Historical Data Period
1980–1990
Eleven annual momentary peak-flow observations were available for the statistical flood analysis.
HEC-HMS · Engineering check
Validation & 1995 Comparison
The original 1995 Hare Irrigation Project estimated design floods from the available annual momentary peak-flow record using statistical frequency analysis. The present study independently reconstructs flood response using a GIS-based HEC-HMS rainfall-runoff model.
The objective of this comparison is not to force HEC-HMS to reproduce the historical values. Instead, the two approaches are compared to understand their agreement, differences and uncertainty, and to establish a defensible hydrologic basis for reassessing the Hare diversion works.
Important distinction:
This comparison is an engineering validation check rather than a formal
calibration against observed hydrographs. The 1995 estimates were derived
statistically from observed annual peak discharges, while HEC-HMS
reconstructs runoff from rainfall, catchment properties and routing.
1. Historical 1995 Flood Analysis
The 1995 study used a series of annual momentary peak flows covering 1980–1990. The eleven annual maximum observations were subjected to flood-frequency analysis to estimate peak discharge for selected return periods.
The historical analysis considered several approaches, including Log-Pearson Type III. The Log-Pearson results provide the most useful return-period benchmark for direct comparison with the present HEC-HMS design events.
Frequency Method
Log-Pearson Type III
Used to estimate design peak flows associated with selected statistical return periods.
2. 1995 Design Peak-Flood Benchmarks
| Return Period | 1995 Design Peak Q (m³/s) | HEC-HMS Q (m³/s) | Difference | Status |
|---|---|---|---|---|
| 5 years | 25.289 | — | — | HEC-HMS result not yet finalised |
| 10 years | 33.372 | — | — | HEC-HMS result not yet finalised |
| 25 years | 46.655 | — | — | HEC-HMS result not yet finalised |
| 50 years | 59.183 | 114.4 |
+55.217 m³/s
+93.3% |
Completed comparison |
| 100 years | 74.397 | — | — | HEC-HMS result not yet finalised |
Current comparison:
The completed modern design run available for direct comparison is the
50-year HEC-HMS event. Other return periods should only be added after
their final simulations have been accepted.
3. 50-Year Design Flood Comparison
The most important completed comparison at the present stage is the 50-year event. The original 1995 Log-Pearson Type III analysis estimated a 50-year design peak of 59.183 m³/s.
The final HEC-HMS rainfall-runoff simulation gives a peak discharge of 114.4 m³/s at Sink-1, representing the Hare Weir and irrigation headworks.
1995 Q50
59.183 m³/s
Log-Pearson Type III frequency estimate based on the historical annual momentary peak-flow series.
Modern HEC-HMS Q50
114.4 m³/s
Event-based rainfall-runoff simulation for the reconstructed 162.035 km² Hare watershed.
Difference
+55.217 m³/s
The HEC-HMS estimate is approximately 93.3% higher than the historical Q50.
4. Why the Two Estimates Differ
A difference between the historical statistical estimate and the modern rainfall-runoff simulation is not unexpected. The two approaches use fundamentally different information and assumptions.
1995 Statistical Method
The historical result was derived from a relatively short record of annual momentary peak discharges.
- 11 annual peak observations
- 1980–1990 record period
- Statistical extrapolation to larger return periods
- No explicit spatial rainfall-runoff representation
- No explicit subbasin or reach routing model
Modern HEC-HMS Method
HEC-HMS reconstructs the physical rainfall-runoff process over the entire watershed.
- 162.035 km² GIS-delineated watershed
- Five separate subbasins
- Design rainfall applied explicitly
- SCS Curve Number rainfall losses
- SCS Unit Hydrograph transformation
- Muskingum channel routing
5. Influence of the Historical Record Length
The historical flood-frequency calculation was based on only eleven annual peak-flow observations. Such a record can describe the floods that occurred during the observed period, but estimates of much rarer events require statistical extrapolation beyond the measured record.
This becomes increasingly important for the 25-, 50- and 100-year return periods. The historical frequency curve therefore remains valuable evidence, but the uncertainty associated with extrapolating a short record should be recognized when the values are used for hydraulic design.
