CliMA Land Benchmarks

This blog tracks the performance of the CliMA Land global simulations in terms its spatial patterns in the following aspects:

Gross primary productivity (GPP)
Vegetation transpiration (T)
Normalized differential vegetation index (NDVI)
Enhanced vegetation index (EVI)
Near infrared reflection of vegetation (NIRv)
Solar induced chlorophyll fluorescence (SIF) at 683, 740, 757, and 771 nm

Per release, the major improvements are:

a14_gm2_wd1: GPP error slightly increased (back to GM2 with no Vcmax scaling)
a14_gm3_wd1: GPP error decreased
a14_gm4_wd1: GPP error decreased
a13_gm4_wd1: GPP error decreased (BUT savana has higher GPP)
a13_gm3_wd1: GPP error increased (due to C4; AND savana has higher GPP)
a12_gm3_wd1: GPP and VIs error increased (more grids now!) and SIF errors is not impacted on average
a11_gm3_wd1: GPP and SIF errors decreased (scale Vcmax to 60% of the dataset)
a11_gm2_wd1: GPP and SIF errors decreased (carotenoid contribution to PPAR removed)
a10_gm2_wd1: GPP error decreased (new photosynthesis model and adding boundary layer conductance) slightly while SIF errors increased slightly (due to adding UV radiation)
a9_gm2_wd1: GPP and SIF errors decreased slightly (due to adding Kd temperature dependency)
a8_gm2_wd1: GPP error increased slightly while reflectance and SIF errors decreased (due to adding CI seasonality)
a7_gm1_wd1: GPP error decreased while SIF errors increased slightly (due to adding a smoothing algorithm for J calculation)
a6_gm1_wd1: GPP error increased while SIF errors decreased slightly (fixed a typo of the coefficient 1.6, was 1 before)
a5_gm1_wd1: GPP and SIF errors increased (fixed the SIF yield scaling bug)
a4_gm1_wd1: GPP error increased slightly (included FR contribution to PPAR, namely 700-750 nm)
a3_gm1_wd1: GPP error increased slightly (add Vcmax profiles)
a2_gm1_wd1: GPP and SIF errors changed a little bit (SIF wavelengths more accurate, have Vcmax25 TD based on moving average T)
a1_gm1_wd1: very first global simulation

The RMSE per variable per simulation are

Version	GPP	T	NDVI	EVI	NIRv	SIF 683	SIF 740	SIF 757	SIF 771
a14_gm2_wd1	1.785	NA	0.173	0.151	0.498	0.043	0.134	0.095	0.077
a14_gm3_wd1	1.731	NA	0.173	0.151	0.498	0.041	0.132	0.095	0.079
a14_gm4_wd1	1.849	NA	0.173	0.151	0.498	0.039	0.131	0.096	0.081
a13_gm4_wd1	1.980	NA	0.173	0.151	0.498	0.040	0.132	0.097	0.080
a13_gm3_wd1	2.284	NA	0.173	0.151	0.498	0.042	0.134	0.096	0.079
a12_gm3_wd1	1.764	NA	0.173	0.151	0.493	0.044	0.141	0.097	0.079
a11_gm3_wd1	1.368	NA	0.135	0.160	0.257	0.051	0.154	0.095	0.063
a11_gm2_wd1	2.276	NA	0.135	0.160	0.257	0.057	0.178	0.110	0.066
a10_gm2_wd1	2.616	NA	0.135	0.160	0.257	0.071	0.224	0.139	0.076
a9_gm2_wd1	2.751	NA	0.134	0.161	0.257	0.068	0.220	0.136	0.075
a8_gm2_wd1	2.783	NA	0.134	0.161	0.257	0.070	0.231	0.144	0.079
a7_gm1_wd1	2.561	NA	0.139	0.173	0.259	0.074	0.261	0.167	0.091
a6_gm1_wd1	3.371	NA	0.139	0.173	0.259	0.073	0.257	0.163	0.088
a5_gm1_wd1	2.619	NA	0.138	0.174	0.259	0.074	0.260	0.166	0.090
a4_gm1_wd1	2.567	NA	0.138	0.174	0.258	0.057	0.185	0.117	0.068
a3_gm1_wd1	2.382	NA	0.138	0.174	0.258	0.057	0.185	0.117	0.068
a2_gm1_wd1	2.368	NA	0.138	0.174	0.258	0.057	0.185	0.117	0.068
a1_gm1_wd1	2.374	NA	0.138	0.174	0.258	0.047	0.185	0.131	0.065

