Have Improvements in Hurricane Intensity Forecasts Been Driven by Better Track Predictions or Achieved on Their Own?

Tropical Cyclone (TC) forecasting has long been one of meteorology’s major challenges. Technological advancements over the past few decades have led to major improvements in TC track forecasting, while TC intensity forecasting has shown relatively smaller improvements. However, it remains unclear if these improvements in TC intensity forecasting are a byproduct of improved TC track forecasts or if improved TC modeling and understanding of TC intensity change have led to improvements in intensity predictions independently of track improvements. This study uses the National Hurricane Center Official Forecast Error Database to examine the causes of recent progress in TC forecasting. We analyze TC track and intensity errors across multiple decades by binning the track forecast errors based on the degree of error. By binning by the track error, we can isolate the improvements in intensity forecasting from the track improvements by keeping the track error constant. We perform this analysis for all forecasts, as well as for forecasts near the North American coast, and the eastern and western tropical Atlantic Basin. We find that when improvements in TC intensity forecasts are separated from track forecasts, there remains improvement in TC intensity, indicating that intensity forecasting is improving independently of track forecasting. This suggests that improvements in TC intensity are the result of improved modeling, data assimilation, and better understanding of TC dynamics.