I have collaborated with colleagues Mary Downtown and Zoe Miller on a project sponsored by the National Oceanic and Atmospheric Administration to reanalyze the historical record of flood damages in the United States. The results of our project can be found here: http://www.flooddamagedata.org. Last week a journal article describing the project came out in Natural Hazards Review:

Downton, M., J. Z. B. Miller and R. A. Pielke, Jr., 2005. Reanalysis of U.S. National Weather Service Flood Loss Database, Natural Hazards Review, 6:13-22. (PDF)

As far as historical data on flood damages in the United States, I believe that this dataset represents the best information available. The data can be used for research purposes, but caution is advised (read the paper). We have a companion article in press describing limitations in disaster loss estimates:

Downton, M. and R. A. Pielke, Jr., 2005 (in press). How Accurate are Disaster Loss Data? The Case of U.S. Flood Damage, Natural Hazards Review. (pre-publication PDF)

Have a look at the project site: (http://www.flooddamagedata.org/ ) And have a look at the paper: (PDF) Here is the abstract:

“Abstract: To understand the nature of increasing flood damage in the United States, accurate data are needed on costs and vulnerability associated with flooding. The National Weather Service (NWS) is the only organization that has maintained a long-term historical record of flood damage throughout the country. The NWS estimates are obtained from diverse sources, compiled soon after each flood event, and not verified by comparison with actual expenditures. This paper presents results of a comprehensive reanalysis of the scope, accuracy, and consistency of NWS damage estimates from 1926 to 2000 and recommends appropriate methods for data use and interpretation. Estimates for individual flood events are often quite inaccurate, but when estimates from many events are aggregated the errors become proportionately smaller. With the precautions described in this paper, the reanalyzed NWS damage estimates can be a valuable tool to aid researchers and decision makers in understanding the changing character of damaging floods in the United States. The reanalyzed data are available at http://www.flooddamagedata.org/.”