Hydrologic and Hydraulic Uncertainties

Purpose

In this hands-on workshop participants are introduced to HEC-FDA Version 2 to familiarize students with entering data for defining exceedance-probability and stage-discharge functions, and associated uncertainties.  This workshop pertains to the Hydraulic data and frequency functions.

Last Updated

Last Modified: 2026-07-28 13:18:39.803

Downloadable Materials

Workshop Overview

The goal of this workshop is to familiarize students with using HEC-FDA Version 2, including navigating the user interface for importing hydrologic data and for defining exceedance-probability and stage-discharge functions, and their associated uncertainties.

  • Steps already completed for the students include:
    • Creating the study
    • Importing the terrain and impact area set.
  • Students will complete the following steps in this workshop:
    • Import the hydraulic data
    • Enter the exceedance probability functions
    • Enter the stage-discharge functions with uncertainty

Introduction

The hydrologic and hydraulic data you're working with are from a section of the White River in Muncie, Indiana. The example data set draws on an example data set used in HEC-RAS Training. The 0.02 annual chance exceedance (50-year) floodplain is illustrated in the image below. 

HEC-RAS modeling for the without-project 2% annual exceedance probability (AEP) event.

Instructions

Task 1 – Open Study for Workshop 2

  1. Open the HEC-FDA Version 2 software. 
  2. From the File menu, click Open. The Open Study dialog opens as a tab.
  3. Click the ellipse button in the Study Path box to open the Select File browser window.
  4. Navigate to the Workshop 2 starting project (e.g., C:\Workshops\FDA\Workshops\Muncie), select the *.sqlite HEC-FDA file and click Open.
  5. From the Open Study tab, click OK.

Task 2 – Import Hydraulic Data

Considerations

Hydraulic modeling is used in HEC-FDA to identify the water surface elevation at each structure for, ideally, at least 8 unique events. This information is used in calculating the stage-damage functions for each structure. The structure-level stage-damage functions (aka consequence functions) are then aggregated to stages at the index point, which are defined by the combination of the input summary relationships. The hydraulics data can also be used to develop graphical stage-exceedance probability functions. Hydraulic engineers will use HEC-RAS or other hydraulic modeling software to develop the spatial data sets like the ones used in this workshop.

Steps

  1. From the Study Tree under Hydraulics right-click on the appropriate spatial data sets type, and click Import Hydraulics. The Import Hydraulics dialog opens as a tab.

    Test Your Knowledge - Question 1

    Which import option should be selected to import the hydraulic data?

    1. Unsteady HDF
    2. Steady HDF
    3. Gridded Data

    Review the Example Study Data to determine the correct spatial data set type for the Muncie, Indiana Hydraulic study data.

     

  2. From the importer enter a name for the hydraulics profile (e.g., Existing Conditions) and enter a description for the profile.
  3. From the Source File box select the ellipse button to open the Select File browser window.
  4. Navigate and select the folder containing the existing conditions hydraulic data (e.g., C:\Workshops\FDA\MuncieData\Hydraulics\Muncie_Existing_Conditions), and click Select Folder.
  5. The importer tab updates with a table displaying the name of the profiles, the annual exceedance probability and the return interval. The Annual Exceedance Probability will need to be corrected. Manually enter the appropriate AEP value for each profile name based on Table 1.

    Table 1
    NameAnnual Exceedance ProbabilityReturn Interval (Return Period)
    20%ACE0.2





    Updates based on the entered AEP using the equation (1/AEP)

    10%ACE0.1
    5%ACE0.05
    50%ACE0.5
    2%ACE0.02
    1%ACE0.01
    0.5%ACE0.005
    0.2%ACE0.002

    Copy and paste is not an option in the import table. Select each cell to type in the appropriate return interval.

  6. Note that the Return Interval column updates based on the AEP. Check that the AEP and return interval values are correct and click Save. The hydraulic data is added to the Study Tree.
  7. Click Close to close the Import Hydraulics tab.

Task 3 – Enter the Flow-Frequency Function

Steps

  1. From the Study Tree, right-click on Frequency Functions and select Create New Frequency Function. The Create Frequency Function opens as a tab.
  2. Enter a name for the function in the Name box (e.g., Flow-Frequency LPIII), and enter a description for the frequency function (e.g., Flow-Frequency LPIII distribution for a period of record (Record Length) of 48 years).
  3. Select Analytical and enter the LP3 Parameters listed below. 

