Very very strong.
Outstanding guidance.
Statement as of 5:00 PM EDT on August 19, 2007
The reconnaissance aircraft completed its mission...finding peak flight-level winds of 144 kt...with peak SFMR winds of 121 kt. One of the last eyewall dropsondes gave a surface estimate of 118 kt derived from the lowest 150 meters of the sounding. In spite of a gradual pressure rise...these data indicate that 125 kt is still a good intensity estimate. Concentric eyewalls are apparent in microwave...conventional satellite...and aircraft data...with the strongest winds now associated with the outer eyewall. Short-term intensity changes will be governed by inner-core processes that are difficult to time...but overall conditions in the western Caribbean would favor further intensification...and Dean could very well become a category five hurricane during the next day or two. The initial motion is 280/17. There is no major change to the forecast thinking...although model guidance has been trending southward...with nearly all of the large-scale models building the subtropical ridge westward in tandem with Dean. The upper low in the Gulf of Mexico continues westward and is not expected to be a major player in the future track of the hurricane. The official forecast has been shifted a little to the south...but still lies to the north of all the major models except for the hwrf. Although I cannot explain why...the deterministic run of the GFS is along the southern end of the guidance envelope...while the GFS ensemble members are almost uniformly along the northern edge. Because most of the guidance models lie to the south of the official forecast... some additional southward adjustments to the track are possible. Forecast positions and Max winds initial 19/2100z 17.3n 76.8w 125 kt 12hr VT 20/0600z 17.8n 79.4w 130 kt 24hr VT 20/1800z 18.4n 83.0w 135 kt 36hr VT 21/0600z 19.0n 86.5w 140 kt 48hr VT 21/1800z 19.9n 90.0w 80 kt...inland 72hr VT 22/1800z 21.5n 96.5w 100 kt 96hr VT 23/1800z 23.0n 101.0w 35 kt...inland 120hr VT 24/1800z...dissipated $$ forecaster Franklin/Rhome