b'Global EcologyContinued34New Way to Assess. . .SIF outperformsCarbon Uptake Patterns existing vegetation indicators Yoichi Shiga, Anna Michalak, Joe Berry, andand a majority of sophisticated colleagues used a dense network of atmosphericterrestrial biosphere models carbon dioxide (CO 2 ) observations to provide, forintracking net CO 2the first time, a regional scale evaluation of a new satellite-based measurement called solar-inducedexchange. . . chlorophyll fluorescence (SIF). SIF is a photosynthetic by-product, which occurs as sunlight excites chlorophyll during photosynthesis. SIF offers theUsing three years of data from the dense North potential to track photosynthesis from space withAmerican network of atmospheric CO 2observation unprecedented resolution and coverage. To date,towers, the researchers found that SIF outperforms analysis of SIF had been limited to either plot (~1existing vegetation indicators and a majority of square kilometer) or global/hemispheric scales.sophisticated terrestrial biosphere models in tracking Regional to continental scales critical for policynet CO 2exchange by exhibiting regional patterns more decisions and informing carbon-climate feedbacksconsistent with the atmospheric data. remained unexplored, until now.They used an atmospheric transport model to relate the patterns of different surface variables to the a fluctuations observed in the atmospheric CO 2data. To identify the impact of SIF on carbon flux patterns, they compared two sets of inverse-modeling estimates of carbon uptake, one with SIF and one without SIF. A new satellite-based measurement called solar-induced chlorophyll fluorescence (SIF) measures a photosynthetic by-product, which occurs as sunlight excites chlorophyll during photosynthesis. When SIF information is included in observations, there is a redistribution of the growing season carbon dioxide uptake from needleleaf forests to croplands, with green indicating a stronger net carbon uptake and brown a weaker net carbon uptake. The difference between June, July, and August flux estimates are obtained from inverse models with and without SIF, averaged over three years (2008-2010). Inverse modeling 0.50 0.25 0 0.25 0.50 starts with the results and then calculates the causes. [mol m -2s -1 ] Image courtesy Yoichi Shiga'