    MeanStandard DeviationSkewRecord Length
    3.7070.24-0.475 (see Test Your Knowledge - Question 2)
    Test Your Knowledge - Question 2

    What is the correct Record Length (period of record) value for the flow-frequency LPIII distribution?

    Review the Example Study Data for Version 2 to determine the correct Record Length for the Flow Frequency Function.

  4. Enter the correct period of record in the Record Length box.
  5. Click Save. The flow-frequency function is added to the Study Tree.
  6. Click Close to close the Create Frequency Function tab.

Task 4 – Enter the Without-project Condition Stage-Discharge Function

Steps

  1. From the Study Tree, right-click on Stage-Discharge Functions and select Create New Stage-Discharge Relationship. The Create New Stage-Discharge Relationship opens as a tab.
  2. Enter a name for the relationship in the Name box (e.g., Existing Condition Relationship, Rating Curve Without Project, etc.), and enter a description for the relationship (e.g., This stage-discharge function represents the relationship between stage and discharge for the without-project condition.).

    Test Your Knowledge - Question 3

    What is the correct Distribution Type for the without-project stage-discharge relationship?

    Review the Example Study Data for Version 2 to determine the correct without-project condition stage-discharge function.

  3. Select Triangular from the Distribution Type dropdown box.
  4. Use the information in Table 3 to create the new stage-discharge relationships for the without-project condition. 
    Table 3

    Loading

  5. Click Save.
Test Your Knowledge - Question 4

Based on the data in Table 3, what is the maximum probable stage for a discharge of 15,185.46cfs?

Test Your Knowledge - Question 5

In the Study Tree, right-click on the frequency function you just created and select edit. Hit the pop-out icon to the left of the close file icon at the top of the tab. Click inside the plot to get the tooltip (yellow callout box) to inspect the plot a little closer (you can also use the mouse scroll wheel to zoom in or out). Use the plot of the flow-frequency function and the table for the rating curve to identify the most likely stage that has an approximate 0.4% chance of being exceeded. 

Review the Flood Risk Assessment and HEC-FDA for an example of how to work through this problem (specifically the Getting to Damage-Frequency section).

Test Your Knowledge - Question 6

Imagine you are evaluating flood risk over a two mile stretch of river. How will you decide which hydrologic and hydraulic summary relationships you'll require and for how many locations? 

Test Your Knowledge - Question 7

What is the methodology called for quantifying uncertainty about a graphical frequency function in HEC-FDA and how does it work?

Test Your Knowledge - Question 8

What is the methodology called for quantifying uncertainty about an analytical frequency function in HEC-FDA and how does it work?

Test Your Knowledge - Question 9

How do hydraulic modelers develop uncertainty estimates about the stage-flow function?

Test Your Knowledge - Question 10

HEC-FDA does not extrapolate. Why is this detail crucial for specification of summary relationships?

Test Your Knowledge - Question 11

What is the equivalent record length and how is it used?

Test Your Knowledge - Question 12

What types of distributions can you use to specify uncertainty in stage for the stage-discharge function? 

Test Your Knowledge - Question 13

If you're using stage-frequency to represent your hazard, can you also use a regulated-unregulated flow transform function?

Test Your Knowledge - Question 14

If you're entering a graphical frequency function, what probabilities should be represented? 

Conclusion

Download the solution project below to compare to your project:

Review the best answers to test your FDA knowledge questions and compare with yours to see how you did!

Return to top of page.

Best Answers to Test Your FDA Knowledge Questions
Test Your Knowledge - Question 1

Which import option should be selected to import the hydraulic data?

  1. Unsteady HDF
  2. Steady HDF
  3. Gridded Data
Test Your Knowledge - Question 2

What is the correct Record Length (period of record) value for the flow-frequency LPIII distribution?

Answer: 48.

Test Your Knowledge - Question 3

What is the correct Distribution Type for the without-project stage-discharge relationship?

Answer: Triangular.

Test Your Knowledge - Question 4

Based on the data in Table 3, what is the maximum probable stage for a discharge of 15,185.46cfs?

Answer: 947.38 feet.

Test Your Knowledge - Question 5

In the Study Tree, right-click on the frequency function you just created and select edit. Hit the pop-out icon to the left of the close file icon at the top of the tab. Use the toolbar tools (located on the right-side of the plot) to inspect the plot a little closer. Use the plot of the flow-frequency function and the table for the rating curve to identify the most likely stage that has an approximate 0.4% chance of being exceeded. 

