603 - 5885 1250 (4 lines) info@mes100.com
View Categories

Storm Sync Time Area Horton Infiltration Benchmark

Storm Sync Time Area can now take pervious losses from Horton infiltration, MSMA1 Equation 14.9. Before relying on it, an engineer will want to know that the loss MiTS removes from each rainfall block is the integral of that equation and nothing else. In this article we pick the inputs from MSMA1 Table 14.4, work a two-block storm by hand, reproduce the same numbers in the MiTS loss parameter, check MiTS calculation for edge cases (light rainfall blcok and a dry gap between rainfall blocks, a zero decay coefficient and a split rainfall block), and finally compare a real Penang MSMA1 project, block by block, against values derived from the manual.

This is an analytical check of the equation as implemented in MSMA1. It is not a comparison with EPA SWMM or another program, and not a field calibration.

For setup, see the Storm Sync Time Area hydrology guide. Additional benchmarks are documented for the decaying-loss model in Storm Sync Time Area Comparison with MSMA2 Appendix 2.F2 and proportional-loss model in Storm Sync Time Area Comparison with Legacy PondCAD.

The equation and its units #

MSMA1 Equation 14.9 gives the infiltration capacity of a pervious surface as it wets up:

f(t) = f_c + (f_0 – f_c)\, e^{-kt}

where,

  • f(t) = infiltration capacity (mm/hr) at a given time t
  • f_0 = initial infiltration capacity (mm/hr)
  • f_c = final or steady-state infiltration capacity (mm/hr)
  • k = decay coefficient (1/hr),
  • t = elapsed time (in hours) from the start of the first wet rainfall block.

A rainfall block runs from t_1 to t_2. The cumulative potential infiltration over the block is the area under the Horton infiltration capacity curve, f(t) between these two time interval:

F = \int_{t_1}^{t_2} f(t)\,dt = f_c (t_2 – t_1) + \dfrac{f_0 – f_c}{k}\left(e^{-k t_1} – e^{-k t_2}\right)

The applied rainfall loss is the smaller of F and the rainfall depth of the block. The rainfall excess is the rainfall depth minus the rainfall loss.

Inputs from MSMA1 Table 14.4 #

We take a dry clay soil with little vegetation, in hydrologic soil group C. In the MiTS soil parameter these presets give:

Soil Parameter Choice Value it sets Source
Soil texture Clay f_0 = 25 mm/h (dry) Table 14.4, dry clay soils with little or no vegetation
Vegetation Little or no vegetation No change to f_0
(For reference: for dense vegetation, the dry f_0 will be doubled)
Table 14.4
Moisture state Dry f_0 stays at 25 mm/h Table 14.4
Hydrologic soil group C f_c = 1.3 mm/h, the lower bound of group C’s 1.3–3.8 mm/h Table 14.4
Decay k (1/hr) default k = 4 per hour Table 14.4

For reference, the following moisture condition adjustments are based on the Table 14.4. The initial infiltration capacity f_0 for each condition is adjusted as follows:

  1. Drained but not dried divides the dry f_0 by 3 (8.33 mm/hr for clay).
  2. Partially dried out divides the dry f_0 by the partial-drying divisor which ranges from 1.5-2.5. Default is 2 (12.5 mm/hr for clay).
  3. Near saturation uses the saturated hydraulic conductivity Ks that you enter.

The preset values are automatically filled in based on the selected moisture condition. Every value remains editable.

MSMA 1st Edition, Chapter 14, Table 14.4

Worked storm by hand #

The storm is two consecutive 30-minute blocks of 20 mm each (40 mm/h). With the inputs above, (f_0 – f_c)/k = (25 − 1.3) / 4 = 5.925 mm. We need e^{-2} = 0.135335 and e^{-4} = 0.018316.

Block 1, 0 to 0.5h

F_1 = 1.3 \times 0.5 + 5.925 \times (1 – 0.135335) = 0.65 + 5.123138 = 5.773138 \text{ mm}

Block 2, 0.5 to 1.0h

F_2 = 1.3 \times 0.5 + 5.925 \times (0.135335 – 0.018316) = 0.65 + 0.693341 = 1.343341 \text{ mm}

Both are below the 20 mm of rainfall in their block, so the whole potential infiltration is lost:

Block Rainfall (mm) Loss (mm) Excess (mm)
0–30 min 20 5.773 14.227
30–60 min 20 1.343 18.657

MiTS’s returns 5.773 mm and 1.343 mm of loss and 14.227 mm and 18.657 mm of excess, the same to every digit shown. The infiltration capacity falls quickly: after one hour the decaying part has shrunk to e^{-4}, under 2% of its starting value (e^{0}), so the second block loses barely more than f_c alone would (1.3 mmm/h × 0.5h = 0.65 mm).

