Scaling Relationships between Extreme Precipitation and Local Temperature: Contrasting for Binning Scaling and Trend Scaling

Trend scaling relationships between extreme precipitation and temperature are often used to represent the influence of long-term warming on the intensity of extreme precipitation. Indeed, such scaling relationships are often regarded as providing more reliable precipitation projections than direct projection, owing to higher confidence for temperature projections in model simulations. Due to limited data availability, especially for the sub-daily rainfall, so-called binning scaling relationships, which relate extreme precipitation to temperature at the time of occurrence and are estimated empirically either through a binning technique or quantile regression, have been considered as a substitute for trend scaling to project the long-term response of local extreme precipitation to temperature change (Lenderink and Van Meijgaard, 2008; Wasko and Sharma, 2014). Estimates of binning scaling rates are generally based on seasonal subdaily precipitation observations, and thus they are influenced by factors other than temperature that change systematically within a season, synchronously with the seasonal cycle (Zhang et al., 2017). In contrast to trend scaling, binning scaling often suggests faster than Clausius-Clapeyron intensification of sub-daily precipitation extremes with temperature.

We explore this apparent contradiction between binning and trend scaling using a large ensemble of moderate resolution regional climate simulations for North America. The large amount of data that is available from this ensemble allows us to confidently estimate both trend and binning scaling rates for the climate that is simulated by that model. Specifically, we use a 35-member initial conditions ensemble of regional climate simulations produced with the Canadian CanRCM4 regional climate model for the period 1950-2100, with historical forcings for the period ending 2005 and RCP8.5 forcing subsequently. Each CanRCM4 ensemble member was driven by a corresponding member of a similar large ensemble of global simulations produced with the Canadian global Earth system model CanESM2 (Scinocca et al., 2016).

We compare binning and trend scaling of precipitation extremes across different durations (1-hour, 3-hour, and 24-hour), considering annual and seasonal values, and both local and regional spatial scales. We provide strong evidence to clarify that binning scaling cannot project the long-term change in precipitation extreme, with substantial disagreement in the spatial pattern and magnitude of scaling rates between binning and trend scaling regardless of the duration, season, and spatial scale. Using the daily dew point temperature as scaling variable rather than dry air temperature does not eliminate the differences between binning and trend scaling rates. While shorter-duration extreme precipitation does appear to intensify faster with warming in CanRCM4, we only find super-adiabatic intensification of annual precipitation extremes in isolated regions regardless of accumulation durations. Compared with annual maximum results, winter extremes intensify more strongly over the western and southeastern North America across all timescales. A decreasing tendency of summer extremes is projected over the north and central Great Plains. The seasonal timing of the occurrences of precipitation extremes are expected to shift towards the cold season, reflecting the different changing tendencies in summer and winter extremes.

Lenderink, G., and Van Meijgaard, E. 2008: Increase in hourly precipitation extremes beyond expectations from temperature changes. Nat. Geosci., 1, https://doi.org/10.1038/ngeo262.

Scinocca, J. F., Kharin, V. V., Jiao, Y., Qian, M. W., Lazare, M., Solheim, L., and Flato. G. M., 2016: Coordinated Global and Regional Climate Modeling. J. Climate, 29, 17-35, https://doi.org/10.1175/Jcli-D-15-0161.1.

Wasko, C., and Sharma, A. 2014: Quantile regression for investigating scaling of extreme precipitation with temperature. Water Resour. Res., 50, 3608-3614, https://doi.org/10.1002/2013WR015194.

Zhang, X. B., Zwiers F. W., Li, G. L., Wan, H., Cannon, A. J., 2017: Complexity in estimating past and future extreme short-duration rainfall. Nat. Geosci., 10, 255-239, https://doi.org/10.1038/NGEO2911.