Development was done by using data from 335 stations; 300 were from the contiguous U.S., 30 from Alaska, and 5 from Puerto Rico and Hawaii. This was a single station development.

**
TESTING SAMPLES
The full cool season sample was from:
**

**September 16, 1997, to April 15, 1998****September 16, 1998, to April 15, 1999**

- the last 15 days of October, November, and December 1998
- the last 15 days of January, February, and March 1999

- the remainder of the test sample listed above

- April 1, 1997, to October 15, 1997
- March 16, 1998, to October 15, 1998

- the last 15 days of April, May, June, July, August, and September 1998

- the remainder of the test sample listed above

PREDICTANDS

2-m temperature

- Valid every 3 hours from 6 to 72 hours after initial model time

- Valid every 3 hours from 6 to 72 hours after initial model time

- Is defined as 0700LST - 1900LST
- Valid approximately 24, 48, and 72 hours after 0000 UTC and 36 and 60 hours after 1200 UTC

- Is defined as 1900LST - 0800LST
- Valid approximately 36 and 60 hours after 0000 UTC and 24, 48, and 72 hours after 1200 UTC

Model Predictors

- Forecasts of temperature and dew point at isobaric levels in the lower troposphere, and low-level thicknesses
- Sea level pressure tendency
- Relative humidity and thermal stability over layers in the lower troposphere
- Mean relative humidity from the surface to approximately 500 mb
- Precipitation amount
- u- and v- wind components, wind speed, and vertical velocity at isobaric levels in the lower troposphere

- 2-m temperature and dew point observed at initial model time + 3 hours, that is, 0300 or 1500 UTC; observations are potential predictors for the 6- through 39-h forecast projections.

- The sine and cosine day of the year and twice day of the year were used.

Max/Min Temperature, 2-m temperature and 2-m dew point

- Mean Absolute Error

Conclusion

- The AVN MOS max/min temperature and 2-m temperature guidance were comparable in skill through 48 hours to the NGM MOS guidance. Beyond 48 hours, the AVN MOS were generally more accurate than the NGM MOS. Generally, the MOS guidance significantly improved over the direct model output.
- The AVN MOS dew point guidance was generally more accurate than the NGM MOS or the direct model output.

Please select a product from the menus on the left.