Observed context:
The largest annual momentary peak in the 1980–1990 historical series was
approximately 41.45 m³/s. Both the historical
Q50 and Q100 therefore depend on extrapolation
beyond the maximum observed annual peak.
6. Difference in Hydrologic Representation
HEC-HMS introduces information that was not explicitly represented in the historical frequency calculation. The modern model accounts for the spatial arrangement of the watershed and simulates the movement of runoff through the drainage system.
| Characteristic | 1995 Frequency Analysis | Modern HEC-HMS |
|---|---|---|
| Primary input | Observed annual peak discharge | Design rainfall |
| Catchment representation | Statistical flow series | Five GIS-derived subbasins |
| Rainfall losses | Not modelled explicitly | SCS Curve Number |
| Runoff transformation | Not modelled explicitly | SCS Unit Hydrograph |
| Channel routing | Not modelled explicitly | Muskingum routing |
| Primary output | Return-period peak discharge | Complete flood hydrograph |
| Spatial land-use influence | Indirectly reflected in historical flow | Explicitly included through weighted CN |
7. Interpretation of the 50-Year Difference
The modern HEC-HMS Q50 is approximately 1.93 times the historical 1995 Log-Pearson estimate. This difference should not automatically be interpreted as evidence that either analysis is incorrect.
Instead, the difference demonstrates the sensitivity of design-flood estimation to the available data, frequency-analysis record length, adopted design rainfall, runoff-loss assumptions, watershed response times and channel-routing representation.
For the present Hare reconstruction, the HEC-HMS result is retained as the modern rainfall-runoff estimate while the 1995 value is preserved as an important historical benchmark.
Engineering principle:
The modern model is not adjusted simply to make its peak discharge equal
to the 1995 value. Any modification of rainfall, Curve Number, lag time
or routing parameters must be supported by hydrologic or physical
evidence.
8. Historical Terminology
The original study used the expression “maximum probable flood” when discussing some frequency-based flood estimates. That wording is retained when referring directly to the historical document.
In the present reassessment, however, return-period results are described more specifically as:
T-year design peak flood, QT
This avoids confusing a statistical 50- or 100-year flood estimate with the technically distinct concept of a Probable Maximum Flood (PMF).
9. Current Validation Status
| Check | Status | Interpretation |
|---|---|---|
| Watershed delineation | Completed | 162.035 km² watershed represented by five subbasins. |
| Land-use / Curve Number analysis | Completed | Spatially weighted CN values assigned to S_1–S_5. |
| Runoff transform | Completed | SCS Unit Hydrograph using final adopted lag times. |
| Reach routing | Completed | Muskingum routing through R_1 and R_2. |
| 50-year HEC-HMS event | Completed | Q50 = 114.4 m³/s at Hare Weir. |
| 1995 Q50 comparison | Completed | Historical estimate = 59.183 m³/s. |
| Observed hydrograph calibration | Not available | The present model should therefore be described as an engineering rainfall-runoff reassessment rather than a formally calibrated hydrologic model. |
| Independent SWAT comparison | Next stage | SWAT will provide an additional independent hydrologic assessment of the Hare watershed. |
10. Engineering Conclusion
The historical 1995 flood analysis and the modern HEC-HMS model provide two independent views of flood behaviour in the Hare watershed.
The 1995 study estimated a 50-year design peak of 59.183 m³/s, while the present HEC-HMS model gives 114.4 m³/s. The difference is substantial and is therefore explicitly retained rather than concealed or artificially removed through parameter adjustment.
For the reconstructed Hare Irrigation Project, both values remain part of the engineering evidence. The modern HEC-HMS hydrograph provides the rainfall-runoff design estimate, while the 1995 frequency analysis provides an important historical benchmark against which the modern result can be judged.
The next independent hydrologic check is the SWAT watershed model. The combined evidence from the historical flow record, HEC-HMS and SWAT will strengthen the hydrologic basis for the final assessment of the Hare diversion weir and headworks.