...
- Megi, Bopha: super-typhoons with far too high minimum pressure
The figures above shows the MSLP forecast for Haiyan (4 day forecast valid 7 Nov 12 UTC), Megi with T1279 (5 day forecast) and Megi with T3999 (also 5 day forecast). All three figures have a 5x5 degree box. For Megi, the observed minimum pressure was 911 hPa. For T1279 it was 990 hPa and T3999 920 Pa. The figure below shows time-series of the minimum pressure for Megi for these two runs (T1279 -red, T3999 - blue) ) and the operational forecast 12 hours before the landfall (green), together with the Best Track estimate (black). The cyclone made landfall on the Philippines 19 October 2010. We clearly see that the minimum pressure improves with resolution. After passing the Philippines the cyclone deepened in the two forecasts from the 13 October, which is a clear forecast error (probably of random nature).
However, we need to do T3999 simulations also for weak tropical cyclones to verify that we not have a negative pressure bias.
- Francesco: Why did the model produce such a low pressure for Francesco and not for Haiyan?
The figures above shows examples of MSLP of Francesco (left) and Haiyan (right), 36 hours into the forecast. Here we see that Francesco was a much more large-scale system, so the pressure gradients look similar even if Francesco had a minimum pressure of 911 hPa and Haiyan 944 hPa.
...