CLIMACS: A method for stochastic generation of continuous climate projected point rainfall for urban drainage design

Climate projected continuous rain series are required for urban drainage design and analysis. Outputs from regional climate models are yet insufficient in quality and resolution for this purpose. Here, we introduce a novel method, CLIMACS (CLImate projection of MeAsured preCipitation Series) for stochastic climate projection of point rain series for urban drainage design. It is developed and evaluated to represent current climate conditions as well as projections for the period 2071–2100 with RCP scenarios 4.5 and 8.5. CLIMACS includes seasonal stochastic resampling of individual rain events and rainfall intermittency as well as climate scaling. Fifteen climate variables representing multiple changes in the future rainfall are selected and projected to target future conditions. Realizations of the resampled and stochastically generated rain series are ranked and selected according to the minimum relative error between realization and target. To give an insight into the uncertainty of the climate projection as well as year-to-year variability of precipitation climate variables, an ensemble of the top 100 realizations for each climate scenario is selected for further analysis. It is concluded that the ensembles represent the future conditions well, however, with a large variability due to the uncertainties in climate projection. With the aim of showing the impact in urban drainage system design, a stormwater detention pond design with multiple design parameters and climate scenarios is demonstrated. The pond design results show a significant difference in required pond volume depending on the design parameters and the choice of climate scenario, which emphasizes the need for climate projected continuous rain series for urban design purposes as an alternative to design storms.