Relationship between Surface Temperature and Extreme Rainfalls: A Multi-Time-Scale and Event-Based Analysis

Recent studies have examined the relationship between the intensity of extreme rainfall and temperature. Two main reasons justify this interest. First, the moisture-holding capacity of the atmosphere is governed by the Clausius-Clapeyron (CC) equation. Second, the temperature dependence of extreme-intensity rainfalls should follow a similar relationship assuming relative humidity remains constant and extreme rainfalls are driven by the actual water content of the atmosphere. The relationship between extreme rainfall intensity and air temperature (P-extr-T-a) was assessed by analyzing maximum daily rainfall intensities for durations ranging from 5 min to 12 h for more than 100 meteorological stations across Canada. Different factors that could influence this relationship have been analyzed. It appears that the duration and the climatic region have a strong influence on this relationship. For short durations, the P-extr-T-a relationship is close to the CC scaling for coastal regions while a super-CC scaling followed by an upper limit is observed for inland regions. As the duration increases, the slope of the relationship P-extr-T-a decreases for all regions. The shape of the P-extr-T-a curve is not sensitive to the percentile or season. Complementary analyses have been carried out to understand the departures from the expected Clausius Clapeyron scaling. The relationship between dewpoint temperature and extreme rainfall intensity shows that the relative humidity is a limiting factor for inland regions, but not for coastal regions. Using hourly rainfall series, an event-based analysis is proposed in order to understand other deviations (super-CC, sub-CC, and monotonic decrease). The analyses suggest that the observed scaling is primarily due to the rainfall event dynamic.