An Improved Covariate for Projecting Future Rainfall Extremes?

Abstract

Projection of extreme rainfall under climate change remains an area of considerable uncertainty. In the absence of geographically consistent simulations of extreme rainfall for the future, alternatives relying on physical relationships between a warmer atmosphere and its moisture carrying capacity are projected, scaling with a known atmospheric covariate. The most common atmospheric covariate adopted is surface air temperature, as it exhibits great consistency across climate model simulations into the future and, as per the Clausius‐Clapeyron relationship, has a well‐established link to atmospheric moisture capacity. However, empirical assessments of this relationship show that it varies with latitude, surface temperature, atmospheric temperature, and other factors, suggesting there may be more stable “global” atmospheric covariates that could be used instead. We argue that a better‐suited covariate would be one that captures the relationship between extreme rainfall and temperature but exhibits greater consistency in the relationship across regions as well as climatic zones. Our analysis identifies plausible atmospheric indicators of changes to future extreme rainfall, which now proliferate literature and compare their suitability based on the variability they exhibit across multiple geographical, topographic, and climatic zones within Australia. It is shown that surface air temperature exhibits a regionally inconsistent relationship with extreme rainfall and hence is not suitable for projecting to future conditions. The study identified integrated water vapor and surface dew point temperature as promising alternatives, with the former showing greater consistency in space but at the cost of reduced temporal coverage.

A Unified Framework for Extreme Subdaily Precipitation Frequency Analyses Based on Ordinary Events

“…Ordinary events are better resolved in climate models than extremes, and detection and understanding of trends and changes in ordinary events is easier due to the smaller inherent stochastic uncertainty. Changes in short‐duration extremes, or in any duration in general, can be thus included in the formulation by exploiting synoptic‐scale information from climate models (e.g., increased/decreased occurrence of precipitating synoptic conditions), covariates explaining extreme precipitation intensities (e.g., Roderick et al, 2020), and projected changes in intrastorm structure (e.g., Peleg et al, 2018; Wasko et al, 2016).…”

Resolving Inconsistencies in Extreme Precipitation‐Temperature Sensitivities

“…The use of atmospheric temperatures instead of surface temperatures has showed modest improvement (Ali & Mishra, 2017; Bui et al, 2019; Golroudbary et al, 2019). Alternatively, the relationship between atmospheric water vapor has also been examined (Neelin et al, 2009; Roderick et al, 2019; Schiro et al, 2016), resulting in sensitivities more consistent with climate model predictions of future rainfall intensification (Roderick et al, 2020). However, by far, the replacement of surface dry‐bulb temperature with surface dew point temperature is the most common variable substitution, with results shower greater consistency with the CC relation (Ali et al, 2018; Ali & Mishra, 2017; Barbero et al, 2018; Lenderink & van Meijgaard, 2010; Park & Min, 2017; Wasko et al, 2018).…”

Estimating design hydrologic extremes in a warming climate: alternatives, uncertainties and the way forward

“…There are also locations where the observed day-to-day variability of rainfall intensity to temperature is, in fact, negative [13,14], though the negative scaling appears to be an aberration (see [15]) forced by the unusual relationship temperature change exhibits with the occurrence of rainfall in tropical climates [16,17]. Observed increases in tropical precipitation extremes [18] have been confirmed in such locations using a more suitable warming surrogate (atmospheric moisture) instead of temperature [19,20], though large-scale circulation changes will affect this response regionally [21–23].…”

“…Questions also remain on whether a single covariate is sufficient to characterize the complex relationship warming poses with extreme rainfall, and whether the same covariate is useable for short duration rain (important for the design of urban stormwater systems) versus long-duration rain (important for the design of spillways of major water supply reservoirs), with different scaling rates observed for different rainfall mechanisms [22,47,50,51]. While there may be merit in using an atmospheric moisture-based covariate instead of temperature [19], the lack of long observational records across the world makes this a less attractive choice.…”