Stationarity assessment of hydroclimatic variables for a tropical river basin

Understanding the space-time variations and dependency between hydroclimatic variables is vital for predicting future changes towards adaptive water resources management in a changing climate. Globally, it is observed that the mean and extreme hydroclimatic conditions are experiencing a significant shift under a changing climate, which affects the spatio-temporal distribution of floods and droughts. However, climate change studies seldom talk about the time invariance of the characteristics of a hydroclimatic time series. This research assesses the time invariance of the statistical properties of hydroclimatic variables (such as rainfall, temperature, streamflow and their derived indices) for a tropical river basin (i.e., the Tapi River basin) in India. Climate change profoundly impacts tropical river systems. Hence, assessing and detecting stationarity in hydrologic processes for such a river basin is imperative to predicting future changes. In this study, we have analyzed nine hydroclimatic variables representing the mean and extreme rainfall, temperature and streamflow. The hydroclimatic indices have been statistically examined to detect stationarity, homogeneity, and trends. A non-overlapping block stratified random sampling approach has been applied to identify the time invariance of hydroclimatic indices. The stationarity assessment approach investigates the similarities in the median, variance, distribution, and statistical moments of the continuous time series data. Based on the results of this study, weak and strict stationarity can be identified. The findings have significant ramifications for the planning and design of hydraulic structures, stormwater networks, flood mitigation, and disaster management for the Tapi basin.