Description: Root Mean Square Error (RMSE) is a measure of the difference between model forecasts and observations. The closer that a point is to an RMSE value of 0.0, the more accurate the 2-m temperature forecasts were over the contiguous U.S. (CONUS) during a given year. Years are shown along the x-axis. RMSE values for Day 1, Day 3, Day 5, and Day 10 forecasts are shown in black, red, blue, and green lines, respectively. Note that short-range GEFS forecasts (e.g., Day 1) have less RMSE (i.e., more skill) than longer-range GEFS forecasts (e.g., Day 10). A decrease in RMSE from left to right indicates that GEFS forecasts have improved over time.