You are on page 1of 8

Workshop 3 Solution

Calibrating a Continuous Simulation Model


Disclaimer: The United States Army Corps of Engineers has granted access to the
information in this model for instructional purposes only. Do not copy,
forward, or release the information without United States Army Corps of
Engineers approval.

Task 1. Calibrate the EF Russian River Deficit & Constant basin model for WY 2005.

Question 1: Populate Table 1 with summary results from your calibrated model.

Summary results for calibrated simulation “WY 2005 DC” are presented in Table 1.

Computed Observed
Peak Discharge (CFS) 6254.2 6552.5
Volume (AC-FT) 231497.9 213599.1
Table 1. Summary Results for Calibrated EF Russian 20 Subbasin

Question 2: The Nash-Sutcliffe efficiency for your calibrated model is ~0.830.


Question 3: Populate Table 2 with calibrated Deficit and Constant parameters for
Subbasin EF Russian 20.

Table 2. Calibrated EF Russian 20 DC” Basin Parameters for Simulation “WY 2005 DC”
Element Parameter Value
Canopy Crop Coefficient 0.3
Loss Initial Deficit (IN) ~6.5
Maximum Deficit (IN) ~6.5
Constant Rate (IN/HR) ~0.125
Transform Time of Concentration (HR) 5.5
Storage Coefficient (HR) 4.5
Baseflow GW 1 Coefficient (HR) ~28
GW 1 Reservoirs 1
GW 2 Coefficient (HR) 400
GW 2 Reservoirs 1

1
WS – Calibrating a Continuous Simulation Model/Brauer
Task 2. Calibrate the EF Russian River Soil Moisture Accounting basin model for WY
2005.

Question 4: Populate Table 3 with summary results from your calibrated model.

Summary results for calibrated simulation “60yr SMA” are presented in Table 3.

Computed Observed
Peak Discharge (CFS) 6236.4 6552.5
Volume (AC-FT) 201815.6 213599.1
Table 3. Summary Results for Calibrated EF Russian 20 Subbasin

Question 5: The Nash-Sutcliffe efficiency for your calibrated model is ~0.905.


Question 6: Populate Table 4 with calibrated Soil Moisture Accounting
parameters for Subbasin “EF Russian 20.”

Table 4. Calibrated “EF Russian 20 SMA” Basin Paramters for Simulation “WY 2005 SMA”
Element Parameter Value
Canopy Crop Coefficient 0.7
Loss Soil Storage (IN) 10
Tension Storage (IN) ~8
Max Infiltration Rate (IN/HR) ~0.35
Soil Percolation (IN/HR) 0.1
GW 1 Storage (IN) 4
GW 1 Percolation (IN/HR) 0.1
GW 1 Coefficient (HR) 24
GW 2 Storage (IN) 10
GW 2 Percolation (IN/HR) 0.001
GW 2 Coefficient (HR) 150
Transform Time of Concentration (HR) 5.5
Storage Coefficient (HR) 4.5
Baseflow GW 1 Coefficient (HR) 0.5
GW 1 Reservoirs 1
GW 2 Coefficient (HR) 24
GW 2 Reservoirs 1

2
WS – Calibrating a Continuous Simulation Model/Brauer
W2 - Task 3

Task 3. Calibrate the EF Russian River Deficit & Constant basin model to observed
annual maximum flow values at Calpella.

The objective of this calibration is to have computed annual maximum flows reproduce
observed annual maximum flows. On average there should be a 1:1 relationship
between computed and observed values. The scatter plot in Figure 1 shows the
relationship between computed and observed values. A best fit line with slope intercept
set equal to zero is plotted next to a line with slope = 1 to provide a graphical indication
of fit. The equation of the best fit line provides a quantifiable indication of fit.
Question 7: The calibrated value for “Constant Rate” is 0.11 in/hr.

Figure 1. Calibrated Annual Maximum Flows for Deficit and Constant Basin Model

3
WS – Calibrating a Continuous Simulation Model/Brauer
W2 - Task 3

Question 8: Is the calibration from the 1 year simulation (Task 1) a good predictor
for the 60 year simulation (Task 3)?
Yes, calibrated values from a shorter duration continuous simulation often provide a
good indicator for a longer duration continuous simulation. For really long simulations, it
may be beneficial to calibrate a shorter duration continuous simulation prior to
calibrating for the longer duration. The shorter duration allows the modeler to take
advantage of faster run times while familiarizing with the mechanics of the study area
and modelling methods.

4
WS – Calibrating a Continuous Simulation Model/Brauer
W2 - Task 3

Task 4. Calibrate the EF Russian River Soil Moisture Accounting basin model to
observed annual maximum flow values at Calpella.

