Summary

Depth grids were modeled using stream gauges for the 2017 California/Nevada flooding disaster event.

Data Sources

NextMap Elevation Data (5m spatial resolution)


NOAA NWS AHPS Stream Gauges

Methodology

  1. Obtain gage points: Initial gage point data was obtained on Monday, January 10th for Nevada and California.  Gage data was selected for the streams/rivers of interest.
  2. Collect “gage 0” information: Using the link found in the attribute table, each gage was checked for complete “gage 0” information so that stage readings could be referenced to mean sea level.  Gages with no information or incomplete “gage 0” information were omitted.
  3. Collect updated stage readings: Again, using the link in the attribute table the stage readings were updated on a daily basis until all gages had reported a crest at their location.  Gages that stopped reporting or malfunctioned during the course of the exercise were eliminated.  Gages that had not crested were populated with the highest recorded stage until a crest occurred at which time they were no longer updated.
  4. Calculate water surface elevations: Once updated stage readings were recorded the “gage 0” value was added to the reading to standardize the water surface elevations.  For example, a particular gage may read 26.1 feet and may have a “gage 0” value of 7.1 feet.  The water surface elevation at that gage relative to mean sea level would be 33.2 feet (26.1+7.1=33.2).  Additionally, gages utilizing the NGVD29 datum were converted to NAVD88 by adding 3.6 feet to the water surface elevation.
  5. Create rough flow lines in order to create gage reinforcement points: Stream/river gages can be sparse in some areas.  When there are great distances between gage points the values located in between those points can be problematic.  In order to help prevent the interpolation process from creating these problematic areas gage reinforcement points are created.  The first step to create these points is to look at the existing gages in relation to the streams and rivers.  In locations where a stream or river has more than one gage located along its path, a rough flow line is created to represent the stream path.  This can also be done where a tributary with one gage along it flows into a stream with an existing flow line.  Flow lines cannot be created for streams or rivers with only one gage if that stream does not flow into another stream with a flow line.  Flow lines also cannot be used if there is a dam on a river between gage locations.
  6. Create gage reinforcement points: Once the rough flow lines have been created gage reinforcement points can be generated by creating points at equal intervals along the flow lines.  The gage reinforcement points are then populated with interpolated values based on the distance between the upstream and downstream gage values.
  7. Since the streams in this event were relatively far apart, the stream lines were buffered and the gage points and reinforcement points were converted to cross-sections which spanned the buffered area perpendicular to the stream line.
  8. Create TIN: Using the water surface elevations at the cross sections and the stream buffer, generate a TIN to cover the affected area with interpolated water surface elevation values.
  9. Convert TIN to a raster: In order to generate a depth grid the TIN needs to be converted to a raster image.  In order to ensure the accuracy and correctness of the final depth grid the TIN is converted to a raster that matches the resolution of the ground elevation data.  The output raster was therefore set to 10 meters cells and was also set to snap to the terrain DEM to ensure that all grid cell edges aligned.
  10. Calculate depth: Using the “Minus” tool, the ground elevation was subtracted from the water surface elevation (water surface elevation – ground elevation = flood depth).
  11. Identify flooded areas: Use the “Greater Than” tool to classify the depth grid into areas greater than zero and areas equal to or less than zero. Areas in the output raster identified as being greater than zero are assigned a value of 1, areas equal to or less than zero are assigned a value of 0.
  12. Generate flood extents polygon: Convert the raster output from the “Greater Than” tool to a polygon feature class. Delete all features where gridcode=0 which indicates that it is not flooded.  All remaining polygon features represent the approximate flood extents.
  13. Eliminate disconnected false flooding areas: Due to the way a TIN interpolates values there may be disconnected areas that are falsely identified as flooding. These areas a usually low lying areas far from the flooded streams with no gage data along them.  In order to remove these areas a spatial selection is performed on the polygon feature class that represents the flooding extents using the gage and gage reinforcement points.  All polygons that intersect a gage or gage reinforcement point are kept, while all polygon features that do not intersect a gage or gage reinforcement point are deleted.
  14. Produce final depth grid: In order to produce a final depth grid that only includes flooded areas (and no negative depth values) it is necessary to remove the areas that did not flood.  In order to do this we use the “Extract by Mask” tool to extract the portion of the depth grid created in step 9 that is overlapped by the polygons created in step 11 that represent the approximate flood extents.
  15. Repeat: This process will be repeated daily until all gages report a crest.  Gage readings will be updated every morning and affected gage reinforcement points will also be recalculated.  Once the values for the gages and gage reinforcement points were updated the process would need to be run through again starting at step 7.

Quality Control

Quality control is included at every step in the process.  This ranges from excluding gage points that lack complete “gage 0” information to performing spot checks on the data at every step.  Below is a listing of techniques used to verify the data and outputs.

  1. Eliminate gages without complete “gage 0” information
  2. Eliminate gages which are not reporting or appear to have malfunctioned.
  3. Double check all gage reading values especially when updating every morning
  4. Double check all gage reinforcement point values after updating to ensure they make sense in relation to the upstream and downstream gages they are based on.
  5. Perform visual inspection of the water surface TIN. The TIN makes it easy to identify problem spots where a value has been incorrectly entered.  This will manifest itself as either a very high point or a very low point that looks out of place among the rest of the data.
  6. Compare raster cells after converting the TIN to a water surface elevation grid to ensure that the cells size matches the ground DEM and that the cells are properly snapped.
  7. After calculating depth perform spot checks to ensure that water surface value and ground elevation value do equal the depth value shown.
  8. Examine flood extents polygons and compare to ground elevation data to identify anomalies that may suggest a source data problem (areas of more widespread flooding or areas where flooding seems unusually small).

Final Products

All delivered geodatabases include the following layers:

  • Depth grid
  • Approximate Flood Extents polygon feature class
  • Analysis Footprint polygon feature class
  • Gage points used (includes stage readings, “gage 0” value, and calculated water surface elevation)
  • Areas of uncertainty

Access & Use Information

Public: This dataset is intended for public access and use.

Downloads & Resources

Carson River Download


Truckee River Download

Contact

Disclaimers

References


 

Tags: , , ,