The contribution of internal climate variability to climate change impacts on droughts.

The assessment of climate change impacts is usually done by calculating the change in drought conditions between future and historical periods by using multiple climate model simulations. However, this approach usually focuses on anthropogenic climate changes (ACCs) while ignoring the internal climate variability (ICV) caused by the chaotic nature of the climate system. Recent studies have shown that ICV plays an important role in the projected future climate change. To evaluate that role, this study quantifies the contribution of ICV to climate change impacts on regional droughts by using the signal-to-noise ratio (SNR) and the fraction of standard deviation (FOSD) as metrics for China. The internal climate variability or noise (i.e. ICV) is estimated as the inter-member variability of two climate models' large-member ensembles; the signal (i.e. ACC) and the climate model uncertainty (or inter-model uncertainty, IMU) are estimated as the ensemble mean and inter-model variability of 29 global climate models, respectively. The drought conditions are characterized by drought frequency, duration and severity, which are quantified by using the theory of run based on the standardized precipitation evapotranspiration index (SPEI). The results show that deteriorated drought conditions induced by ACCs are projected to occur over China. From the perspective of the SNR, the ICV impacts are less significant compared to the ACC impacts for drought metrics. Remarkable spatial variations of SNRs for future drought metrics are found, with values varying from 0.001 to exceeding 10. In terms of the FOSD, ICV contributions relative to the IMU are large, as FOSDs are >1 for around 22% grids. These results imply the significance of taking into account the impacts of ICV in drought assessment, any study ignores the influence of ICV may be biased.