Evaluation of change factor-based statistical downscaling methods for impact analysis in urban hydrology

ABSTRACT

Climate change impact analysis in urban hydrology involves downscaling of coarse climate model outputs. This study evaluates the downscaling skill of four change factor methods for impact analysis on urban hydrology in Belgium. The downscaling methods are applied to precipitation observations. For this, the 100-year Uccle time series is split in observations, in pseudo-climate model runs and in an evaluation period. Conceptual models for a retention basin and sewer system are thereafter forced with the downscaled time series and the time series for the evaluation period. The downscaling skill is determined based on the reproduction of precipitation and impact statistics. Results show that the skill depends on the stormwater system and the impact variable. Due to climate variability, none of the methods is found outperforming. It is therefore recommended to apply a more thorough analysis, which also employs the long-term projections.

KEYWORDS:

• Urban stormwater systems
• overflow events
• perfect predictor experiment
• skill
• validation
• assumptions

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

E. Van Uytven was supported by the Research Foundation-Flanders, Fonds Wetenschappelijk Onderzoek - Vlaanderen (FWO) under Grant [11ZY418N].