CrazyC83 wrote:Looking through the data for Jose, I agree that it should be at least 140 kt. I think they were trying to avoid a posthumous upgrade as much as possible.
The SFMR of 142 looked very reasonable to me, and that *alone* would support 140-145 kt. Even if that data is discarded, the Dvorak ratings were higher at 1800Z (between the Recon passes) than at either 1200Z or 0000Z. If you use 135 kt at 0000Z, then at 1800Z it should be 140 kt just on satellite interpretation.
I would have gone with a peak intensity of 145 kt at 1800Z as a result (with a reasonable range being 140-150 kt). The minimum pressure I would have set at 935 mb at that time on the same basis.
From page 4 of the TCR:
Jose’s estimated peak intensity of 135 kt from 1800 UTC 8 September to 0000 UTC 9 September is based on a blend of SFMR and flight-level aircraft data from two consecutive Air Force Reserve Hurricane Hunter missions. During the first mission, the aircraft measured peak 700-mb flight-level winds of 146 kt and 142 kt at 1208 UTC and 1356 UTC 8 September, respectively, which equate to an intensity of about 130 kt. The highest “unflagged” SFMR measurement during the same flight was 125 kt. On the next flight, the plane measured a slightly lower peak 700-mb flight-level wind of 135 kt at 0026 UTC 9 September. The highest SFMR measurement was 142 kt a few minutes later, however this observation was surrounded by multiple flagged values and data dropouts, making the measurement somewhat questionable. Without entirely discounting this measurement, Jose’s maximum intensity is estimated to be 135 kt, which is a blend of the SFMR data and the highest flight-level-adjusted winds.
I actually think the reasoning for the intensities at the time of the two aircraft fixes (at 12:08–13:56 UTC/8 September and 00:26 UTC/9 September, respectively) is sound. The SFMR readings were not as reliable as I originally suspected, as I was unaware of the flagged readings, which give added context. I do support the intensities of 130 and 135 knots at 12:00 UTC/8 September and 00:00 UTC/9 September. The major issue, as you mentioned, is that the satellite estimates were higher around 18:00 UTC than at 00:00 UTC, which would support an increase of about five knots. To cite one past example from the Atlantic basin: the Hurricane Center used satellite estimates to derive Danny's 110-knot peak intensity in 2015, as the reconnaissance data supported 100 knots several hours after the satellite presentation had degraded substantially. The NHC also followed a similar reasoning in its assessment of Patricia's (2015) maximum intensity over the eastern North Pacific. Thus, the failure to apply similar scientific reasoning in this case is inconsistent with past practice, and may harm efforts to introduce consistency and accuracy to global tropical-cyclone records. It corrupts the data.
Why do you think "they were trying to avoid a posthumous upgrade"? The NHC has upgraded a number of storms, including Emily of 2005 (to 140 knots) and Otto of 2016 (to 100 knots).