Our ability to observe extreme, mesoscale cyclones over the tropical oceans - hurricanes - has improved dramatically during the last century. Up until the 1940s, hurricanes were only observed by unlucky/unskillful ship captains and the inhabitants of coastlines when these cyclones made landfall. During World War II came the advent of aircraft reconnaissance, while still quite primitive in their measurement capabilities compared with today, these aircraft still provided a huge advance for the monitoring of Atlantic basin hurricanes and Northwest Pacific basin typhoons. The 1970s and 1980s heralded the satellite era with geostationary satellites providing another giant step forward with coverage over entire basins as well as vastly improved temporal monitoring. The 1990s and the most recent decade have seen the development of imagery and data from both passive and active instruments on low earth orbiting satellite, providing more direct measures of hurricane positions, intensities, and sizes.

All of these changes - generally improvements (but not always with, for example, the discontinuance of aircraft reconnaissance in the Northwest Pacific after 1987) - have affected our climate records of the frequency, intensity and duration of hurricanes. There are efforts underway to reanalyze the hurricane databases by obtaining all of the original observations and reassess existing cyclones (and searching for previously unrecognized systems) by using today's understanding and analysis methodology. However, such work will never be able to overcome the lack of sufficient measurements taken decades previously to fully account for what has occurred in the past. Fortunately, there have been innovative new methods devised to account for undersampling of hurricane frequency, intensity, and duration. These allow for a much more homogeneous estimate of the actual long-term trends and variability in hurricane activity.