This study investigates the nonstationary behavior of extreme temperatures in Pakistan across six diverse climatic zones over two distinct periods (1951–1985 and 1986–2020). The nonstationarity assessment is done by incorporating time and antecedent soil moisture (AMS) levels as nonstationarity covariates in Generalized Extreme Value (GEV) distribution. Results show signifcant intra-zonal, intra-seasonal, and intra-periodic variabilities in nonstationarity impacts on extreme temperatures. With few exceptions, maximum temperatures exhibited an increase in annual return levels particularly for second period while annual minimum temperature extreme gave mixed pattern during both periods. Seasonal analysis reveals notable intraperiod variabilities, with most zones shows higher increase in maximum temperature return levels during second period up to+2.3 °C and+0.91 °C for spring and summer respectively. Conversely, minimum temperature analysis shows higher increase in return levels during spring (up to+2.47 °C) and summer (up to+2.02 °C) across most of the zones with decreased (up to -1.16 °C) return levels in central and eastern-western regions during winter. Except for two coastal zones, increased soil moisture levels were associated with signifcant reductions in maximum temperatures across most zones, with zonal average decrease in maximum temperature ranging from -0.01 °C to -0.05 °C per 1% increase in antecedent soil moisture. Positive association between minimum temperature and antecedent soil moisture has been found with zonal average range of+0.002 °C to+0.277 °C per 1% increase in antecedent soil moisture. The fndings of this study have important implications for climate modeling, risk assessment, and adaptation strategies in the region for improved climate change assessmen
