COS 273-2 - Global Photosynthesis Response to Rainfall Frequency and Magnitude: Contrasting Dryland and Mesic Ecosystems

Abstract: Rain events are becoming fewer, but stronger under a changing climate. These intra-seasonal rainfall features have received less attention than changes in annual temperature and rainfall in the context of how they influence plants and the carbon cycle. However, field rainfall manipulation experiments show notable 10%-20% changes to annual plant function in response to rainfall frequency alterations, while holding annual rainfall amount constant. Nevertheless, field and modeling experiments show little consensus on the sign and magnitude of annual mean plant response to changing rainfall frequency. In this study, we ask: based on satellite observations, how is global photosynthesis responding to changes in rainfall frequency? How and why do photosynthesis responses vary across drylands and mesic ecosystems?



Using linear and panel regression techniques, we isolate the effects of rainfall frequency on annual gross primary production (GPP) while controlling for the effects of total annual rainfall. We perform these methods on several global satellite observation-based proxies of GPP including solar induced fluorescence, vegetation optical depth, and normalized difference vegetation index. We additionally validate our results with flux tower field sites and compare our results with previous field and modeling studies in how drylands and mesic ecosystem responses differ.

We find that the photosynthesis response to rainfall frequency is as high in magnitude and spatial extent as its response to total annual rainfall across the globe. However, the satellite-based photosynthesis proxies and field tower sites indicate that dryland and mesic environments respond differently to less frequent rainfall. With fewer, stronger storms, dryland GPP increases while mesic ecosystem GPP decreases. These results largely agree with our review of previous modeling and field manipulation studies. We find that the difference of the responses is due to lower sand fraction and stronger plant response magnitude to individual rain pulses in drylands, which appear to favor GPP increases under stronger storms. By contrast, mesic ecosystems show widespread increases in mean atmospheric aridity response with fewer, stronger storms, which appear to drive GPP decreases with longer interstorm periods. Overall, changing rainfall frequency will result in average annual photosynthesis changes of 10-20% over the next century over at least a third of global vegetated surfaces, with the largest relative changes in drylands. Ultimately, our results indicate that intra-annual rainfall variability drives global annual photosynthesis variability similarly to inter-annual rainfall variability.