COS 138-2 - Observed Global Photosynthesis Response to Changing Rainfall Frequency and Intensity

Background/Question/Methods

Rain events are becoming more intense and less frequent under a changing climate. These intra-seasonal rainfall features have received less attention than changes in annual mean temperature and total rainfall in the context of how they influence global ecosystems. However, field manipulation experiments consistently show non-negligible (10%-20%) changes to annual gross primary production (GPP), even after controlling for changes in total annual rainfall amounts. Nevertheless, after at least twenty years of field and modeling experiments, there is little consensus on the sign change of annual GPP due to changing rainfall frequency and intensity. In this study we ask: based on available global satellite observations, how are the water and carbon cycles responding to changes in rainfall frequency and intensity? What is driving the magnitude and sign of annual GPP response to these rainfall characteristics?

Using linear and panel regression techniques, we isolate the effects of rainfall intensity and frequency on annual GPP by controlling for total annual rainfall effects. We perform these methods on several global satellite observations of vegetation photosynthesis including annual averaged OCO-2 solar induced fluorescence, SMAP vegetation optical depth, MODIS normalized difference vegetation index, and FluxSat gross primary productivity as well as rainfall observations (CPC, GPM).

Results/Conclusions

GPP in humid ecosystems (mean annual precipitation (MAP) >500mm) tend to respond negatively to less frequent, more intense rainfall. These responses appear to be related to a soil response rather than a direct vegetation response: mean annual soil moisture tends to decrease with less frequent, more intense storms in these humid environments. Rain pulse characteristics are therefore likely altering mean soil moisture availability in these ecosystems rather than directly impacting GPP responses. By contrast, dryland ecosystems (MAP< 500mm) appear most sensitive to rainfall frequency and intensity but show both positive and negative signs of GPP response to these rainfall characteristics across the globe. We show how these diverging responses in drylands appear to be due to complex vegetation responses to individual rain pulses. Namely, this includes plant species differences in their strategies of pulse plant water uptake and dry period survival. With global rainfall becoming more intense and less frequent, we ultimately find that increasing rainfall intensity and decreasing frequency will decrease global mean annual GPP. However, there is uncertainty in these changes given the large absolute magnitude and differences in trends in global drylands.