EMICs land surface intercomparison

Description of the experiments

Part 1. Equilibrium simulations

Experiment	Acronym	Vegetation/ land cover	Atmosphere	Ocean	Atmospheric CO2
CONTROL - pre-industrial equilibrium	CTRL	Potential vegetation cover (without land-use)	-	-	280 ppmv
Complete deforestation - atmosphere only	CD-A	Forests/trees are replaced with grassland. The rest of vegetation cover is prescribed from (CTRL). Biophysical parameters (e.g. leaf area index, canopy conductance) are interactive with the atmosphere	Interactive	fixed (CTRL)	fixed (CTRL)
Complete deforestation - atmosphere+ocean	CD-AO	The same as in (CD-A)	Interactive	Interactive (dynamic or slab)	fixed (CTRL)
Complete afforestation - atmosphere only	CA-A	All ice-free land is covered with forests, forest/tree type (e.g. needleaved/broadleaved) is simulated by the vegetation model. Biophysical parameters (e.g. leaf area index, canopy conductance) are interactive with the atmosphere	Interactive	fixed (CTRL)	fixed (CTRL)
Complete afforestation - atmosphere+ocean	CA-AO*	The same as in (CA-A)	Interactive	Interactive (dynamic or slab)	fixed (CTRL)
Complete desertification - atmosphere only	CS-A	All vegetation is replaced with desert (bare soil)	Interactive	fixed (CTRL)	fixed (CTRL)
Complete desertification - atmosphere+ocean	CS-AO*	The same as in (CS-A)	Interactive	Interactive (dynamic or slab)	fixed (CTRL)

^*)obligatory for participating in the Part 1 of the intercomparison

Model Output

Difference (Exp - CTRL)

Variables: air surface temperature (degr.C), precipitation (mm/day), surface albedo, evapotranspiration (mm/day).
Temporal resolution: annual, summer (JJA), winter (DJF), spring (MAM).
Spatial resolution: model grid.
Precision: *.***.

Participants (updated 17.04.02)

Experiment	CLIMBER-2	ECBILT-BIOME	ECBILT-CLIO-VECODE	IAP RAS CM	MIT-TEM	MoBidiC-VECODE	UVic
CD-A	+				+
CD-AO	+		+	+	+		+
CA-A	+				+
CA-AO	+	+	+	+	+	+	+
CS-A	+				+
CS-AO	+	+		+	+	+	+

Part 2. Transient simulation of Historical Land Cover Changes (THLCC)

Experiment	Acronym	Vegetation/ land cover	Atmosphere	Ocean	Atmospheric CO2
Historical deforestation	HD-AO*	Forests/trees are replaced with grassland (cropland) in accordance with R&F 1700-1992 scenario. The rest of vegetation cover is prescribed from (CTRL). Biophysical parameters (e.g. leaf area index, canopy conductance) are interactive with the atmosphere	Interactive	Interactive (dynamic or slab)	fixed (CTRL)
Historical CO2 changes	HC-AO	Vegetation cover is prescribed from (CTRL). Biophysical parameters (e.g. leaf area index, canopy conductance) are interactive with the atmosphere	Interactive	Interactive (dynamic or slab)	Prescribed from the filtered Siple station ice core data and MaunaLoa observations, yr 1700-1992. Other GHG (e.g.CH4) are not accounted for
Historical deforestation + CO2 changes	HDC-AO	The same as in (HD-AO)	Interactive	Interactive (dynamic or slab)	The same as in (HC-AO)
Historical deforestation - interactive carbon cycle	HD-AOC	The same as in (HD-AO)	Interactive	Interactive (dynamic or slab)	Interactive. Oceanic/terrestrial carbon interacts with the atmospheric CO2

^*)obligatory for participating in the Part 2 of the intercomparison

Experimental setup

Download Ramankutty & Foley (R&F)dataset at http://www.sage.wisc.edu/pages/datamodels.html#Anchor-Historical-35882 and scale it to the resolution of the land surface model.
"Warm start" period. Use a linear interpolation for cropland fraction from 0 at the start year to R&F data at the yr 1700. The rationale is avoiding a kick in model response at the yr 1700. With CLIMBER-2, warm start period of 700 years is used, so deforestation is started at the year 1000AD. The same 700-yr warm start period is suggested for the other 2.5 EMICs. For 3-D models, the warm start period phase can be shorter.
During the yr 1700 to 1990, the changes in cropland fraction from R&F dataset are used as the external forcing. In each grid cell, tree fraction is decreased (increased) in accordance with increase(decrease) in cropland fraction. If crops are absent in land surface types, they are treated as grassland.

Model Output

Time slices difference: 1990 - pre-industrial.

Variables: air surface temperature (degr.C), surface albedo, evapotranspiration (mm/day).
Temporal resolution: annual, winter (DJF), spring (MAM), summer (JJA),autumn (SON).
Spatial resolution: zonally averaged, with model latitudional resolution.
Precision: *.****.

Transient output, 1000 (beginning of warm start) - 1990.

Variables: air surface temperature (degr.C), precipitation (mm/day).
Temporal resolution: annual.
Spatial resolution: global, northern hemisphere (temp. only).

Participants (updated 17.04.02)

Experiment	CLIMBER-2	ECBILT-BIOME	ECBILT-CLIO-VECODE	MIT-TEM	MoBidiC-VECODE	UVic
HD-AO	+	+	+	+	+	+
HC-AO	+	+	+	+	+	+
HDC-AO	+	+	+	+	+	+
HD-AOC	+

Intercomparison of the model results

Part 2. Transient simulation of Historical Land Cover Changes (THLCC)

1. A template EXCEL file for result submission is available upon request. The file includes results for HD-AO, HC-AO, HDC-AO simulations in two formats: time series and zonal average. Please write results of your model in the same format and submit to victor@pik-potsdam.de.
2. We need from every participated group a short description of processes in land surface models; particularly, a description of mechanisms of forest impact on climate (masking of snow albedo, changes in transpiration) is very important. Please download as an example a description of land surface processes in CLIMBER-2.

Contact: victor@pik-potsdam.de

Last change: 18.10.2002