Attribution of temperature changes in Western China

Annual mean temperature anomalies (°C, relative to 1961–1990) for the WC region from observations (black line) during 1958–2015 and multi‐model ensembles under ALL (red line), GHG (green line), ANT (brown line) and NAT (blue line) forcings during 1958–2012. Pink and blue shadings show the 5–95% ranges of the individual model simulations from ALL and NAT experiments, respectively. ABSTRACT Western China (WC) is located in the arid and semi‐arid belt of the mid‐latitudes. Understanding the causes of the large temperature increases experienced in this region is of great importance because both the fragile ecological environment and the human societies in WC are highly vulnerable to changes in climate. We conducted a detection and attribution analysis on the annual mean temperature using an optimal fingerprinting method by comparing observations based on homogenized station data with simulations conducted using the models that participated in the Coupled Model Intercomparison Project Phase 5 (CMIP5). The results show that the increase in the mean temperature can be attributed to human activities. The effect of anthropogenic (ANT) forcing can be detected in single‐signal detection analyses when the response to the ANT forcing is compared directly with the observations. This effect can also be separated from the influence of the natural external forcing in two‐signal analyses. The warming attributable to the ANT forcing explains most of the observed regional warming. The models appear to have underestimated the observed warming, suggesting that the projected future temperature increases based on raw output from the model simulations may have also been underestimated.