Decline in total precipitation volumes and increase in the intensity, duration and frequencies of extreme precipitation events pose surmounting socioeconomic and engineering challenges. While recent studies have reported that there is threefold increase in extreme precipitation events over India in 20th century, the implications of such extremes on built systems such as transportation networks, natural systems such as ecological networks and societal systems are seldom studied. The statistics obtained from extreme daily precipitation such as 100–year return level (RL 100) and 30-year return level (RL 30) is vital for designing and managing the infrastructure systems. While these infrastructure systems are designed for T year return period, the magnitude of T-year return period is likely to change due to non-stationary features (either natural or anthropogenic) which is observed in various hydrometeorological processes. The hypothesis of this study is to detect the non-stationarity in statistics of extreme precipitation for homogeneous zones of India. We analyze the trends in Precipitation Extremes Volatility Index (PEVI), which is a stability indicator for the return periods. Higher values of PEVI indicated higher instability of return period. Trends in PEVI is analyzed using 40-year moving windows for each grid in all five homogeneous regions (North-Western India, North Central India, East Peninsular India, Western Peninsular India and Southern Peninsular India) and risk assessment of critical infrastructure facilities with specific focus on transportation networks is performed. Our findings suggest that nearly 50% of grid-points across the nation exhibit significant volatility posing new challenges to infrastructure managers in the management of such extremes. We motivate a network science based framework that can examine risk and suggest recovery strategies for networked systems under the scenario of non-stationary PEVI.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner