The SPIChanges R-package: Improving the interpretation of the standardized precipitation index under changing climate conditions

Abstract

The Standardized Precipitation Index (SPI) assumes the frequency of droughts remains constant over time. Given that climate change may violate this assumption, nonstationary versions of the SPI (NSPI) have been proposed. However, the NSPI presents a contradictory behavior, which may indicate fewer drought events under changes towards drier conditions. To overcome this issue, the SPIChanges package uses nonstationary distributions to quantify the effect of changes in precipitation patterns on the probability of SPI estimates. The package includes 16 candidate models that can describe a wide range of changes in precipitation patterns. The selected model calculates the probability of precipitation amounts under altered climate conditions and compares them with those of the SPI algorithm, revealing how the frequency of droughts has changed over time. The results of Monte Carlo simulations and case study applications allowed us to conclude that the package enhances the interpretation of SPI estimates under climate change conditions.