The animation shows a sequence of 31 frames representing the state of the 500mb flow at 00Z on successive days. The first 15 frames are past analyses, and are denoted by the green title bar and green portion of the clock in the upper left corner. The analyses are based on observations (weather balloons, aircraft measurements, satellite measurements) assimilated into the computer model of the atmospheric circulation. The animation pauses at day 0, representing the most recent observed state of the atmosphere, and the start of the current forecast cycle. The next 7 frames (yellow color keyed) are the first week of the AVN high-resolution forecast. The final 8 frames (red keyed) are the continuation of the AVN forecast at a lower resolution.

The red and blue shading indicate positive and negative vorticity respectively. The black contours show the 500mb geopotential height with a contour interval of 120 meters. The heavy contour is 5400m. The form of these maps are essentially the same as the 500mb forecast and analysis maps - see the key for the forecast maps for more details.

Things to notice:

• The features tend to "spin" counterclockwise around the North Pole, reflecting the pervasive west-to-east circulation in the troposphere. The rate of motion is considerably slower at low latitudes, with features in the tropics appearing rather stationary, or even moving clockwise.
• The picture becomes very "smooth" in the interval from Day 7 to Day 8. This reflects the reduction in resolution of the model at that point. You may also notice that the character of the circulation appears to change at this point (e.g. features wash out, and the rate at which the vorticity features traverse the map abruptly changes). This indicates the effect that horizontal resolution may have on model simulation. Presumably, the higher the resolution, the better the forecast.
• Sometimes there appears to be an abrupt change in the character of the circulation at the transition from analyses to forecast. This may indicate that the model is not representing the circulation well, and may indicate a poor forecast. However, this effect is not well understood, and is a topic of ongoing study. One motivation for conducting ensemble forecasts (with many realizations of the same forecast period) is to isolate and disregard such inconsistent forecasts by comparing each member of the ensemble to the whole.