Predicting landscape-scale CO2 flux at a pasture and rice paddy with long-term hyperspectral canopy reflectance measurements

[摘要] Measurements of hyperspectral canopy reflectance provide a detailed snapshotof information regarding canopy biochemistry, structure and physiology. Inthis study, we collected 5 years of repeated canopy hyperspectralreflectance measurements for a total of over 100 site visits within the fluxfootprints of two eddy covariance towers at a pasture and rice paddy innorthern California. The vegetation at both sites exhibited dynamicphenology, with significant interannual variability in the timing ofseasonal patterns that propagated into interannual variability in measuredhyperspectral reflectance. We used partial least-squares regression (PLSR)modeling to leverage the information contained within the entire canopyreflectance spectra (400–900 nm) in order to investigate questions regardingthe connection between measured hyperspectral reflectance andlandscape-scale fluxes of net ecosystem exchange (NEE) and gross primaryproductivity (GPP) across multiple timescales, from instantaneous flux tomonthly integrated flux. With the PLSR models developed from this largedata set we achieved a high level of predictability for both NEE and GPP fluxin these two ecosystems, where the R² of prediction with an independentvalidation data set ranged from 0.24 to 0.69. The PLSR models achieved thehighest skill at predicting the integrated GPP flux for the week prior tothe hyperspectral canopy reflectance collection, whereas the NEE flux oftenachieved the same high predictive power at daily tomonthly integrated flux timescales. The high level of predictabilityachieved by PLSR in this study demonstrated the potential forusing repeated hyperspectral canopy reflectance measurements to helppartition NEE into its component fluxes, GPP and ecosystemrespiration, and for using quasi-continuous hyperspectral reflectancemeasurements to model regional carbon flux in future analyses.

[学科分类] 地球化学与岩石