Assessing Future Changes in Climate Extremes using the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) Dataset

Abstract:

Heat waves will become more frequent in the future under a warming climate. Their duration and intensity will increase as well. In order to get a better understanding of the trend of heat wave occurrence and intensity over the US, this study systematically analyzes daily maximum temperatures from an ensemble of 21 CMIP5 GCMs from the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) datasets. We define the occurance of a heat wave as a period of 3 (or more) consecutive days with daily maximum temperature above a threshold, and the intensity of the heat wave is the accumulated daily maximum temperatures over its duration. We calculated the two metrics for the period 1950-2099 over US for all models under Representative Concentration Pathway 8.5 (RCP8.5). We also test two different approaches to estimate the ensemble means of the heat wave metrics. The first approach calculates the numbers and intensity of heat waves for each individual model and then calculated the ensemble mean over 21 models of these two metrics, while the second approach first calculates the ensemble mean of the maximum temperatures over 21 different models and then estimates the annually heat wave numbers and heat wave intensity. The two approaches lead to remarkably different results: The results from the first method detect significant increasing trend of annually heat wave frequency and intensity in each individual GCMs. The mean durations of the heat waves are also found to be steadly increasing during the rest of the 21st century. The results from the second method show less-significant increasing trend of annually heat wave numbers. However the intensity of heat waves in the second approach is much larger than that calculated from the first way. These results confirms our hypothesis that both the occurance and the intensity of heat waves will increase in the future. They also emphasize that caution must be taken in processing high-frequency (daily or above) climate data in order to provide consistent and coherent assessments of future changes of climate extremes.