1. MOS does not account well for precipitation totals from the result of deep convection. This is because deep convection occurs on a smaller scale than MOS can account for. In reality, some places in the same forecast region will have a magnitude or more precipitation than other nearby location within the forecast region. MOS will average the precipitation over a synoptic region if deep convection is favorable over that region.

2. Be wary of extended forecast MOS numbers. MOS trends toward climatology in extended forecasts, especially beyond the 48 hour time period. Beyond 48 hours, the MOS precipitation values will rarely be near 100% or 0% due to the trending toward the climatological precipitation probability for any given day in that forecast area. MOS temperatures have trouble beyond the 48 hour time period when rapidly changing or unusual weather is expected to occur.

3. Take note of mesoscale effects important to the forecast region. The MOS equations have difficulty accounting for localized changes in vegetation, topography, lakes, sea breezes, mountain/valley winds, urbanization, pollution, snow cover, influences from deep convection, fog, soil moisture, and cloud cover thickness among others.

4. MOS does poorly in ageostrophic environments. This is to be expected since the weather is most difficult to predict when it is changing rapidly.

5. The MOS highs tend to be more accurate than the lows. During the day wind speeds tend to be higher and the result is well mixed air. Well mixed air is uniform in temperatures. At night when wind speeds become less, the slacking of mixing results in pronounced mesoscale temperature changes.

6. The current weather conditions are a more reliable indicator than what MOS forecasted them to be at that time. Make sure the MOS data being used is the latest data. Compare the predicted MOS values to the current trends in temperature, dewpoint, wind, etc. This will give the forecaster an idea of the accuracy trend of the MOS data.

7. MOS does poorly when rare weather events occur. Much of the MOS equation is based on climatological factors. Examples of rare weather events include hurricanes (tropical systems), cold air outbreaks, persistent drought.

8. If the models are slow or fast with a storm system, the MOS POP values can be in great error. Often MOS POPs will have a tight gradient between POP values. Therefore, it is important for the forecaster to recognize when the tight gradients may shift due to the model underestimating or overestimating the motion of a precipitation producer. Also, slight variations in the data inputed into the MOS equation can produce dramatically different results. For example, a slight overestimation of low level moisture or temperature can increase the MOS POP by 20% in some situations.

9. Use MOS high and low temperature values correctly in cases where the high or low temperature does not occur at the traditional times of the day. This can occur when a cold front moves through in the middle of the day. Use the 3-hour temperature trend to determine when a frontal passage is expected to occur. Inform the public in situations where the temperature trend during the day will be abnormal.

10. MOS is more accurate with average wind direction than the average wind speed. MOS tends to underestimate wind speed on a typical day (it tends to make light winds lighter).