Science Center Objects

Accurate estimations of flood discharges at bridge or culvert sites is required to provide cost-effective design of that structure. Streamflow-gaging stations, for which flood data are available, are usually located in major perennial drainage basins that are not representative of sites where common bridge and culvert designs are needed.

In 1942, the USGS, in cooperation with the New Mexico Department of Transporation, began operating a network of crest-stage gaging stations to provide and increase coverage of flood-frequency information used to estimate flood characteristics at ungaged sites.

The New Mexico Department of Transportation (NMDOT) has the responsibility for the design of bridges and culverts for New Mexico’s highways. Flood characteristics such as the magnitude and frequency of peak discharges are a major consideration in the design process. Accurate determination of flood characteristics at a bridge or culvert can contribute greatly to the cost-effective design of that structure. However, in New Mexico there is a general lack of consistent, long-term peak-discharge data concerning floods, making it necessary to develop methods to accurately estimate the magnitude and frequency of peak discharge at many sites. Variations in regional storm patterns and basin characteristics result in significant variations of peak discharge temporally and spatially. Streamflow-gaging stations, for which flood data are available, are usually located in major perennial drainage basins that are not representative of sites where the most common bridge and culvert designs are needed. Therefore NMDOT, in cooperation with the U.S. Geological Survey (USGS) operates and maintains a network of crest-stage streamflow-gaging stations predominately in ephemeral drainage basins, to provide an increased area where flood-frequency information can be used to estimate flood characteristics at ungaged sites. The flood-analysis program of New Mexico began in 1951 as a joint effort between the NMDOT and the USGS. This program provided a cost effective means for obtaining flood data and interpretive information that helped benefit both agencies to accomplish their missions. The program initially consisted of 10 crest-stage gaging stations to supplement annual peak discharge information collected at continuous-record streamflow-gaging stations, mainly at sites with drainage areas of less than 50 square miles. The network gradually expanded to about 145 crest-stage gaging stations by 1965. Floods that occurred in the Canadian and Rio Hondo River basins in 1965 increased the awareness of severe flood potential, which resulted in an expansion of the network from 1966 to 1969 to include flood-hydrograph gaging stations for rainfall-runoff modeling. The collection of continuous records of discharge and rainfall at 41 stations for calibration of a deterministic rainfall-runoff model continued through 1978. Subsequent analyses showed that large variances existed in the modeling effort, and the use of rainfall-runoff modeling to generate long-term synthetic flood discharge records was suspended.


The general lack of systematic, long-term peak-discharge data at most streamflow-gaging stations in New Mexico substantiate the need for continual data collection and the improvement in the frequency analysis of peak discharge for bridge and culvert design. The average period of record for New Mexico streamflow-gaging stations is 56 years. This time period includes two extreme flood events: (1) the 1965 flood, when severe flooding occurred in the eastern half of New Mexico, and (2) the 1904 flood, which also affected eastern New Mexico, but was about twice the magnitude of the 1965 flood. The stage-discharge relations at the crest-state gaging stations require continual improvement. Insufficient indirect discharge measurements exist for some stations and require updated stage-discharge ratings. Initially, peak discharges at crest-stage gaging stations were determined by slope-area measurements. Later, new hydraulic techniques were developed for theoretical stage-discharge ratings. Current USGS policy is to prepare stage-discharge ratings using step-backwater techniques and update the ratings as more flood information becomes available. Regional regression equations have unexplained variation that can be reduced with the inclusion of statistically significant variables. One method to reduce this variation is to develop additional or new basin and climatic characteristics. One basin characteristic is active-channel width. Mean annual discharge and flood discharges with selected recurrence intervals were significantly related to active channel width, which reduced the variation in regional equations for the western United States. Equations for New Mexico that reduced the variation by use of active-channel width were developed; but, only 79 streamflow-gaging stations in the State were used. These investigations indicate that active- channel width data for more streamflow-gaging stations may improve regional flood-frequency analysis. Highway department engineers use updated flood-frequency regression equations to ensure that design criteria are adequate when new flood information is acquired. Currently, USGS policy is to provide updated flood-frequency information every10 years using 10 or more years of gaging-station record. Peak discharge for the period of record was recorded at many crest-stage gages during water year 1999 and 2004; additionally substantial flood discharges were measured at continuous record streamflow-gaging stations.


The primary objective of this program is to provide updated flood flow frequency information for estimation of peak discharges that can be readily incorporated into design of culverts and bridges by engineers. Objectives of the 2006 – 2010 project plan are to:

  1. Provide updated peak discharge data at crest-stage gaging stations.
  2. Provide analysis of floods at miscellaneous bridge waterway sites.
  3. Provide regression equations to estimate flood characteristics based on basin/climatic variables of unregulated gaging stations.
  4. Provide world-wide web based streamflow statistics at gaging stations and ungaged sites.


The Federal Highway Administration (FHWA) and American Association of State Highway and Transportation Officials (AASHTO) have adopted guidelines that recommend the use of regional regression equations developed by the USGS to estimate flood discharges for the design of bridges and culverts at ungaged sites with drainage basins greater than 5 square miles. The guidelines also recommend the use of unit-hydrograph methods to estimate flood discharges in basins with areas greater than 5 square miles, as a secondary method for comparison.

Objectives of the Fy 2006-2010 program will be achieved by:

1. Collection of annual peak discharge data at 120 crest-stage gages. Twenty crest-stage gages will be reactivated on July 1, 2005. Twenty new crest-stage gages installations will occurr during FY 2006 – 2007.

2. Collect flood-frequency data at streamflow-gaging stations on unregulated streams for use in updating regression models. New basin/climatic characteristics will be updated and developed using USGS 7.5 minute (30 meter) Digital Elevation Models (DEM). The USGS GIS-Weasel and the National Elevation Dataset (NED) will be used to update basin characterisitcs. The updated NOAA Atlas 14 precipitation raster data will be used to update average basin precipitation. USGS 7.5-minute digital elevation models (DEM) have been developed into the National Elevation Dataset (NED), which is a more accurate dataset. Mean monthly precipitation (1961-90) raster data also have been developed for the United States. The raster data now includes high-elevation precipitation data from the NRCS and the National Weather Service. NOAA Atlas 14, an update of Atlas 2, presents 24-hour and 6-hour maximum precipitation for various recurrence intervals, which also includes high-elevation precipitation data from the NRCS. USGS has developed the GIS Weasel, an interactive program that uses the NED. The GIS Weasel, in conjunction with raster modeling, will be used to determine significant variables in regional flow frequency regression equations.

Average basin slope is the most promising variable expected to improve regional-regression equations because flood response is expected to be associated with basin slope. Any improvements from the average basin 24-hour or 6-hour maximum precipitation (NOAA Atlas 14) are unknown for New Mexico at this time. The latest flood-flow frequency report used a point interpolation at the gaging station location to determine variables, whereas the new variables will be determined by average basin values.

3. StreamStats is a national USGS Web application for streamflow information. Streamflow and basin/climatic characteristics for streamflow-gaging stations will be compiled into a Access data base. Streamflow characteristics will be updated using USGS application programs. Basin and climatic characteritics have been updated for unregulated gaging stations and are available for regulated gaging stations. Regression models for low-flow frequency, flood-flow frequency, and average discharge are available for detemining streamflow characteritics at ungaged sites. GIS coverages or layers are available from prior effort.