Extended Data Fig. 4: The adjusted response of log (τ) to MAT when the hydrometeorological effect is controlled in multiple regression models. | Nature Geoscience

Extended Data Fig. 4: The adjusted response of log (τ) to MAT when the hydrometeorological effect is controlled in multiple regression models.

From: Global apparent temperature sensitivity of terrestrial carbon turnover modulated by hydrometeorological factors

Extended Data Fig. 4: The adjusted response of log (τ) to MAT when the hydrometeorological effect is controlled in multiple regression models.

Solid black lines represent quadratic fitting curve (log10(τ) ~ MAT2+MAT+1/MAP+1/PET+PSF) while red lines represent linear fit (log10(τ) ~ MAT+1/MAP+1/PET+PSF). The density plot (overlaid) shows the adjusted response of log(τ) to MAT when the average conditions of the other predictors are taken into account. The embedded smaller figure at the lower left corner is Q10 values (derived from quadratic fitting) against temperature. Different data is used to estimate τ with two different carbon stocks (Ctotal and CHWSD) and two carbon flux products (GPPMODIS and GPPFLUXCOM).

Source data

Back to article page