Model Versions

Version Tag	Linear Change History
A14	Compute beta using SWC
	Add option to remove Rd from GPP
A13	Fix a bug in PFT fractions
	Add C4 photosynthesis support
A12	Fix a bug in f_sunlit (RT impacted)
A11	Not counting Car absorption in PPAR+SIF
	Add new SIF spectrum option (V7)
A10	Extend spectra down to 300 nm
	Balance FR and VIS PAR in photosyntheis
	Bug fix: account for wind in Gb
	Remove f_PSII option
	Add option to prescribe leaf refl+trans
	Add option to use broadband soil albedo
	Set f_PPAR to 1 in broadband RT mode
A9	Add option for a TD Kd and maxPSII
	Add function to prescribe Kf
A8	Fix a bug in sunlit APAR
A7	Add quadratic colimitation to C3 J (0.7)
	Add option to prescribe f_PSII
	Add option to set SIF excitation WL
A6	Fix a bug in Medlyn model (coeff 1.6)
Note	History in ClimaLand-0.1
A5	Fix SIF yield scaling bug
	Add ePAR scenario
	Add prescibed CI option
A4	Extend PAR definition to 400-750 nm
A3	Add vertical Vcmax profile
A2	Use minimum beta instead of mean beta
	Use moving average optimal T from CLM5
A1	Version used for initial run

Weather Drivers

Version Tag	Description
WD1	ERA5 Single Levels

GriddingMachine Selections

GM4

Variable	GriddingMachine Tag
s_cc	SC_2X_1Y_V1
s_α	SOIL_VGA_12X_1Y_V1
s_n	SOIL_VGN_12X_1Y_V1
s_Θr	SOIL_SWCR_12X_1Y_V1
s_Θs	SOIL_SWCS_12X_1Y_V1
p_ch	CH_20X_1Y_V1
p_chl	CHL_2X_7D_V1
p_ci	CI_2X_1M_V3
p_lai	LAI_MODIS_2X_8D_YYYY_V1
p_sla	SLA_2X_1Y_V1
p_vcm	VCMAX_CESM_1X_1M_V3
t_ele	ELEV_4X_1Y_V1
t_lm	LM_4X_1Y_V1
t_pft	PFT_2X_1Y_V1

GM3

Variable	GriddingMachine Tag
s_cc	SC_2X_1Y_V1
s_α	SOIL_VGA_12X_1Y_V1
s_n	SOIL_VGN_12X_1Y_V1
s_Θr	SOIL_SWCR_12X_1Y_V1
s_Θs	SOIL_SWCS_12X_1Y_V1
p_ch	CH_20X_1Y_V1
p_chl	CHL_2X_7D_V1
p_ci	CI_2X_1M_V3
p_lai	LAI_MODIS_2X_8D_YYYY_V1
p_sla	SLA_2X_1Y_V1
p_vcm	VCMAX_2X_1Y_V2 (60%)
t_ele	ELEV_4X_1Y_V1
t_lm	LM_4X_1Y_V1
t_pft	PFT_2X_1Y_V1

GM2 (default)

Variable	GriddingMachine Tag
s_cc	SC_2X_1Y_V1
s_α	SOIL_VGA_12X_1Y_V1
s_n	SOIL_VGN_12X_1Y_V1
s_Θr	SOIL_SWCR_12X_1Y_V1
s_Θs	SOIL_SWCS_12X_1Y_V1
p_ch	CH_20X_1Y_V1
p_chl	CHL_2X_7D_V1
p_ci	CI_2X_1M_V3
p_lai	LAI_MODIS_2X_8D_YYYY_V1
p_sla	SLA_2X_1Y_V1
p_vcm	VCMAX_2X_1Y_V2
t_ele	ELEV_4X_1Y_V1
t_lm	LM_4X_1Y_V1
t_pft	PFT_2X_1Y_V1

GM1

Variable	GriddingMachine Tag
s_cc	SC_2X_1Y_V1
s_α	SOIL_VGA_12X_1Y_V1
s_n	SOIL_VGN_12X_1Y_V1
s_Θr	SOIL_SWCR_12X_1Y_V1
s_Θs	SOIL_SWCS_12X_1Y_V1
p_ch	CH_20X_1Y_V1
p_chl	CHL_2X_7D_V1
p_ci	CI_2X_1Y_V1
p_lai	LAI_MODIS_2X_8D_YYYY_V1
p_sla	SLA_2X_1Y_V1
p_vcm	VCMAX_2X_1Y_V2
t_ele	ELEV_4X_1Y_V1
t_lm	LM_4X_1Y_V1
t_pft	PFT_2X_1Y_V1