Abstract

In recent decades, the impact of climate change on urban flooding has increased, along with an increase in urban population and urban areas. Hence, historical design storms require revisions based on robust intensity–duration–frequency (IDF) relationships. To this end, the development of an urban rain-gauge network is essential to yield the spatiotemporal attributes of rainfall. The present study addresses two objectives: (a) to reconstruct sub-daily rainfall time series for the historical period over an urban gauge network, and (b) to investigate the spatiotemporal variation in extreme rainfall distribution within a city. This study considers Bangalore, India, where rainfall has been historically monitored by two stations but a dense gauge network has recently been developed. The study applies random forest regression for rainfall reconstruction, finding that the performance of the model is better when the predictand and predictor stations are near to one another. Robust IDF relationships confirm significant spatial variations in extreme rainfall distribution at the station and the city-region levels. The areal reduction factor (ARF) is also estimated in order to understand the likely impact of the reconstructed time series on hydrological modeling. A significant decrease in the ARF is observed as the area grows beyond 450 km2, indicating a substantial reduction in the volume of the design floods.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call