Variables

The variable library fixes how each Earth observation variable is named, labelled and coloured, so NDVI or land surface temperature look the same in every figure of a paper.

import eofigs

ndvi = eofigs.variable("NDVI")  # by name, alias or CF standard name
ndvi.label  # 'Normalized Difference Vegetation Index'
ndvi.cmap  # 'cmc.bamako_r'

gpp = eofigs.variable("gpp")
gpp.short_label  # 'GPP (gC m$^{-2}$ d$^{-1}$)'
gpp.trend("year").units  # 'gC m$^{-2}$ d$^{-1}$ yr$^{-1}$'
gpp.anomaly().cmap  # 'eofigs.terra_slate', zero-centred

eofigs.variable(dataarray)  # recognised from standard_name, long_name or name

Anomalies and trends of any variable switch to a zero-centred diverging colour map, with the direction that reads naturally:

Variables	Map	Negative	Positive
vegetation indices and productivity (NDVI, EVI, kNDVI, NIRv, LAI, FAPAR, GPP, NPP, SIF)	`eofigs.terra_slate`	terracotta (browning)	slate teal (greening)
vegetation structure (VOD, canopy height, biomass, tree cover)	`eofigs.terra_slate`	terracotta (loss)	slate teal (gain)
water (precipitation, SPEI, soil moisture, evapotranspiration)	`cmc.vik_r`	brown-red (drying)	blue (wetting)
temperature (LST, air temperature)	`cmc.vik`	blue (cooling)	red (warming)
terrain (elevation, slope) and burn severity (dNBR)	`cmc.vik`	blue	red (higher, more severe)

Units

Units are written in matplotlib mathtext (gC m$^{-2}$ d$^{-1}$), which every journal font can typeset; Unicode superscripts such as "⁻¹" are missing from some fonts. The library units are the units the data are expected in: eofigs never converts units. When a DataArray has a units attribute that differs from the library, variable() uses the data's units instead (CF strings such as kg m-2 s-1 are written in mathtext).

Your own variables

eofigs.register_variable(
    "cci",
    long_name="Chlorophyll carotenoid index",
    symbol="CCI",
    cmap="cmc.bamako_r",
    vmin=-0.5,
    vmax=0.5,
)

Registered variables are found by variable() and every recipe for the rest of the session. A Variable object can also be passed directly wherever a variable name is accepted.

Bundled variables

Colour maps follow perceptually uniform cmcrameri maps: bamako_r for vegetation, davos_r for water, lajolla for temperature (bright = hot) and lajolla_r for burn severity (dark = severe), batlow for terrain.

Name	Long name	Symbol	Units	Colour map	Anomaly and trend map	Default range
`agb`	Aboveground biomass	AGB	Mg ha^-1	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 400
`air_temperature`	Air temperature	T_a	°C	`cmc.lajolla`	`cmc.vik`	from data
`canopy_height`	Canopy height	h	m	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 50
`dnbr`	Differenced Normalized Burn Ratio	dNBR	–	`cmc.lajolla_r`	`cmc.vik`	-0.25 to 1
`elevation`	Elevation	z	m	`cmc.batlow`	`cmc.vik`	from data
`evapotranspiration`	Evapotranspiration	ET	mm d^-1	`cmc.davos_r`	`cmc.vik_r`	0 to data
`evi`	Enhanced Vegetation Index	EVI	–	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 1
`fapar`	Fraction of absorbed photosynthetically active radiation	FAPAR	–	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 1
`gpp`	Gross primary productivity	GPP	gC m^-2 d^-1	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 15
`kndvi`	Kernel NDVI	kNDVI	–	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 1
`lai`	Leaf area index	LAI	m² m^-2	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 7
`lst`	Land surface temperature	LST	°C	`cmc.lajolla`	`cmc.vik`	from data
`ndvi`	Normalized Difference Vegetation Index	NDVI	–	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 1
`nirv`	Near-infrared reflectance of vegetation	NIR_v	–	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 0.5
`npp`	Net primary productivity	NPP	gC m^-2 yr^-1	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 1500
`precipitation`	Precipitation	P	mm	`cmc.davos_r`	`cmc.vik_r`	0 to data
`sif`	Solar-induced chlorophyll fluorescence	SIF	mW m^-2 nm^-1 sr^-1	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 2
`soil_moisture`	Soil moisture	SM	m³ m^-3	`cmc.davos_r`	`cmc.vik_r`	0 to 0.5
`spei`	Standardised Precipitation-Evapotranspiration Index	SPEI	–	`cmc.vik_r`	`cmc.vik_r`	-3 to 3
`terrain_slope`	Terrain slope	Slope	°	`cmc.batlow`	`cmc.vik`	0 to 60
`tree_cover`	Tree cover	TC	%	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 100
`vod`	Vegetation optical depth	VOD	–	`cmc.bamako_r`	`eofigs.terra_slate`	0 to 1.5

The table is generated from eofigs/data/variables.json, which the R version of eofigs will share.