This app explores temperature variability in North America. We have used a novel technique to model temperature as a smooth function of both day of the year and number of years since the start, 1850. Thus, we obtain accurate estimates of various temperature quantiles. The data set is obtained from a CESM (Community Earth Simulation model) and is an ensemble of 50 independent climate simulations spanning the years 1850 to 2100 (courtesy of Ryan Sriver) .
Select a coordinate point using "Latitude" and "Longitude" and a day in the year, using "Day", to look at the initial vs. final shifts in temperatures as shown in the top right panel. The map on the top left shows the given quantile difference change between 1850 and 2100 with a cross if the change is less than 4 standard deviations away from the initial value. Select two quantiles, using "q1" and "q2" to see how these quantiles change in time. In the bottom two panels, each line corresponds to a time-slice in the 1850-2100 model, with a 40-year interval between each line. The quantile difference gives an estimate of the variability between the given quantiles. Thus, one can use the quantile difference as an estimate of high and low extremes variability. The "Data Type" variable refers to the input data: Daily average (will expand to max/min). There are locations that have unstable estimates as seen by the large confidence intervals