Temperature

BCRMSE

    • In general, there is an increase in BCRMSE with forecast lead time. For both the 00 and 12 UTC initializations, a diurnal trend is displayed in the distribution of BCRMSE values, which generally range between 2.0 - 3.0 C, depending on the temporal aggregation.
    • For the spring and summer aggregations, there is a maximum in forecast error magnitude around 21-00 UTC and a minimum around 12-15 UTC valid time. For the fall aggregation, the maximum in BCRMSE is found at 12 UTC. An exception is winter, where the errors tend to remain high throughout the daytime hours.

Bias

    • A diurnal signal is seen regardless of initialization or temporal aggregation. There is generally a SS cold bias for all forecast valid time, with the magnitude of the cold bias being largest during 21-03 UTC and smallest during 09-15 UTC.
    • Larger cold biases as well as more pronounced diurnal signals are seen in the fall and winter aggregations. Depending on temporal aggregation, the magnitudes of the diurnal maximum cold bias range from -0.8 to -2.3 C, and the magnitudes of the diurnal minimum range from -0.2 to -1.0 C.

Dew Point Temperature

BCRMSE

    • All temporal aggregations reveal an increase in BCRMSE values with lead time. A strong diurnal trend is noted in all temporal aggregations except fall and winter, where the signal is weak.
    • For both the 00 and 12 UTC initializations, BCRMSE values generally range from 2.2 - 3.5 C. The largest errors are generally seen around 21 - 00 UTC for all temporal aggregations. The smallest errors vary in exact time for each season, but generally occur during the overnight hours between 03 and 12 UTC.

Bias

    • A strong diurnal signal is noted regardless of initialization or temporal aggregation. Generally, there is a SS wet bias for forecasts valid between 18 and 03 UTC, and a SS neutral (CIs encompassing zero) or dry bias for forecasts valid between 09 and 15 UTC. Depending on temporal aggregation, the diurnal maximum bias values range from 0.6 to 2.2 C, and the diurnal minimum bias values range from -0.8 to 0.5 C.
    • The largest SS wet biases are seen in the spring aggregation, and more SS neutral (CIs encompassing zero) or dry bias values are noted in the fall aggregation. In addition to the diurnal variation, there is also a trend of the bias drifting toward the dry side with increasing lead time for most aggregations.

Wind

BCRMSE

    • For both initializations, the BCRMSE increase with forecast lead time, with values ranging from approximately 1.6- 2.6 m s-1.. A diurnal trend is noted for all temporal aggregations except fall and winter.
    • For the annual, summer and spring aggregations, the largest BCRMSE values are seen at forecast lead times valid at 00-03 UTC, while the smallest errors are around the 12 UTC valid time, regardless of initialization time.
    • Due to the weak diurnal signal seen for the fall and winter aggregations, the error is fairly consistent across valid times, especially when CIs are considered.

Bias

    • A SS high wind speed bias is noted for all forecast lead times, initializations and temporal aggregations, with the exception of a few neutral values (CIs encompassing zero) for spring and summer aggregations during daytime hours (15-21 UTC).
    • For both initializations, a very large diurnal cycle is noted with SS larger high bias during the overnight hours (03 - 12 UTC) compared to the daytime hours (15 - 21 UTC). The overnight largest bias values are 1.1-1.3 m s-1 and the daytime smallest bias values are 0.0-0.4 m s-1 depending on temporal aggregation. There is a also a trend of the bias drifting toward the high wind side with increasing lead time.

3-hour Precipitation Accumulation

GSS

    • Regardless of initialization time or forecast lead time examined, the highest GSS values are seen at the lowest precipitation threshold and steadily decrease to near-zero for higher thresholds. The number of observed events (base rate) by threshold has a similar trend. The largest CIs are typically noted for the winter aggregation which also has the lowest base rate, and for summer lead times valid at 12 UTC. In general, the GSS values are higher in the winter and fall than in the spring and summer aggregations.

