Satellite-derived winds and Tropical Desk verification

An interesting concept arose over the past few day. Those at the mock Tropical desk are exploring new product options to take the place of the soon-to-be-retired Area Forecast (FA) text products that are currently issued both domestically and internationally. This is more of a concern for the latter as unlike domestic, they do not issue the graphical AIRMETS; only FAs and SIGMETs. So, they've spent the past week exploring some graphical-based alternatives.

The challenge arises in verification. There are very few observations over the oceans and even less radar coverages. This does not leave a whole lot of data to work with, particularly for verifying low-level wind forecasts. It may seem like low-level winds aren't of too much significance since there are not many terminals in this forecast area and aircraft typically are found at cruising altitude. However, the Caribbean and particularly the Gulf of Mexico are home to a very large number of helicopter operations, all of which take place at low levels. It is for this reason the wind forecasts are issued. Check out an experimental graphical forecast for the tropics issued yesterday, including all hazards.

20150812 0000 UTC forecast. Cyan scallops indicate convection, red filled indicate turbulence, blue filled indicate icing, and white filled indicate areas of forecast winds.

In an effort to verify the experimental forecast wind areas, participants examined satellite derived atmospheric motion vectors from GOES-E and scatterometer data from ASCAT.

20150812 1800 UTC Tropical desk forecast and ASCAT winds at ~1524 UTC

The forecast called for winds around 20-25 knots just north of South America. Scatterometers indicated a large area of winds near 20 knots. There were also some showing 20-25 knots. It is important to note though, that the forecast was for 1800 UTC and the nearest ASCAT pass came at ~1524 UTC. Winds were forecast to decrease in this area throughout the afternoon, and could have easily done so in the three hours between the ASCAT pass and 1800 UTC time period. This considered, using these winds was actually a very useful tool for verification of this type.

AMVs were also explored for verification. This however, is somewhat more difficult to use for surface winds given that they are estimated from cloud objects. However, the lowest layer, from roughly 1000-950mb, may be of use for this concept and will be looked at later in this experiment. Additionally, it was noted that the higher level AMVs may be of value in verifying some the global graphics wind forecasts issued by the AWC's international forecasters.