Answer: approximately 947.22 feet. (Based on discharge of 17,023.02 at .004 AEP)

Test Your Knowledge - Question 6

Imagine you are evaluating flood risk over a two mile stretch of river. How will you decide which hydrologic and hydraulic summary relationships you'll require and for how many locations? 

Answer: First, investigate whether flow can be included in the frequency-based risk assessment model. Flow can be included where there is a one-to-one relationship between peak flow and peak stage that can be modeled with a simple graphical function with approximately 8 coordinates. In situations where there is not this one-to-one relationship such as situations with considerable backwater, flow cannot be included in the risk assessment model due to conflict with the frequency-based methodology. In situations where flow cannot be included, modelers depend on stage-frequency functions to entirely describe the uncertainty. Next, evaluate what measures exist or are being proposed, and how those measures are modeled. For example, channelization will be modeled using a stage-discharge function that reflects the lower channelized stage for a given discharge. Finally, identify for how many locations you will need to model the hydrologic and hydraulic summary relationship. A given relationship, identified at a point in the river, should be representative of the entire impact area, requiring that a given impact area reflect homogenous hydrologic and hydraulic conditions over the impact area. 


Test Your Knowledge - Question 7

What is the methodology called for quantifying uncertainty about a graphical frequency function in HEC-FDA and how does it work?

Answer: The methodology for quantifying uncertainty about a graphical frequency function is called the Less Simple Method. The Less Simple Method uses the slope of the frequency curve with the equivalent record length using an asymptotic approximation of the Normal distribution to calculate standard deviation of the hazard for each coordinate. 

Test Your Knowledge - Question 8

What is the methodology called for quantifying uncertainty about an analytical frequency function in HEC-FDA and how does it work?

Answer: The methodology for quantifying uncertainty about a Log Pearson Type III analytical frequency function is called the parametric bootstrap. The parametric bootstrap generates a sample from the original user-input Log Pearson Type III of the same size as the equivalent record length; then, the mean, standard deviation, and skew are calculated for the sample and used to create a new instance of a Log Pearson Type III distribution, allowing for variability in the mean, standard deviation, and skew. 

Test Your Knowledge - Question 9

How do hydraulic modelers develop uncertainty estimates about the stage-flow function?

Answer: Hydraulic modelers often create a triangular distribution of stages by varying manning's n, using a lowest estimate, a most likely estimate, and a highest estimate. 

Test Your Knowledge - Question 10

HEC-FDA does not extrapolate. Why is this detail crucial for specification of summary relationships?

Answer: Summary relationships must be specified to fully cover the modeled domain for proper concordance across the summary relationships. Issues with concordance occur when there exists poor overlap across model domain. For example, if a flow-frequency curve includes flows above 100,000 cfs for the 0.01 AEP but the stage-flow curve has a maximum flow of 80,000 cfs, then there will not be stages that concord to flows above 80,000 cfs and the combination of the two curves results in a truncated stage-frequency function and ultimately a biased estimate of EAD.  This concordance of modeled conditions is critical for proper counterfactual analysis with a well-defined intervention and because the risk assessment math depends on concordance.

Test Your Knowledge - Question 11

What is the equivalent record length and how is it used?

Answer: The equivalent record length is the number of years of a systematic record of recorded peak discharges at a stream gage. For probability functions derived at ungaged locations using model or other data, the equivalent record length is estimated based on the overall “worth” or “quality” of the frequency function expressed as the number of years-of-record. This parameter is important in risk-based analysis because it relates directly to the uncertainty of the probability function, and is used in the calculation of quantifying uncertainty about frequency functions. The greater the equivalent record length, the less uncertainty. 

Test Your Knowledge - Question 12

What types of distributions can you use to specify uncertainty in stage for the stage-discharge function? 

Answer: The available distributions are Normal, Triangular, Uniform, and Deterministic 

Test Your Knowledge - Question 13

If you're using stage-frequency to represent your hazard, can you also use a regulated-unregulated flow transform function?

Answer: No. If flow-frequency is not included in the risk assessment model, there is not capability to model the transformation of flow. 

Test Your Knowledge - Question 14

If you're entering a graphical frequency function, what probabilities should be represented? 

Answer: The probabilities entered should span the probability domain to the greatest extent practicable. Typically, users enter coordinates for probabilities between 0.5 and 0.002. However, if there is damage at the 0.5, then users must enter a coordinate for a probability that does not cause damage to avoid biasing EAD upward. 

Related Topics