The Reasons for a Reanalysis of the Typhoons Intensity in the western North Pacific

At the time where a recent article (Webster and al., 2005) claimed that the number of intense tropical cyclones have significantly increased since 1975, it was very interesting to have a view as objective as possible on the quality of the best track data. We chose the western North Pacific because this basin is the more active, it has the greatest number of cyclones at the hurricane/typhoon's intensity (at least 33 m/s) and the more important number of cyclones with a sustained wind at or over 60 m/s (at least Category 4 of Saffir-Simpson). This basin is original too because the aerial reconnaissance missions did exist in the cyclones until August 1987 (ATCR, 1987). During this period, the cyclones intensity was primarily based on these reconnaissance data. Since August 1987, the satellite data analysis via the Dvorak's Technique is the only tool to estimate the cyclones intensity. There are two databases concerning the cyclones of the western North Pacific, respectively made by the official Regional Specialized Meteorological Center of Tokyo (RSMC) and by the Joint Typhoon Warning Center (JTWC) now based in Hawaii. Even if the RSMC of Tokyo uses the sustained wind over 10 minutes and JTWC the wind over 1 minute, this does not explain frequent substantial difference in the cyclones intensity estimate between these two agencies since the end of the aircraft reconnaissances. For example, in October 1998, Tokyo RSMC estimated the Typhoon Babs at its peak at 940 hPa-44 m/s over 10 minutes or 50 m/s over 1 minute whereas JTWC gave 905 hPa-70 m/s over 1 minute. This matched with an intensity's difference of almost 1.5 T-number on the Dvorak's scale. And in the 1995-1998 years, the intensity of cyclones could have been overestimated when the Objective Dvorak Technique was experimentally used in the WPAC by JTWC (ATCR, 1995 to 1998). Moreover, extreme typhoons like Gay (Nov. 1992) or Angela (Nov. 1995) displayed a stronger satellite signature than Typhoon Tip (870 hPa and 85 m/s in October 1979) but this is not reflected in the databases (Hoarau and al., 2004). Therefore, it is obvious that a reanalysis of the cyclones intensity after the reconnaissance era is needed. Related to this, there is still an important issue: is the database of the reconnaissance era before August 1987 more reliable than the data after this month? Here, one must to give the example of Super Typhoon Tip : on 11th October 1979 at 0000Z, JTWC and the RSMC of Tokyo gave an intensity respectively of 67 m/s over 1 minute and 67 m/s over 10 minutes or 77 m/s over 1 minute. These sustained winds matched with a Dvorak T-number of 6.6 for JTWC and a T-number of 8.0 for the RSMC of Tokyo whereas the reconnaissance aircraft gave a sea level pressure of 900 hPa and the maximum wind at the flight level of 700 hPa reached only 65 m/s in the strongest semi-circle (ATCR, 1979). It was very frequent that the radar signal of the aircraft was attenuated by the strong rain. So, the reconnaissance missions did not measured the true maximum flight wind. Therefore, the more reliable data recorded was the sea level pressure. And the wind was estimated from the satellite data and mainly from the Atkinson and Holliday (1977) wind-pressure relationship. This shows that all the cyclones of the WPAC, including those during the reconnaissances era, need to be reanalysed with the satellite pictures and the Dvorak's Technique. We already began a short reanalysis and found interesting cases to present.