From its formation as a tropical storm in the Caribbean Sea to its dissipation over the North Atlantic Ocean, this satellite imagery showcases the eventful eleven-day journey of Hurricane Melissa.
I'm just going to point out that there is no quality scientific analysis being presented here.
1. Kantar is a market research company specializing in advertising, branding, and public opinion, without expertise in meteorological analysis or tropical cyclone datasets.
2. The provided "report," unlike peer-reviewed research papers, obfuscates the analysis and provides no data or comprehensive process with which to replicate and vet the claimed results.
3. The report states: "the
forecast time steps analyzed were hours 12, 24, 36, 48, 60, 84 and 108 from the issued time." But @NHC_Atlantic issues forecasts at lead times of 12, 24, 36, 48, 60, 72, 96, and 120 hours, making it unclear exactly how this analysis was done. Was interpolation of some data to a different set of lead times performed? Was mean absolute error or some other metric used? We don't know, because the report is opaque.
4. Anyone in this field knows that a single season of hurricane statistics is not robust. The sample size is not large enough to yield statistically rigorous results, given the variability between individual storms and years. This is why multi-year samples are typically necessary to show statistically significant performance gaps between different types of forecasts.
If you think I'm overtly taking @NHC_Atlantic's and @NOAA's side here, yes, I am. American publicly funded hurricane research, computer modeling, and expertise are unrivaled worldwide. And, by the way, @Accuweather wouldn't be able to issue their own forecasts without any of those things.
The full Kantar report: https://t.co/f47GELgRly