Datensatz

Zusammenfassung

Combined droughts and heatwaves are among those compound extreme events that induce severe impacts on the terrestrial biosphere and human health. A record breaking hot and dry compound event hit western Russia in summer 2010 (Russian heatwave, RHW). Events of this kind are typically studied either from a hydrometeorological perspective, or with a focus on impacts in the terrestrial biosphere such as reductions of the terrestrial carbon storage. These different perspectives might not only require different strategies for event detection, but also change interpretations and impact assessment. To exemplify this issue, we revisit the RHW both from a biospheric and a hydrometeorological perspective. We consider several hydrometeorological and biospheric variables agnostically as inputs to a recently developed multivariate anomaly detection approach. Our analysis of biospheric variables reveals that the RHW was preceded by increased gross ecosystem production in spring that partly compensated the reduced summer production, but remained unconsidered in earlier impact oriented studies. We also find that the region of reduced summer ecosystem production does not match the area identified as extreme in the hydrometeorological variables. The reason is that forest-dominated ecosystems in the higher latitudes respond with unusually high productivity to the RHW, leading overall to a compensation of 54 % (36 % in spring, 18 % in summer) of the reduced gross primary production (GPP) in southern agriculturally dominated ecosystems. Our results show that an ecosystem-specific and multivariate perspective on extreme events can reveal multiple facets of extreme events by simultaneously integrating several data streams irrespective of impact direction and the variables’ domain (here "biosphere" or "hydrometeorology"). Focusing on negative impacts in specific variables e.g. a vegetation index, leads to a spatiotemporally delineation of extreme events that is inconsistent with the hydrometeorological conditions and and can limit the interpretation of their impacts on the terrestrial biosphere. Our study exemplifies the need for robust multivariate analytic approaches to detect extreme events in both hydrometeorological conditions and associated biosphere responses to fully characterize the effects of extremes, including possible compensatory effects in space and time.