Precipitation Variability

Hyrologic Regionalization using wavelet entropy

The concept of entropy, when applied in conjunction with wavelet analysis, can be used to determine the randomness (i.e. level of uncertainty) of a time series at different timescales. At a given scale, maximum entropy is possible when the information is evenly spread across time, and minimum entropy occurs when all the information is contained in a single location. The Wavelet based Multiscale Entropy (WME), which is a measure of the degree of order/disorder of the signal and carries information associated with multi-frequency signal, can provide useful information about the underlying dynamic processes associated with the signal and can help in regionalization studies. This provides the motivation to develop a robust regionalization tool based on WME.

Understanding the variability in natural processes using entropy.

Understanding the spatiotemporal variability of rainfall is vital for water resources planning and management, flood and drought mitigation, and erosion control, among others. Despite the progress in this direction, through proposal of many different approaches and their applications to rainfall data at various regions around the world, our knowledge of the spatiotemporal variability of rainfall remains limited. Studies in this direction have focused mainly on the amount of rainfall and its spatial patterns, and investigations of spatiotemporal variability at multiscale are limited. In our group, we introduce a novel measure, Standardized Variability Index (SVI), based on the concept of entropy to investigate the spatiotemporal variability of gridded rainfall in the Indian subcontinent at different timescales.

Further, we have developed a novel method to estimate the predictability of time series using wavelet entropy concept and to evaluate the performance of the hydrological models.