A New Method for Detecting Urban Impact on Precipitation and Application to Three Cities in Mainland <scp>China</scp>

ABSTRACT Observing urban effects on precipitation remains challenging, since the detected urban signal mainly relies on urban–rural differences and is inevitably confounded by background noise from non‐urban factors. To address this limitation, we propose a new method to detect the urban impact on precipitation and introduce a signal‐to‐noise ratio (SNR) index for the detected urban signal. For this aim, we classify non‐urban stations into interpolated and validation subsets. The urban signal is defined as the difference between urban stations and the interpolation field, while non‐urban noise is quantified by comparing validation stations to the interpolation field. Applying this method to three cities in mainland China based on the station data, we observe different urban impacts on extreme precipitation across cities. In Shanghai city, the detected urban signals exhibit high SNR. Our study highlights that introducing the SNR provides an effective means to quantify the reliability of conclusions regarding urban impacts on precipitation. Moreover, the framework of this method is adaptable for assessing the urban impact on other spatially continuous meteorological variables, such as temperature and humidity.