The Frequency Duration is a variable in Hypothetical Storm modeling that needs to be considered.  The Frequency duration defines the length of time over which rainfall intensity and volume are accumulated and directly influences the temporal distribution of precipitation inputs used in hydrologic analyses. Accurate representation and selection of the storm duration can be important when generating hypothetical storms—inadequate consideration of storm duration can lead to underestimation of runoff volumes.  This analysis looked at comparing the 12-hr, 24-hr, 48-hr, and 96-hr durations.  The historic temporal distribution and depth area reduction were adjusted to match the duration.  

Initial Model:

SanLorenzo_durationalAnalysis_inital.zip

Excel Files:

Bootstrap_12hr.xlsx

Bootstrap_24hr.xlsx

Bootstrap_48hr.xlsx

Bootstrap_96hr.xlsx

Duration Results.xlsx

Review HEC-HMS Model

  1. Open project SanLorezno.hms and expand the Meteorologic Model in the Watershed Explorer.  Four Meteorologic Models have been created for the 48 hour duration.
  2. Expand the Paired Data folder in the Watershed Explorer.  Review the Percentage Curve, Area-Reduction Functions, and Parameter Value Samples for the 48 hour duration. 
  3. Head over to the Compute tab and review the Uncertainty Analysis compute for 1percentAEP-48hr.  The analysis has a Total Sample of 500 and three Parameters.  Parameter 1 samples rainfall depth. Parameter 2 samples temporal pattern, and Parameter 3 samples area-reduction curves.
  4. Run all four Uncertainty Analysis computes and ensure the compute completes.

Add Paired Data for Duration

  1. Navigate to the Watershed Explorer and expand the Paired Data Folder.  In the Parameter Value Samples, right click and select Create Copy for the 1percentAEP-48hr.  Name the copy 1percentAEP-12hr.  Repeat the copy again two more times and name the Paired Data 1percentAEP-24hr and 1percentAEP-96hr
  2. In the excel files corresponding to the duration, navigate to the Depths to Copy tab and select all 500 values in the 1% column. Copy and paste the values into Paired Data Parameter Value Sample table corresponding to the duration.  Repeat this for remaining durations. 

Graph showing 12-hr precipitation-frequency depthsGraph showing 24-hr precipitation-frequency depthsGraph showing 48-hr precipitation-frequency depthsGraph showing 96-hr precipitation-frequency depths

Add Percentage Curves

  1. Navigate to the Paired Data - Percentage Curves. Right click on 1998-48hr Percentage Curve and select Create Copy. Name the file 1998-12hr.  Repeat this process for the remaining Percentage Curves.
  2. Select the 1998-12hr Percentage Curve and choose the DSS Pathname //1998-12hr/PERCENT GRAPH///TABLE/.  Repeat this process for the remaining Percentage Curves.  
  3. Repeat Steps 1 and 2 for the 24 hour and 96 hour durations.  

While the 1982 and 2023-1st storm patterns were included in the previous tutorial, they were excluded from this analysis. Both storms had relatively short durations (approximately 48 hours) and, therefore, would not be suitable for use with the 96-hour rainfall depth.  Approximately 30% of the temporal pattern within a 96-hr period contained no rainfall. Applying a rainfall depth intended for a longer duration to a shorter storm pattern can artificially inflate the rainfall intensity within that pattern. This over-scaling may result in unrealistically high peak outflows that do not accurately represent the storm’s true hydrologic response.  For this analysis, a threshold was set requiring that rainfall occur for at least 25% of the time within a 96-hour duration.

Add Area-Reduction Function

  1. Navigate to the Paired Data Area-Reduction Functions. Right click on 1998-48hr Area-Reduction Functions and select Create Copy. Name the file 1998-12hr.  Repeat this process for the remaining Area-Reduction Functions.
  2. Select the 1998-12hr Area-Reduction Function and choose the DSS Pathname //1998-12hr/AREA-REDUCTION FACTOR///TABLE/.  Repeat this process for the remaining Area-Reduction Functions.  
  3. Repeat Steps 1 and 2 for the 24 hour and 96 hour durations.  

