global area burned
CLM4.5 (potential release with CESM update late 2012)
– Reconsider soil evaporation and under canopy turbulence parameterizations from water cycle response to land cover change perspective
CLM4.5 (potential release with CESM update late 2012)
Additional potential options (depending on how much we can get done)
– Ecosystem demography option
– VIC hydrology option
– Infrastructure for riverine transport of nutrients, carbon, and sediments
– N2O emissions
Planned urban model developments
• Dataset development
– Enable multiple urban classes
– Improved spatially-explicit representation of present-day and future urban characteristics
• Urban emissions (long term goal, but no active project)
CBD
Medium Density Low Density
High Density
Urban Properties • Height • H/W ratio • Vegetated fraction • Roof fraction
Wall properties • Albedo • Thermal properties • Radiative properties
Roof properties Road properties Interior Tsettings
Tools: Land Uncertainty Quantification (UQ)
- CLM parameter sensitivity analysis (CSSEF)
• Carbon, Hydrology, Crops
– Data assimilation NEON/NCAR
– CLM parameter calibration/optimization (eventually)
Larger sensitivity to parameters of subsurface processes
Challenges
• Increasing complexity
– Increasing number of options, configurations
– Increasingly difficult to maintain and test
– Strengthening interactions with BGCWG, ChemClim WG and SDWG
• Increasing user base
– Number of users around the world growing rapidly
– CLM used in NorESM and WRF, COSMO, and HIRHAM regional models
– Inquiries: Sam Levis estimated over 250 CLM email conversations in 2011
– Tutorials [China (proposed), Cornell (funded) ] / Lectures
• Expanding number of developers
– Huge investment from DOE through CSSEF, IMPACTS, and NGEE (future)
– Increase in NCAR staff, but mostly postdocs
• Other
– Lack of robust and extensible testbed/model evaluation system (ILAMB?)
– Strengthening / defining relationship with NEON
– WRF/CLM support?
Ecosystem Demography Model (ED) “Size-and-age structured approximation of a gap model”
Bare Gd Shrub
C4
C3 NL tree
BL tree
1 y.o. 5 y.o.
15 y.o.
30 y.o. 60 y.o.
90 y.o.
ED
Time since disturbance-based tile structure. PFT-based tile structure.
Slide courtesy Rosie Fisher
50
iESM Experiment – first approach to full coupling
RCP 4.5 Climate
Rep 4.5 Feedback
• Experiment 1: RCP 4.5, with feedback from the CLM-CCSM calculations
– 15-year time steps;
– Lagged adjustment:
Carbon Density and Productivity
1. GCAM(Human
Dimensions Elements only;
15 ghgs, aerosols, SLS; 14
geopolitical regions; 151 Ecoregions)
2. GLM(½ x ½ degree
grid land-use-land-cover.)
ESM1(3. CLM & 4. CCSM)
Fossil Fuel & Industrial Emissions (Gridded)
Land Use
LU-LC
Proposed Multi-scale Land Cover Change Project
Experiments: Each set of resolutions would be run for 30 years with 1850 (potential vegetation) and again with 2000 land cover
Configuration: CAM5 AMIP with present day SSTs, CO2 and aerosol dep (CN: off; irrigation: on for 2000 land cover; high res RTM?)
Focus region: preferably centered slightly to the east of the one shown below to cover midwest crop + eastern US forests
Resolution atm Resolution lnd
ne30 (1o) ne30 (1o)
ne30 (1o) ne240 (0.125o) focus ne30 elsewhere
ne240 (0.125o) focus ne30 (1o) elsewhere
ne30 (1o)
ne240 (0.125o) focus ne30 (1o) elsewhere
ne240 (0.125o) focus ne30 (1o)elsewhere
B.S. and Ph.D. Degrees in CLM group
• B.S.
– Chemical Engineering
– Biology/Geography
– Physics/Geophysics/Geology
– Environmental Science
– Math/Geography
– Chemical Engineering
– Computer Science
– Physics
– Finance/Climate & Resource Management
– Atmospheric Science
– Computer Science
– Physics
– Environmental Science/Geology
– Geology/Physics
– Environmental Science
– Physics/Math
– Physics
– Aerospace Engineering
– Geography
– Environmental Sciences
– Biology
– Horticulture
– Biology
• PhD (highest degree)
– Chemical Engineering
– Climatology
– Geophysics (MS)
– Ecology and Evolutionary Biology
– Meteorology (MS)
– Atmospheric Science
– Environment and Resources
– Atmosphere and Ocean Science
– Geography
– Hydrology and water resources
– Geographical Sciences
– Atmosphere and Ocean Science
– Geology
– Electrical and Computer Engineering
– Geological and Environmental Sciences
– Atmospheric Physics
– Physics (Remote Sensing)
– Aerospace Engineering
– Geography
– Environmental Sciences
– Ecology and Evolutionary Biology
– Ecosystem Ecology
– Environmental Science