Early Wednesday morning, 29 August 2012, Hurricane Isaac made its second landfall at around 0700 UTC per a National Hurricane Center Advisory. While more complete data is unavailable, and specific observational tools may not tell the whole story, I thought I would share a few of the analysis products for Isaac that the Tropical Cyclone Winds CSTAR project has examined with other storms.
The Hurricane Research Division (HRD) continues to participate in the H*Wind Project whose purpose is to develop an integrated tropical cyclone observing system in which wind measurements from a variety of observation platforms could be used to develop an objective analysis of the distribution of wind speeds in a hurricane. The HRD wind analysis requires the input of all available surface weather observations (e.g., ships, buoys, coastal platforms, surface aviation reports, reconnaissance aircraft data adjusted to the surface, etc.). More information on the H*WIND data is available in a previous blog post – HRD Real-time Hurricane Wind Analysis System (H*WIND. The H*WIND data is available every 3 hours and an analysis was available just before Isaac made landfall at 0430 UTC, and is shown above. Archived H*WIND data for Isaac is available at this link.
The Multi-Platform Tropical Cyclone Surface Wind Analysis (MPTCSWA) from CIRA for 0600 UTC 29 August is shown to the right. This product is intended to provide an “estimate” of the 1-minute maximum winds at 10-m elevation. Note that there are several limitations with this data set and more information on the MPTCSWA is available from a previous CIMMSE blog post – Hurricane Rina, Irene, and the Multi-Platform Tropical Cyclone Surface Wind Analysis. Archived MPTCSWA data for Isaac is available at this link.
Short term wind forecasts from the NDFD valid at 0300 UTC (left) and 0600 UTC (right) on 29 August are shown below. The forecasts appear well collaborated and depict a symmetric storm system with a sharp gradient in wind speeds on the perimeter of the storm.
A quantitative verification of wind and wind gust forecasts is not available at this time but some general comments can be made. The figure below shows the short term wind forecasts from the NDFD valid at 0300 UTC (left) and 0600 UTC (center) on 29 August with the 0430 UTC H*WIND analysis (right). The H*WIND analysis is preferred over the MPTCSWA because of the greater number of observations the H*WIND uses and some studies note that the MPTCSWA has a high bias compared to the H*WIND. A recent study notes that for well sampled, stable, TCs over open water, the H*WIND itself has a general positive bias to the wind speed as noted in a blog post – H*Wind Uncertainty Estimates. When comparing the H*WIND analysis with the NDFD wind forecasts, the NDFD winds appear to have some positive bias, especially at longer distances from the storm center. In addition, the over-land wind speed reduction might be under represented. The positive NDFD wind bias and an insufficient over-land reduction of winds are consistent with a verification of NDFD forecasts for tropical cyclones in the Southeast, completed as a part of the TC Wind CSTAR project. More information on the NDFD verification is available in a previous CIMMSE blog post – NDFD Final Verification.
Short term wind gust forecasts from the NDFD valid at 0300 UTC (left) and 0600 UTC (right) on 29 August are shown below. Some minor collaboration issues are noted, especially on the perimeter of the storm. A very large area of wind gusts in excess of 50 kts is shown in the dark red but it is difficult to note whether the wind gust speeds are too strong.