Performance of a two-leaf light use efficiency model for mapping gross primary productivity against remotely sensed...

Estimating terrestrial gross primary production is an important task when studying the carbon cycle. In this study, the ability of a two-leaf light use efficiency model to simulate regional gross primary production in China was validated using satellite Global Ozone Monitoring Instrument - 2 sun-induced chlorophyll fluorescence data. The two-leaf light use efficiency model was used to estimate daily gross primary production in China's terrestrial ecosystems with 500-m resolution for the period from 2007 to 2014. Gross primary production simulated with the two-leaf light use efficiency model was resampled to a spatial resolution of 0.5° and then compared with sun-induced chlorophyll fluorescence. During the study period, sun-induced chlorophyll fluorescence and gross primary production simulated by the two-leaf light use efficiency model exhibited similar spatial and temporal patterns in China. The correlation coefficient between sun-induced chlorophyll fluorescence and monthly gross primary production simulated by the two-leaf light use efficiency model was significant (p<0.05, n=96) in 88.9% of vegetated areas in China (average value 0.78) and varied among vegetation types. The interannual variations in monthly sun-induced chlorophyll fluorescence and gross primary production simulated by the two-leaf light use efficiency model were similar in spring and autumn in most vegetated regions, but dissimilar in winter and summer. The spatial variability of sun-induced chlorophyll fluorescence and gross primary production simulated by the two-leaf light use efficiency model was similar in spring, summer, and autumn. The proportion of spatial variations of sun-induced chlorophyll fluorescence and annual gross primary production simulated by the two-leaf light use efficiency model explained by ranged from 0.76 (2011) to 0.80 (2013) during the study period. Overall, the two-leaf light use efficiency model was capable of capturing spatial and temporal variations in gross primary production in China. However, the model needs further improvement to better simulate gross primary production in summer.