Create Meteorologic Models for Each Duration

  1. Navigate back to the Meteorologic Model in the Watershed Explorer. Right click and copy the 1percentAEP-48hr.  Rename the model 1percentAEP-12hr
  2. In the Hypothetical Storm Component Editor, change the Storm Pattern to 1998-12hr, Storm Duration to 12, and Area Reduction Function to 1998-12hr.  
  3. Confirm the 1percentAEP-12hr Meteorologic Model can work with the UncertaintyBasinModel
  4. Repeat Steps 1 and 2 for the remaining durations.  
The Point Depth, Storm Pattern, and Area-Reduction Functions will be sampled from the Parameter Value Sample and so the actual Paired Data selection does not matter.  The Storm Duration does matter and must be changed to apply the Percentage Curve and Area-Reduction Functions correctly.    

Create Parameter Value Sample for Durations

  1. Navigate back to the Watershed Explorer Paired Data folder and expand the Parameter Value Samples.  Right click on AllTemporalPatterns-48hr and select Create Copy.  Name the copy AllTemporalPatterns-12hr.  Repeat this for the remaining durations.
  2. Select AllTemporalPatterns-12hr and click on the Table tab.  Under the Value column, change the value to the corresponding 12 hour patterns.  Repeat for the other durations.
  3. Right click on AllAreaReduction-48hr and select Create Copy.  Name the copy AllAreaReduction-12hr.  Repeat this for the remaining durations.

Create Uncertainty Analysis for Durations

  1. Navigate to the Compute tab. Right click on 1percentAEP-48hr and Create Copy.  Rename the simulation 1percentAEP-12hr.  
  2. In the 1percentAEP-12hr Component Editor, change the Meteorologic Model to 1percentAEP-12hr.  Click the cog wheel next to Analysis Points.  In the Results window, make sure the AbvBigTrees - Precipitation and BigTrees - Outflow is checked.
  3. Select Parameter 1 and change the Parameter Value to 1percentAEP-12hr
  4. Select Parameter 2 and change the Parameter Value to AllTemporalPatterns-12hr
  5. Select Parameter 3 and change the Parameter Value to AllAreaReduction-12hr
  6. Repeat Steps 1 - 5 for the 24 hour and 96 hour.
  7. Navigate to the Compute | Multiple Compute.  Check the 1percentAEP analysis to run and select Compute.  The compute will run all 4 Uncertainty Analysis computes.

Review Results

  1. Navigate to the Results tab and expand the 1perentAEP-12hr node. 
  2. Review the Results Parameter table.  These tables show the values used for the Monte Carlo analysis in the order it was computed.
  3. Expand the Abv_BigTrees node and select the Precipitation Total table. Select the Parameter 1 and review the precipitation depths.  The precipitation depth from the Parameter 1 table is original frequency depth.  The Precipitation Total table is the precipitation depth after area reduction is applied.
  4. Expand BigTrees node and select Maximum Outflow.  Copy all 500 values under the Statistic Value column.  Open the Duration Results.xlsx spreadsheet and navigate to the Paste tab.  Under the 12-hr column, paste the outputs from HEC-HMS.  
  5. Repeat Steps 1-3 for the remaining durations.  
  6. In the excel spreadsheet Box and Whisker - Histogram tab, a plot of the outflows for each duration is shown.  

Which duration is the most appropriate duration for San Lorenzo?

The graph indicates that using the 12-hour rainfall depth would likely lead to underestimation, as its values are consistently lower than those of the 24-hour, 48-hour, and 96-hour durations. The 24-hour, 48-hour, and 96-hour depths have similar mean and median values. Both the 24-hour and 96-hour durations display a wider range of peak outflows, while the 48-hour duration has a higher lower quartile and min value. Overall, these three durations tend to produce comparable peak flows. When determining the appropriate storm duration for flood estimation, it is more accurate to use the typical storm duration that actually produces flooding rather than relying solely on a basin characteristic like the time of concentration. In this case, the basin’s time of concentration is approximately 4–5 hours, which would have underestimated outflow. This tutorial selects the 48-hour duration because it offers a relatively narrow interquartile range of outflow values and a higher minimum outflow.

Final Model: SanLorenzo_durationAnalysis_final.zip

Excel File: Duration Results_filled.xlsx

Continue to Hypothetical Storm Pattern Comparison