Dependence of Precipitation Extremes on Temperature over United States

Abstract:

Hydrologic disturbances are commonly associated with the phenomenal occurrence of extreme events. The human kind has always been facing problem with hydrologic extremes in terms of deaths and economic loss. Hence, a complete analysis of observed extreme events will have a substantial role in planning, designing and management of the water resource systems. Over the United States, precipitation extremes, temperature and streamflow, have increased during the twentieth century and has been attributed to many natural and anthropogenic influences. The present study examines the association of precipitation extremes on temperature over US for the period of 1950-2000. The annual maxima (AM) precipitation has been extracted for hot and cold years. The spatial mean of surface temperature/ sea surface temperature from 1950 to 2000, so obtained is arranged in ascending order. The corresponding years, with lowest temperature of 25 years are defined as cold years and highest temperature of 25 years are defined as hot years respectively. The spatio-temporal variability of 50 year return level (RL) for the AM is determined considering generalized extreme value (GEV) and non-parametric kernel distributions. To identify the significant changes in the derived RL from cold to hot years, a bootstrap-based approach is implemented. The results exhibited no significant changes in the 50 year RL of AM precipitation between hot and cold years, with 70% of total grids showing no significant changes with respect to both land surface and sea surface temperature at 20% significance level. The scatter plot between the spatial mean of AM precipitation and both land surface and sea surface temperature over US showed no association. Further the comparison with the CMIP5 models revealed that the models are showed significant association between both land surface and sea surface temperature with the AM of precipitation. The major decision making and planning rely on the model predictions, which shows increased global warming with increase in precipitation extremes, however the observed data shows no such significant relation. Hence, the results imply the requirements of serious research on the association between intensification of precipitation extremes and temperature increase.