Converting Broomfield to a DFIRM

by:
Bill DeGroot, P.E., Chief, Floodplain Management Program
Urban Drainage and Flood Control District

Introduction

In 2000, the voters of Colorado established the new City and County of Broomfield.   The City of Broomfield had taken the issue to the voters because it was located in parts of four separate counties, which greatly complicated the delivery of those services that are provided by counties in Colorado.  The new City and County of Broomfield officially came into being on November 15, 2001.

The Urban Drainage and Flood Control District saw the creation of this county as an opportunity to prepare a new countywide Digital Flood Insurance Rate Map (DFIRM), utilizing the Federal Emergency Management Agency's (FEMA) DFIRM specifications.  Our belief was that we could learn a great deal about the process that would be involved in such an effort while completing a countywide map for a small county, which was therefore affordable to the District.  The process we followed and the lessons we learned are discussed below.

After the District had begun its DFIRM conversion effort, FEMA published a draft Implementation Strategy for Flood Map Modernization.  We determined that our Broomfield effort very closely resembled FEMA's definition of a Level 1 Flood Map Upgrade.   This paper is intended to demonstrate how the District has in effect developed a Level 1 map upgrade for Broomfield, and how that DFIRM is vastly superior to the current paper FIRM.

Base Map

One of the key ingredients in FEMA's DFIRM conversion process is the ability to use the local governments' maps as the base map for the DFIRM.  Broomfield has an excellent ArcInfo based Geographic Information System (GIS), and was able to provide a current road centerline base map and a rectified orthophoto background; and permission to use them for this project.

Flood Data Sources

We knew that if we simply digitized the information from Broomfield's paper FIRM (at a scale of 1-inch equals 1000-feet) that we would not be improving upon the quality of the paper FIRM.  We therefore decided to digitize the original source material for each reach of floodplain in order to provide a more accurate DFIRM.

We reviewed the most recent revised Flood Insurance Study (FIS), which was dated September 30, 1997, and were able to determine the original source for each reach of each floodplain as of that date.  We then reviewed Letters of Map Revision (LOMRs) that had been issued after that date to identify more recent changes that we wanted to incorporate into the new DFIRM (Figure 1).

We found that there were four sources of information.  The first was the District's own Flood Hazard Area Delineation (FHAD) studies.  The second source was the work maps from the original FIS.  The third source was the work maps from several LOMRs, and there was one instance where the LOMR information was provided to us by FEMA's Map Coordination Contractor (MCC) in digital format.

It should be noted that the District's FHAD studies are done using projected future developed watersheds for the hydrology.  Because FEMA uses existing watershed conditions hydrology for its maps, there were cases in which the District had the only detailed flood data available, and FEMA could only utilize it as a Zone A (approximate floodplain).  In other instances the watershed was built out to the point where FEMA could use the detailed information and did so.  In those cases there was some information missing from the FHAD maps because the District does not publish the floodway boundaries or the 500-year flood boundaries.  Adjustments to the maps had to be made in these cases, and that process is described below.

The Process

The District contracted with its GIS consultant, Merrick & Company, to digitize the flood data from the various sources, and to otherwise prepare the flood data in conformance with FEMA's DFIRM specifications.  The District obtained the source data for each reach of the floodplain, color-coded the data to help Merrick personnel know what to digitize from each set of source documents, and performed quality control checks on the products completed by Merrick.

The flood data for the Zone A floodplains, which were taken from the District's FHAD studies, were relatively easy to deal with.  The FHADs were originally done utilizing mapping at a scale of 1-inch equals 100-feet with 2-foot contour intervals.  The 100-year flood outline, thalwag and structures were color coded by the District (Figure 2) and digitized accordingly by Merrick.  The topographic mapping, and particularly the roads on the topographic mapping, fit extremely well on the Broomfield base map, thus assuring very close spatial compatibility between the flood data and the base maps.

The Zone AE floodplains, which were based on District FHAD studies, were the most challenging and took the most District staff time to prepare for Merrick.  We color-coded the 100-year flood boundaries, thalwag and structures as we had with the Zone A floodplains.  We also color-coded and identified the cross sections that were shown on the FIRM (not all of the cross sections shown on the FHAD were shown on the FIRM).   Then we added to the source maps Base Flood Elevations (BFEs) at the same locations shown on the FIRM, floodways based on the width of the floodway at each cross section as contained in the "Floodway Data" tables in the Flood Insurance Study and interpolated between those cross sections to duplicate as closely as possible the shape of the floodway on the FIRM.  Finally, in some places a 500-year floodplain had to be added to the source maps.  This was difficult in some areas because the flood profiles in the Flood Insurance Study usually indicated that the 100- and 500-year flood profiles coincided and yet some how a larger 500-year floodplain had been drafted onto the FIRM (Figure 3).

The source maps for the LOMRs were handled in a similar manner.  These maps were also completed at a scale of 1-inch equals 100-feet with 2-foot contour intervals.   The District had most of these source maps available in its files, although the MCC provided work maps for two of the LOMRs and digital flood data for one LOMR.  In a couple of cases the LOMR was based on the construction of channels which contained the 100-year discharge, and the FIRM identified them as Zone A.  For these cases some of the additional information which was available on the work maps, such as BFEs and cross section locations, was not digitized.  For other LOMRs all of the flood data, including BFEs, cross section locations and floodways, were color coded and digitized.   The work maps for one LOMR could not be found by Broomfield, the District, the MCC or the consulting engineer who had originally prepared the LOMR request.  That reach of floodplain has yet to be incorporated into the DFIRM.

There were only two drainageways where we had to request the original FIS work maps.   Ironically, the work maps had been completed on USGS quad maps at a scale of 1-inch equals 2000-feet, whereas the FIRM showing these floodplains was published at a scale of 1-inch equals 1000-feet.  For this case we digitized the Zone A boundaries from the FIRM and fit them to the Broomfield base map.  We found that they fit reasonably well.

Observations

We believe that by digitizing the original source maps for each floodplain, we now have a more accurate portrayal of the flood hazards in relation to the Broomfield road map (Figure 4) and orthophotos (Figure 5) than was possible with the FIRM.  The map is now effectively accurate to 1-inch equals 100-feet rather than 1-inch equals 1000-feet.  Non-technical people can utilize the flood theme overlaid on the orthophotos to make judgments about the relationship between flood hazards and structures, which they were not able to perform in the past (Figure 5).  The DFIRM is definitely a superior product to the FIRM.

The cost to Broomfield for this project was to provide the base map and orthophoto information, which they had already developed for other purposes.  The cost to the District for the services of Merrick & Company was approximately $50,000.   The District's contribution was the staff time needed to obtain the source maps, color code them and perform the quality control checks.

During the quality control checks, when the flood data was overlaid on the orthophotos, we found a few cases where physical changes had obviously been made that had not been reflected in changes to the FIRM (Figure 6).  None of those physical changes were determined to have been inappropriate, but the necessary changes to the FIRM had not been processed.  Now that these areas have been identified we will be able to work with Broomfield to document the changes and revise the DFIRM accordingly.  In this case the conversion process allowed us to find these areas, and the new DFIRM will allow us to more easily and quickly make the necessary adjustments. 

Printable Version