Reproduce it in MiTS #

The preview table shows the calculated integral in 5-minute intervals, so users can check these numbers on screen without setting up a separate storm event.

  1. Open any catchment’s Time Area Settings from the Catchment Spread Input.
  2. Set Rainfall loss model to Horton infiltration.
  3. Set Soil texture = Clay, Vegetation = Little or no vegetation, Moisture state = Dry and Hydrologic soil group = C. Check that Initial f0 reads 25, Final fc 1.3 and Decay k 4.
  4. In the preview, choose Surface = Pervious, Preview starts (min) = 0 and Preview ends (min) = 60.

The preview table should read:

From (min) To (min) Loss depth (mm)
0 5 1.788
5 10 1.312
10 15 0.971
15 20 0.726
20 25 0.551
25 30 0.426
30 35 0.336
35 40 0.271
40 45 0.225
45 50 0.192
50 55 0.168
55 60 0.151

The first six rows add up to 5.773 mm and the last six to 1.343 mm, the two block losses above. That is what the integral promises: integrating over six 5-minute pieces gives the same depth as integrating over the 30-minute block in one go. Each row is F with t_1 and t_2 at the row’s From and To; for the first row, 1.3 \times \tfrac{5}{60} + 5.925 \times (1 – e^{-1/3}) = 0.1083 + 1.6796 = 1.7879 mm.

These are potential depths. In an analysis, the loss in each period is capped at that period’s rainfall.

Potential Horton infiltration by rainfall period

Check MiTS calculation for Edge Cases #

All three use f_0 = 25 mm/h, f_c = 1.3 mm/h and k = 4 per hour unless stated.

Infiltration Loss During Wet and Dry Periods #

The storm consists of four 30-minute blocks, with rainfall depths of 0 mm, 1 mm, 0 mm, and 20 mm repectively. Elapsed time starts at the first wet block (30 min), so that block runs from t = 0 to 0.5 h, the dry gap from 0.5 to 1.0 h, and the last block from 1.0 to 1.5 h.

Block Condition Rainfall (mm) Elapsed t (h) Potential F (mm) Loss (mm) Excess (mm)
0–30 min No rainfall 0 Before rain – 0 0
30–60 min Light rainfall 1 0 – 0.5 5.773 1 0
60–90 min Dry period 0 0.5 – 1.0 1.343 0 0
90–120 min Heavy rainfall 20 1.0 – 1.5 0.744 0.744 19.256

For the last block, F = 0.65 + 5.925 \times (e^{-4} – e^{-6}) = 0.65 + 5.925 \times (0.0183 – 0.0025) = 0.7438 mm.

This example demonstrates two aspects of the Horton infiltration calculation:

Rainfall loss cannot exceed infiltration capacity: During the first wet interval, the potential infiltration capacity is 5.773 mm, but only 1 mm of rainfall is available. Therefore, the infiltration loss is limited to 1 mm, resulting in zero rainfall excess; it won’t go down to negative.

Continuous infiltration decay: During the dry interval, no rainfall occurs, but the infiltration capacity continues to decay with elapsed time. It does not pause or recover between rainfall intervals. As a result, the potential infiltration loss during the final wet interval is 0.744 mm rather than 1.343 mm. If the capacity had paused during the gap, the last block would have lost 1.343 mm instead of 0.744 mm. MiTS does not let the soil recover during dry periods within a storm event.

Zero decay coefficient #

With k = 0 the capacity stays at f_0, and the integral becomes f_0 (t_2 – t_1). For one 30-minute block of 20 mm, the loss is 25 × 0.5 = 12.5 mm. The formula above divides by k, so MiTS evaluates this limit directly; a k as small as 10^{-12} gives the same 12.5 mm.

Splitting a block #

For a 60-minute rainfall event starting at 10 minutes, a single block of 100 mm produces a total infiltration loss of F = 1.3 × 1 + 5.925 × (1 − 0.018316) = 7.116 mm. When the same rainfall is divided into two blocks: 15-minute block of 25 mm and a 45-minute block of 75 mm rainfall depth, the respective infiltration losses are 4.070 mm and 3.046 mm. These add up to the same total infiltration loss of 7.116 mm.

