“…Noticeably, breakpoints within the time series of diurnal temperature range were detected at~176 sites, of which the majority (93%) occurred before the periods with flux measurements (Figures S5a-S5c and S5e-S5g).Most of the aforementioned region years with climatic anomalies and extremes were documented in previous studies, such as large-scale droughts(e.g., 2003(e.g., Europe, 2003(e.g., -2004 western U.S., 2011 southern U.S., and 2012 Midwest U.S.)[Ciais et al, 2005;Coumou and Rahmstorf, 2012;Mallya et al, 2013;Parazoo et al, 2015;Schwalm et al, 2012;Wolf et al, 2016], heat waves (e.g., 2003 summer Europe and 2012 North America)[Ault et al, 2013;Ciais et al, 2005;Karl et al, 2012;Wuebbles et al, 2014], and high precipitation causing flooding (e.g., 2010-2011 Australia and 2007 summer western Europe)[Bastos et al, 2013;Coumou and Rahmstorf, 2012;Haverd et al, 2013]. To the best of our knowledge, only a few of those extreme events have been studied extensively for their effects on CO 2 and energy fluxes across sites and regions [e.g.,Bastos et al, 2013;Chu et al, 2016;Ciais et al, 2005;Schwalm et al, 2012;Wolf et al, 2013Wolf et al, , 2016Zscheischler et al, 2014]. These "hot spots" (region years) of climate extremes pointed out potential areas of future region-wide syntheses by using the new FLUXNET data set.…”