Dr. Nelson
Overview
You have learned the basic modeling techniques and performed various sensitivity studies with the precipitation and infiltration methods in HMS and GSSHA. In this exercise we will examine how HMS develops the runoff hydrograph, which is often termed Transformation.
Our sensitivity studies will again center around the Judy's Branch Watershed and you can use the same data as you have previously. We will use different transformation methods and see how they affect output. You will perform the similar analysis with GSSHA (in the next assignment) and at the end you will compare the response of both HMS and GSSHA to different transformation methods/processes.
Model Set up
Use your basic Judy's branch WMS project with the SSURGO soil information. We have clipped the SSURGO soil shape file in ArcGIS so that it is smaller and easier to work with, but still completely overlaps your watershed. You can get it Here. The HYDGRP field is already joined and it is in th UTM NAD 83 coordinate system.
Use the land use data that you have been using. If you got confused with lots of data set you downloaded, get the Land use data from Here.
Create land use and soil type coverages in WMS and determine the curve number for the watershed as usual.
Processes
1. Transformation method comparison
In this part we will the see variation in output hydrograph as we change the transformation method. Create a HMS model with:
a. SCS Transformation Method
Watershed: Single Basin
Loss Rate Method: Green and Ampt ( Initial loss = 0.2 in, Moisture deficit = 0.2, suction = 12 in, Conductivity = 0.25 in/hr, Impervious = 0%)
Transform method: SCS
Lag time: Kerby Method for Overland Flow for Tc (WMS will convert this Tc to a Lag time as required by SCS)
Precipitation: 2.5 inches with a Type I temporal distribution.
Model Run: 36 hrs @ 6 mins interval
b. Clark Transformation Method
Watershed: Single Basin
Loss Rate Method: Green and Ampt ( Initial loss = 0.2 in, Moisture deficit = 0.2, suction = 12 in, Conductivity = 0.25 in/hr, Impervious = 0%)
Transform method: Clark
Time of concentration: Kerby Method for Overland Flow
Precipitation: 2.5 inches with a Type I temporal distribution.
Model Run: 36 hrs @ 6 mins interval
Compare the results for the two different methods.
2. Sensitivity Analysis of R in Clark Method
The Parameters for the Clark Transformation method are Time of Concentration tc and Storage Coefficient R. In HMS, open the project file that you used for 1b where you used Clark Method (you might want to save it as with new name so that you will have previous project unchanged). Now under Transform tab, leave the time of concentration the same but change the value of storage coefficient:
By increasing by 100%
By decreasing by 50%
Compare the outflow hydrographs with normal value (Determined by WMS in part 1b) and these two cases.
3. Sensitivity of Lag Time
The parameter for SCS Transformation method is lag time TLag. In WMS, open project the file you used for creating HMS file for 1a where you used SCS method. In the Edit Parameter Window, when you compute basin data for SCS Transformation method, change the Lag time equation as:
Use any four different Lag time or Time of Concentration (remember WMS will convert Tc to Tlag if you pick a Tc method) equations available in WMS basin data time computation calculator. Select the equations suitably so as to match the watershed property/Location/Land use and or soil type, but don't be surprised if there is some variation.
Save the HMS Project file (or you can calculate the lag time in WMS, take that value and replace it in HMS) and run for each case.
Compare the outflow hydrographs from all the different cases.
4. MODClark Method
The performance of MODClark method is affected by the parameters like initial abstraction, potential retention factor and grid size resolution. Along with these, the time of concentration and storage coefficient R also play a significant role. But as we already analyzed the sensitivity of time of concentration and R, we will not change these values. Here we will be analyzing the sensitivity of Initial abstraction, potential retention factor and grid size resolution.
From the WMS base project create a 100 by 100 MODClark grid cells. Enter all values to create a MODClark HMS project. Use following:
Watershed: MODClark Grids (100 by 100)
Loss Rate Method: Gridded SCS method
Initial abstraction ratio 0, 0.1 and 0.2
Potential retention factor: 1
Transform method: MODClark method
Time of concentration: Kerby Method for Overland Flow
Precipitation: 2.5 inches with a Type I temporal distribution. NOTE: You must define this as a user hyetograph and import the XY series that has the SCS distributions in them. Use the Type 1 Distribution and enter the 2.5 inches, WMS will multiply when saving.
Model Run: 36 hrs @ 6 mins interval
Run HMS for all three cases and compare the results
b. Now In WMS, delete the grid you had and create grids of 5 by 5, 25 by 25 and 200 by 200. Then each of those models use the following:
Watershed: MODClark Grids
Loss Rate Method: Gridded SCS method
Initial abstraction ratio 0.2
Potential retention factor: 1
Transform method: MODClark method
Time of concentration: Kerby Method for Overland Flow
Precipitation: 2.5 inches with a Type I temporal distribution. NOTE: You must define this as a user hyetograph and import the XY series that has the SCS distributions in them. Use the Type 1 Distribution and enter the 2.5 inches, WMS will multiply when saving.
Model Run: 36 hrs @ 6 mins interval
Run MODClark in HMS for all the different grid cells sizes. you already have results for 100 by 100. compare the results for four different grid resolution.
Comparison between Clark and SCS method (1a and b)
Analyze the sensitivity of R in Clark Method (2)
Analyze the sensitivity of different Lag time equations for SCS method (3).
Analyze the sensitivity of initial abstraction ratio of MODClark Method (4a)
Analyze the grid resolution sensitivity in MODClark Method (4b)