Frequency Bias

    • For the annual, fall and winter aggregations, a general high bias is seen for all but the highest thresholds, regardless of forecast lead time or initialization hour. This basic trend holds for spring and summer lead times valid at 00 UTC; however, for 12 UTC valid times, most results are not significantly different from an unbiased forecast. A only exception is a low bias for the lowest threshold (>0.01 in) in the summer forecasts valid at 12 UTC.

24-hour Precipitation Accumulation

GSS

    • A general decrease in GSS with increasing threshold and lead time, similar to that seen for 3-hour precipitation accumulations, is noted for all temporal aggregations except winter. Again, a similar trend is noted in the base rate, with very few observed events of 3" accumulations in 24 hours.

Frequency Bias

    • Most temporal aggregations exhibit a fairly consistent high bias for all but the lowest (>0.01 in) and highest (>3.0 in) thresholds, which have CIs that encompass one. For the summer aggregation, there is a low bias for the lowest threshold (>0.01 in).
    • In general, as the threshold increases, the width of the CIs increase, indicating more spread and less confidence in the median bias value, likely related to the decreasing base rate.

Temperature

BCRMSE

    • Regardless of forecast lead time or temporal aggregation, a minimum in the BCRMSE values occurs between 500 and 300 hPa and increases to a maximum at both the near-surface and upper levels.
    • A SS increase in BCRMSE values between the 12-hour and 48-hour forecast lead times is noted. Errors are less than 2.2 C for all lead times.

Bias

    • At forecast initialization time, a SS warm bias is noted throughout the vertical levels (except for neutral biases at 150 hPa for spring and winter). The biases at lower vertical levels generally move toward the cold side with increasing forecast lead time.
    • For all temporal aggregations, at lead times beyond 12 h or 24 h, a SS neutral or cold bias is seen at the 850 hPa level, transitioning to a SS warm bias with height. The level of transition occurs near 700 to 500 hPa, but is dependent on temporal aggreagtion and lead time. One exception is often noted at 150 hPa where there is large variability in the bias. A vertical local minimum is noted which translates into a larger cold bias or a smaller warm bias at that level.

Dew Point Temperature

BCRMSE

    • As the pressure level decreases from 850 to 500 hPa and the forecast lead time increases from 00-hour to 48-hour, the BCRMSE values generally increase for all temporal aggregations. The values range from 2.0 C (850 hPa, summer, initial time) to 10.3 C (500 hPa, spring 48-hour).

Bias

    • At forecast initialization time, a SS high bias is noted throughout the vertical levels for all temporal aggregations. The bias values increase with height (as pressure decreases).
    • As the lead time increases, the bias values at 850 hPa for all temporal aggregations and at all vertical levels for the summer aggregation decrease. For the 36- and 48-hour lead times, a SS neutral or cold bias is noted at 850 hPa for most aggregations and at 700 hPa for summer. The bias values range from to -0.8 C (850 hPa, fall 48-hour) to 3.4 C (500 hPa, spring 12-hour and fall 48-hour).

Wind

BCRMSE

    • A general increase from the minimum BCRMSE values at the lowest levels (typically at the 850 or 700 hPa level) to a maximum generally found at 300 to 150 hPa level for all forecast lead times and temporal aggregations, is noted.
    • Overall, the errors generally range from 2.7 - 6.7 m s-1 depending on lead time and temporal aggregation, with a SS increase in BCRMSE values between the 12- and 48-hour lead times seen for several levels.

Bias

    • In general, at forecast initialization time, there is a SS low wind speed bias at all levels below 200 hPa, regardless of the temporal aggregation. As the lead time increases, the forecast bias values at the lowest and upper-most levels becomes more positive, transitioning into a SS neutral or high wind bias around 24-hour lead time.
    • At forecast initialization time, the largest low wind speed bias is found at the lower to mid levels, whereas for the 12- to 48-hour lead times, the largest low bias is observed at 200 hPa for all temporal aggregations. The largest high bias is found at the upper-most levels. The bias values generally range approximately from -0.8 to 1.2 m s-1.