Analysis of precipitation dynamics at different timescales based on entropy theory: an application to the State of Ceará, Brazil

Abstract

Water resource variables are highly complex and vary both spatially and temporally. Understanding the variability and how it evolves has been an important scientific question in Ceara ́, Brazil. However, describing and determining the uncertainty and the variability in precipitation is still a challenge. Assessing the uncertainty around precipitation is key to develop robust and proactive planning. This study’s main aim is to evaluate the underlying spatiotemporal variability of precipitation in the State of Ceara ́ at different timescales by using standardized variability indices computed from different entropy measures. This methodology was applied to analyze 31 meteorological stations with daily time series from 1962 through 2006 while expanding the analysis to the remaining region using an interpolation method. The seasonal timescale analysis revealed that the dry season contributes more to the annual variability, and the change in intra-annual precipitation dynamics could vary with timescales. There were significant upward trends in entropy. Thus, for some stations, there was an increase in the uncertainty of rainfall. Also, there was an increase in variability amount and intensity throughout the decades at the monthly and seasonal timescales. Assessment of precipitation uncertainty within different timescales can benefit a broad community of scientists who are interested in arid-region and natural hazards.