Item First Block Second Block
Rainfall depth 25 mm 75 mm
Start time 0 min 15 min
End time 15 min 60 min
Duration 0.25 h 0.75 h
Calculation 1.3 × 0.25 + 5.925 × (1 – 0.3679) 1.3 × 0.75 + 5.925 (0.3679 – 0.0183)
Infiltration loss 4.070 mm 3.046 mm

The total Horton infiltration loss is independent of how the rainfall is divided into time blocks. Using a finer temporal pattern does not change the total infiltration loss, provided that the total rainfall duration and the continuous infiltration decay period remain unchanged.

Invalid inputs #

MiTS checks the input settings before calculation and rejects invalid values or configurations. These include when f_c is greater than f_0 (for example f_c = 30 > f_0 = 25), when k is not a number, when a block ends before it starts, or when a block has negative rainfall depth.

A real project #

The previous examples use hypothetical rainfall data to illustrate individual Horton calculations. This section applies the Horton infiltration method in Storm Sync analysis using a sample project. The results are then compared with manual calculations based on MSMA1.

The project and settings #

The project sample is MSMA1_TimeArea.7z: the legacy PondCAD MSMA1 detention pond in Penang, re-entered as two Storm Design Time Area catchments. We use catchment Pre: 20 ha, fully pervious, time of concentration 30 minutes, divided into six 5-minute bands of 3.33 ha.

  1. Open the project. Check that Design Rainfall > Malaysia is MSMA 1st Edition (2000), the IDF source is the Penang urban centre, and the temporal pattern is MSMA1 West Coast.
  2. In Storm Design > Storm Analysis, set the analysis type to Analysis, the catchment hydrology method to Time Area, and tick only the 30 min storm duration.
  3. In the Catchment Spread Input, open catchment Pre’s Time Area Settings and set Horton infiltration with the presets in Inputs from MSMA1 Table 14.4: f0 = 25, fc = 1.3, k = 4. Apply.
  4. Execute.
  5. Open the Catchment Hydrology report, set it to Detailed, and select catchment Pre and the 2-year ARI, then the 50-year ARI.

a) Catchment Pre’s Time Area Settings with Horton and the six bands

b) Catchment Hydrology report for Catchment “Pre”, ARI 2 years, showing the per-interval rainfall, loss and excess

c) Catchment Hydrology report for Catchment “Pre”, ARI 50 years, showing the per-interval rainfall, loss and excess

Expected values from the guideline #

Rainfall depth, MSMA1 Equation 13.2. For Penang, Table 13.A1 gives:

ARI a b c d
2 years 4.514 0.6729 −0.2311 0.0118
50 years 2.8429 2.1456 −0.5469 0.0335

MSMA1 Table 13.A1 – Fitting coefficients For Penang

At t = 30 min, \ln 30 = 3.401197. For 2 years, \ln I = 4.514 + 0.6729 × 3.401197 − 0.2311 × 3.401197² + 0.0118 × 3.401197³ = 4.593545, so I = 98.8442 mm/h and the 30-minute depth is 98.8442 × 0.5 = 49.4221 mm. The same working for 50 years gives I = 169.3499 mm/h and 84.6749 mm.

Rainfall per block, MSMA1 Table 13.B1. The West Coast 30-minute pattern splits the depth into six 5-minute blocks with fractions 0.16, 0.25, 0.33, 0.09, 0.11 and 0.06.

Loss per block, MSMA1 Equation 14.9. Each block’s potential infiltration is the integral over that block, the same twelve preview values as in Reproduce it in MiTS for the first 30 minutes. Every block here has more rainfall than its potential loss, so the full potential is lost and the excess is rainfall minus loss.

Comparison, 2-year ARI #

Block (min) Rainfall, guideline (mm) Rainfall, MiTS (mm) Loss, guideline (mm) Loss, MiTS (mm) Excess, guideline (mm)
0–5 7.9075 7.9075 1.7879 1.7879 6.1197
5–10 12.3555 12.3555 1.3118 1.3118 11.0437
10–15 16.3093 16.3093 0.9706 0.9706 15.3387
15–20 4.4480 4.4480 0.7262 0.7262 3.7218
20–25 5.4364 5.4364 0.5511 0.5511 4.8854
25–30 2.9653 2.9653 0.4256 0.4256 2.5398
Total 49.4221 49.4221 5.7731 5.7731 43.6490

MiTS reports a pervious excess of 43.6490 mm. The runoff volume is 10 × 43.6490 mm × 200000 m² = 8,729.79 m³ by hand, and MiTS reports 8,729.79 m³. MiTS gives a peak flow of 4.850 m³/s at 25 min; the peak comes from the Time Area routing, which this article does not rework by hand.

