An analysis suggests that high carbon uptake by US land ecosystems during the warm spring of 2012 offset the carbon loss that resulted from severe drought over the summer — and hints that the warm spring could have worsened the drought.
Warmer springs and drier summers are an expected consequence of climate change1. Warmer springs should increase the carbon uptake of terrestrial ecosystems by lengthening the growing season, whereas drier summers should reduce uptake because of poor plant growth, especially in drought years. In 2012, the continental United States had the warmest spring on record, and one of the worst summer droughts in decades. What did these extremes do to the land carbon budget? The answer matters because terrestrial carbon uptake helps to remove anthropogenic carbon dioxide emissions from the atmosphere. Writing in Proceedings of the National Academy of Sciences, Wolf et al.2 conclude that the increased carbon uptake during the spring essentially offset the carbon lost during the summer — although the details of this phenomenon are rather complex.
The effects of interactions between spring warming and summer drought on carbon budgets at continental and local scales have been reported previously3,4, but it is only in the past few years that multiple data sources with which to evaluate large-scale climate effects and their local variations have become widely available. The authors arrived at their conclusions by comparing three data sets: eddy-covariance data that measure carbon exchange between the lowest part of the atmosphere (the boundary layer) and land biospheres over areas of approximately 1 square kilometre, gathered by 22 towers scattered across the United States; satellite estimates of the timing of plant growth; and regional carbon-budget estimates from CarbonTracker, a modelling system that uses observations of atmospheric CO2 levels and gradients to infer surface fluxes of the gas over land. So what do the data show?
The severe drought that occurred during the summer of 2012 encompassed more than half of the continental United States, with most of the affected regions falling into the two worst categories as defined by the US Drought Monitor (extreme and exceptional)5. Accordingly, most of the towers reported a loss of carbon from their sites during this period, and recorded that the annual carbon budgets did not balance. Meanwhile, CarbonTracker suggested that carbon gain during the spring (0.24 petagrams of carbon; 1 Pg is 1015 grams) and carbon loss during the summer (0.23 Pg) were almost equal for the continental United States as a whole.
However, there was considerable variability within that picture. Eastern temperate forests (Fig. 1a) vigorously sequestered carbon during the spring, and this carbon gain (0.18 Pg) slightly more than offset the summer carbon loss (0.16 Pg) from the Great Plains (Fig. 1b) — the area most affected by drought, and which accumulated significantly less carbon than in an average year. Overall, carbon uptake for the lands of the continental United States had increased, rather than reduced, by the end of the year (a rise of 0.11 Pg C yr−1), with the surplus resulting from increased carbon uptake during the autumn.
Wolf and colleagues propose that the spring warming and summer drought were physically coupled through interactions between the land surface and atmosphere. Simply put, ecosystems entered the summer with a relative water deficit because water was used up earlier than normal during the warmer spring. The deficit led to a reduction in evaporative cooling, which increased the effects of summer heating, causing water stress.
The authors go on to suggest that early warming might even have reinforced weather patterns, increasing the probability or the severity of summer drought. Confirming this will require a more comprehensive analysis and diagnosis, including measurements from more eddy-covariance towers, but is well within the realm of possibility. Clear evidence of such a link would undoubtedly help the public, policy-makers and resource managers to prepare strategies for adapting to droughts in the future.
A strength of Wolf and co-workers' study is that it combines in situ eddy-covariance measurements, atmospheric observations and remote-sensing data. The eddy-covariance data provide the most direct evidence for seasonal changes in terrestrial carbon uptake, and are the only data that directly constrain the CarbonTracker and satellite estimates, by quantifying both the carbon flux and the full energy balance of water–temperature interactions. The remote-sensing data provide the best insight into the timing of biological activity across the continent, whereas the atmospheric analyses allow the local fluxes and processes to be understood in the context of the overall carbon budget. In the future, a more sophisticated synthesis of the different data will greatly improve the accuracy of analyses of carbon and water exchange between the land and atmosphere.
A limitation of the study is that the tower sites weren't specifically placed to sample the dominant carbon-flux anomalies that were revealed by CarbonTracker and the satellite data. For instance, the largest region of springtime carbon-uptake anomalies occurred in the southeastern United States, where there are no flux towers. The largest region of midsummer carbon-loss anomalies occurred in the Great Plains, where the two sites used in the study represent grasslands, rather than the dominant agricultural landscapes of this region.
In addition, the current tower network isn't dense enough to cover climate events such as the extreme year of 2012. A facility called the National Ecological Observatory Network (with which one of us, D.S., was associated for several years), designed to sample climate conditions optimally, will come online in the next few years6 and provide uniform coverage of the continental United States. Climatologists have long designed networks to study spatial patterns, whereas ecologists have tended to rely on local field studies and extrapolated their findings to larger areas on the basis of vegetation types or other classifications. A reference network that covers all spatial components and biomes is essential for this type of extrapolation in future studies.
Wolf and colleagues' work shows how important systematic, continental-scale sampling is, because no one site — and not even several sites — could tell the entire story of a perturbation such as the one that occurred in 2012. As ecologists attempt to understand problems at ever larger scales, they will increasingly direct their creative energies towards problems that require massively more data than individual research laboratories can collect. Information obtained from infrastructural monitoring systems and openly available data will therefore have a crucial role in advancing the science of climate impacts, as they already do in other disciplines. Footnote 1
Melillo, J. M., Richmond, T. C. & Yohe, G. W. (eds) Climate Change Impacts in the United States: The Third National Climate Assessment http://dx.doi.org/10.7930/J0Z31WJ2 (2014).
Wolf, S. et al. Proc. Natl Acad. Sci. USA 113, 5880–5885 (2016).
Angert, A. et al. Proc. Natl Acad. Sci. USA 102, 10823–10827 (2005).
Hu, J., Moore, D. J. P., Burns, S. P. & Monson, R. K. Glob. Change Biol. 16, 771–783 (2010).
Rippey, B. R. Weath. Clim. Extremes 10, 57–64 (2015).
Keller, M., Schimel, D. S., Hargrove, W. W., Hoffman, F. M. Front. Ecol. Environ. 6, 282–284 (2008).
Related links in Nature Research
About this article
Stabilization of carbon sequestration in a Chinese desert steppe benefits from increased temperatures and from precipitation outside the growing season
Science of The Total Environment (2019)
Spatiotemporal Consistency of Four Gross Primary Production Products and Solar-Induced Chlorophyll Fluorescence in Response to Climate Extremes Across CONUS in 2012
Journal of Geophysical Research: Biogeosciences (2018)
Tracking the phenology and expansion of Spartina alterniflora coastal wetland by time series MODIS and Landsat images
Multimedia Tools and Applications (2018)