资源环境科技发展态势分析平台

Geiger et al (Environ. Res. Lett. 2016 11 084012) employ two functional relationships to characterize hurricane damage in the USA-either based on GDP (one exponent) or on per capita GDP and population (two exponents). From the Akaike Information Criterion the authors cannot reject the former kind in favor of the latter. The different approaches, however, lead to divergent projections of future hurricane losses. In this comment, we argue that there is no rigorous evidence in [1] to give preference to one or the other approach, and the conclusion that high-income does not protect against hurricane losses needs to be revisited. As a perspective, it needs to be mentioned that the previously published relationship between GDP and population could unify both approaches.