The dataset merges the FEMA Q3 National Flood Hazard Layer (NFHL) data ("Q3 flood data"), developed in the 1980s and 1990s, with available data from the ongoing NFHL updates ("modern flood data"). Attributes are simplified, given the intended usage of this dataset.

Supplementary Information
Updated FEMA National Flood Hazard Layer data were obtained in February 2009. Areas covered by that dataset were removed from the FEMA 'Q3 Flood Data'. The two datasets were merged, and additional NoData zones was determined from the combined coverage. An ongoing effort to update all FEMA maps throughout the United States will eventually make the Q3 data obsolete. But as of February 2009, only 5 counties had updated coverage for all the areas covered by the Q3 flood data (CLAC, MARR, POLK, LANE, DESC) and 5 counties (MULT, WASH, UMAT, JACK, TILL) had partial updated coverage for all the areas covered by the Q3 flood data in digital form. (Modern Coverage for the entire county is limited to CLAC, DESCH).

Formal definitions of FEMA attributes:

Several notes on the merging of the two datasets:

FEMA flood data are spatially biased in that ample coverage exists around populated areas, yet less-populated or remote areas, especially large blocks of uninhabited federal land, have little or no coverage. The first question when using this data should be: does data exist for a particular area? Such areas are represented with FLD_ZONE = "NoData". Such polygons were created by the complete absence of any FEMA flood data, or where FEMA flood data existed, the FLD_ZONE values were either "ANI", "D", or "AREA NOT INCLUDED".

Modern flood data replaces the Q3 FLD_ZONE attribute of "X500" with "0.2 PCT ANNUAL CHANCE FLOOD HAZARD"; and to be consistent, the Q3 FLD_ZONE values of "X500" were changed to modern usage.

Some polygons in the modern flood data had an FLD_ZONE attribute value of "X PROTECTED BY LEVEE". Upon closer inspection, such a value made no physiographic sense (no levees present), and the value was assumed to be in error. Such polygons were given an "X" value instead.

In Oregon, Q3 Flood Data did not divide a 100-year flood zone into A, AE, AO, AH, etc. The modern flood data maintains such distinctions in Oregon, and the attribute was retained for possible use in ecological modeling. Likewise, the 'FLOODWAY' attribute was retained from both sets of data. If the distinctions between the different 100-year flood polygons are irrelevant for modeling purposes, the 100-year flood areas values can be determined using a wildcard match of "A%".

A 'SOURCE' attribute was added to distinguish polygons obtained from the modern dataset and the Q3 dataset.