Comparison, 50-year ARI #

Block (min) Rainfall, guideline (mm) Rainfall, MiTS (mm) Loss, guideline (mm) Loss, MiTS (mm) Excess, guideline (mm)
0–5 13.5480 13.5480 1.7879 1.7879 11.7601
5–10 21.1687 21.1687 1.3118 1.3118 19.8570
10–15 27.9427 27.9427 0.9706 0.9706 26.9721
15–20 7.6207 7.6207 0.7262 0.7262 6.8945
20–25 9.3142 9.3142 0.5511 0.5511 8.7632
25–30 5.0805 5.0805 0.4256 0.4256 4.6549
Total 84.6749 84.6749 5.7731 5.7731 78.9018

MiTS reports a pervious excess of 78.9018 mm and a runoff volume of 15,780.361 m³, against 10 × 78.9018 × 20 = 15,780.36 m³ by hand. The peak is 8.767 m³/s at 25 min.

The loss is the same 5.7731 mm in both storms. That is expected: Horton loss depends only on time since the rain started, as long as each block has more rain than the soil can take, so the heavier storm’s extra rainfall all becomes excess.

Automated full-analysis check #

A second automated check runs Horton through a complete Storm Sync analysis under MSMA 1st, 2nd and 3rd Edition, using our MSMA1 and MSMA2 Time Area test projects. Every catchment is set to Horton with f_0 = 61 mm/h, f_c = 6 mm/h and k = 3.9 per hour, half of each band impervious, and a 30-minute storm. For every catchment and ARI, the saved result must satisfy:

  • pervious loss + pervious excess = rainfall depth, with no pervious initial loss;
  • impervious initial loss + impervious continuing loss + impervious excess = rainfall depth;
  • runoff volume (m³) = 10 × (pervious excess (mm) × pervious area (ha) + impervious excess (mm) × impervious area (ha)).

After the run, we change f_0 in the project to 99 mm/h. The previously saved results and its detailed Hydrology report continue to show f_0 = 61 mm/h. The report also cites MSMA1 Equation 14.9, confirming that the reported parameters reflect the original analysis configuration rather than the current project configuration. To reflect the updated parameters (f_0 = 99 mm/h) in the results and report, the analysis must be run again.

Assumptions in this model #

  • Rainfall cap. The loss in a block cannot exceed that block’s rainfall.
  • No pervious initial loss. Horton replaces the pervious initial and continuing loss; MiTS adds no separate initial abstraction on top of it.
  • No recovery in dry gaps. Elapsed time starts at the first wet block and continues until the end of the storm.
  • Impervious surface. The impervious part of the catchment uses its own initial loss and constant rate (defaults 1.5 mm and 0 mm/h), not Horton.
  • Event model. This is a single-storm model with fixed parameters. Run-on from one band infiltrating in another is not modelled. In other words, the model treats each band independently when calculating infiltration, without considering runoff received from other bands

Horton is offered under MSMA 1st, 2nd and 3rd Edition, with presets from MSMA1 Table 14.4. However, providing Horton as an option under MSMA2 or MSMA3 does not mean these editions prescribe the MSMA1 presets; the selected parameters should be justified based on the site conditions.

Conclusion #

For each rainfall block, MiTS removes the exact integral of MSMA1 Equation 14.9 over the block, with the loss capped at the block’s rainfall depth. Time is counted from the first wet block until the end of storm duration. No infiltration recovery will be modelled during dry gaps between rainfall blocks.

The worked storm gives 5.773 mm and 1.343 mm of loss by hand, in the calculator and as the sums of the preview table. On the Penang MSMA1 project, the total rainfall, rainfall loss, rainfall excess and runoff volume from MSMA1 Equations 13.2 and 14.9 and Table 13.B1 match MiTS for both the 2- and 50-year storms.

Additional tests confirm that MiTS calculates infiltration loss correctly under all edge cases. A complete Storm Sync analysis also confirms that rainfall is accounted for on both pervious and impervious surfaces, and that the Hydrology report retains the parameters used during the analysis.

This benchmark does not establish whether the Table 14.4 presets are suitable for a particular catchment. Their suitability remains subject to the engineer’s judgement.

Powered by BetterDocs