Question 9: The calibrated value for “Maximum Infiltration” is ~ 0.30 – 0.39 in/hr.
Figure 2 shows a scatter plot for the calibrated Soil Moisture Accounting model.

Figure 2. Calibrated Annual Maximum Flows for Soil Moisture Accounting Basin Model

Question 10: Is the calibration from the 1 year simulation (Task 2) a good
predictor for the 60 year simulation (Task 4)?
Yes, calibrated values from a shorter duration continuous simulation often provide a
good indicator for a longer duration continuous simulation. For really long simulations, it
may be beneficial to calibrate a shorter duration continuous simulation prior to
calibrating for the longer duration. The shorter duration allows the modeler to take
advantage of faster run times while familiarizing with the mechanics of the study area
and modelling methods.

5
WS – Calibrating a Continuous Simulation Model/Brauer
W2 - Task 3

Additional Task 1. Calculate performance metrics for calibrated “EF Russian 20


DC” and “EF Russian 20 SMA” basin models based on daily flows.

Summary statistics are used to quantify model performance compared to observations.


Statistics include Nash-Sutcliffe Efficiency (NSE), Observed Standard Deviation Ratio
(RSR), Coefficient of Determination (R2), and Percent Bias (PBIAS).

Question 11: Populate Table 5 with the summary statistics calculated in the
Calibration App.
Summary statistics for simulations “60yr DC” and “60yr SMA” are presented in Table 5.

Table 5. Summary Statistics for EF Russian River


Simulation NSE RSR R2 PBIAS
60yr DC 0.849 0.388 0.882 -8.489
60yr SMA 0.856 0.379 0.886 -1.42

Summary statistic performance ratings based on Moriasi et al (2007) are presented in


Table 6.
Table 6. Performance Ratings for Summary Statistics
Performance
NSE RSR R2 PBIAS
Rating
Very Good 0.65<𝑁𝑁𝑆𝑆𝐸𝐸≤1.00 0.00<𝑅𝑅𝑆𝑆𝑅𝑅≤0.60 0.65< 𝑅𝑅2≤1.00 𝑃𝑃𝐵𝐵𝐼𝐼𝐴𝐴𝑆𝑆< ±15
Good 0.55<𝑁𝑁𝑆𝑆𝐸𝐸≤0.65 0.60<𝑅𝑅𝑆𝑆𝑅𝑅≤0.70 0.55<𝑅𝑅2≤0.65 ±15≤𝑃𝑃𝐵𝐵𝐼𝐼𝐴𝐴𝑆𝑆<±20
Satisfactory 0.40<𝑁𝑁𝑆𝑆𝐸𝐸≤0.55 0.70<𝑅𝑅𝑆𝑆𝑅𝑅≤0.80 0.40<𝑅𝑅 ≤0.55
2
±20≤𝑃𝑃𝐵𝐵𝐼𝐼𝐴𝐴𝑆𝑆<±30
Unsatisfactory 𝑁𝑁𝑆𝑆𝐸𝐸≤0.40 𝑅𝑅𝑆𝑆𝑅𝑅>0.80 𝑅𝑅2≤0.40 𝑃𝑃𝐵𝐵𝐼𝐼𝐴𝐴𝑆𝑆≥±30

Question 12: Based on the summary statistics which calibrated basin model
better models average daily flow?

The Soil Moisture Accounting basin model performs slightly better on summary statistics
although both basin models fall within the “Very Good” range for all summary statistics.

6
WS – Calibrating a Continuous Simulation Model/Brauer
W2 - Task 3

Additional Task 2. Create a Bulletin 17B flow frequency curve using HEC-SSP.
Figure 3 shows the flow frequency curve for simulation “60yr SMA” in HEC-SSP.

Figure 3. Bulletin 17 Flow Frequency Curve in HEC-SSP

7
WS – Calibrating a Continuous Simulation Model/Brauer
W2 - Task 3

Figure 4 shows view the computed flow-frequency curve plotted against the observed
flow-frequency curve for simulation “60yr SMA”.

Figure 4. Flow-Frequency Curve Plotted in Excel

Question 13: For a study analyzing rare flood events, where is it most critical for
the computed curve to reproduce the observed?

For rare flood events it is important for the high flow end of the flow frequency curve to
reproduce the observed. In this case the model performs well across the spectrum of
flows but while fine tuning during calibration emphasis was placed on the high flow end
of the curve rather than middle or low flow portions.

8
WS – Calibrating a Continuous Simulation Model/Brauer

You might also like