A CONUS-wide assessment of the climate change impact on low probability precipitation events: Combining information from the past and scaling arguments to estimate future trend
Abstract
The protection and design of critical infrastructure, as well as the characterization of flood risk, require accurate knowledge of current and future hydroclimatic conditions. Consequently, in view of the rapidly changing climate, there is an urgent need to acquire accurate information on the evolution of low probability precipitation events. Various research efforts have investigated the impacts of climate change on the frequency of intense rainfall, mostly by fitting different theoretical distribution models to data with parameters that vary (in most cases linearly) with time. However, the increase of model parameters increases also the uncertainty of the estimated quantiles, which becomes particularly important for low probability events. Furthermore, long-term climate model projections can be quite inaccurate, thus their use to describe the frequency of extreme events may not be conclusive due to epistemic uncertainties. In this study, we utilize the high-resolution hourly precipitation product developed by Emmanouil et al. (2021), which spans back to 1979, in order to investigate the influence of climate change on the evolution of low probability rainfall events. To do so, we apply the parametric multifractal (MF) approximation developed in Langousis et al. (2009), which has been proven robust in describing the intensity and frequency of extreme rainfall from small rainfall samples (Emmanouil et al., 2020). More precisely, we: a) split the continuous hourly rainfall timeseries into sequential 10-year segments, where climate conditions can be assumed stationary; b) use each segment from (a) to fit a stationary multifractal model; c) use the fitted models from (b) and the MF approximation to estimate the rainfall intensity at various return levels (i.e. return periods T from 2 years to more than 100 years), and d) use the results from (c) to estimate the trend magnitude in the intensity of precipitation extremes at various return levels over the considered 40-year period. The obtained results show that the effects of climate change on rainfall extremes may severely impact the current infrastructure, as they vary spatially over CONUS, are significantly influenced by the topography and rainfall climatology of each region, while depending on the duration d of rainfall events and the return period T of interest.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFM.H33I..08E