Spatial-temporal estimation of maximum temperature high returns periods for annual time series considering stationary/nonstationary approaches in Iran urban area

Abstract

The maximum temperature in the annual time series is a variable with a broad impact on heatstroke, energy consumption, cooling system stability, peak water consumption, and cardiovascular patients. Due to climate change, variations in the high return period of maximum temperature in the annual time series are expected. Due to the significant impact of the high return period of maximum temperature in the annual series in urban areas, it is necessary to study its Spatial-Temporal variations in these areas. In this study, using long-term data from 41 urban areas in Iran, the Generalized additive Models for Location, Scale, and Shape (GAMLSS) have been used to estimate changes in the return period of 100 and 50 of maximum temperatures. Findings show that in 83% of urban areas in Iran, the nonstationarity in the maximum temperature time series is statistically significant. Ignoring the nonstationarity has led to underestimating the maximum temperature in the high return period up to 2.6 °C and overestimating up to 7 °C. Also, the conventional stationary approach in most Iran urban areas has led to underestimating the hazard of annual maximum temperature. Compared to other areas, the western and southern urban areas of Iran are more clearly affected by the nonstationarity of maximum temperature in the annual time series, which means that the related designs in these areas should be corrected.