During the last 150 years, there has been a tremendous increase in the amount of information available to the forecaster for detection, analysis, forecasting, and post-analysis of tropical cyclones. Originally, real-time observations were limited to reports relayed by telegraph from sparse land-based weather stations. Available observations gradually increased over time in response to technological advances such as radio, radar, aircraft, and rocketry. Real-time observations today include a vast array of surface, aircraft, satellite, radar, and other types of data, which have increased the information provided to forecasters for detection and analysis by orders of magnitude. No part of this advancement was more important that the development of the geostationary weather satellite, which enables forecasters to routinely monitor the oceanic breeding areas. In terms of post-analysis, the primary data source used to be ship reports that were delivered to the forecasters well after the storm was over. In today's world, there can be an overwhelming amount of data that can only be thoroughly examined after the fact, and observations from experimental or non-convention data sources can arrive months after the cyclone. These data, combined with advances in computers, mathematics, data assimilation, and theoretical understanding of how the atmosphere works, have led to equally tremendous advances in the forecasting and warning processes for tropical cyclones. Older forecasters applied more art than science and often struggled to provide even 12-24 hours of warning of an impending storm. Today's forecasters can often provide five days of lead time of the approach of a tropical cyclone, with precautionary watches and warnings being issued 36-48 hours in advance of the onset of the dangerous conditions. Two key aspects of these advances have been the development of global dynamical forecast models that have improved tropical cyclone track forecasts, and the development of storm surge models that have improved the identification of the areas most at risk from inundation. Many challenges remain for the tropical cyclone forecast and warning process. First, the coastal population in the United States and other cyclone-prone regions are growing faster than the ability to forecast cyclones. Second, while much progress has been made in track forecasting, intensity forecasts remain far less skillful, and much work remains to be done in forecasting cyclone size and rainfall. Finally, our improved ability to observe has led to the discovery of cyclones that are marginally of tropical cyclone character. The continuous spectrum of cyclones produced by nature often poses problems for the discrete warning classification systems used by operational forecast centers.