Model help:TopoFlow-Infiltration-Smith-Parlange

TopoFlow-Infiltration-Smith-Parlange

This module is the infiltration process component (Smith-Parlange method) for a D8-based, spatial hydrologic model

Model introduction

This process component is part of a spatially-distributed hydrologic model called TopoFlow, but it can now be used as a stand-alone model.

Model parameters

Parameter	Description	Unit
First parameter	Description parameter	[Units]

Parameter	Description	Unit
First parameter	Description parameter	[Units]

Uses ports

This will be something that the CSDMS facility will add

Provides ports

This will be something that the CSDMS facility will add

Main equations

Infiltrability (max infiltration rate)

[math]\displaystyle{ f_{c}= K_{s} + Y \left ( K_{s} - K_{i}\right) / \left ( exp \left ( Y F / J \right) - 1 \right) }[/math]

(1)

[math]\displaystyle{ J= G \left ( \theta_{s} - \theta_{i}\right) }[/math]

(2)

Infiltration rate at surface (K_s < (P + M))

[math]\displaystyle{ v_{0}= min \left ( \left ( P + M \right), f_{c}\right) }[/math]

(3)

Infiltration rate at surface (K_s > (P + M))

[math]\displaystyle{ v_{0}= \left ( P + M \right) }[/math]

(4)

Cumulative infiltration depth (from 0 to t)

[math]\displaystyle{ F= \int v_{0}\left ( t \right) dt }[/math]

(5)

Nomenclature

Symbol	Description	Unit
K_s	saturated hydraulic conductivity	m / s
K_i	initial hydraulic conductivity (typically much less than K_s)	m / s
θ_s	soil water content at ψ = 0 (typically set to the porosity, ψ)	-
θ_i	initial soil water content	-
G	capillary length scale	m
P	precipitation rate	mm / sec
M	snowmelt rate	mm / sec
Y	Smith-Pariange method parameter (between 0 and 1, near 0.8)
f_c	infiltrability	-
F	cumulative infiltration depth	m
J		-
v0	infiltration rate at surface	mm / sec
t	time	s

Notes

Note on input parameters

For a detailed discussion of these variables and infiltration theory, see the References below.

For each variable, you may choose from the droplist of data types. For the "Scalar" data type, enter a numeric value with the units indicated in the dialog. For the other data types, enter a filename. Values in files must also use the indicated units.

Single grids and grid sequences are assumed to be stored as RTG and RTS files, respectively. Time series are assumed to be stored as text files, with one value per line. For a time series or grid sequence, the time between values must coincide with the timestep provided.

Choosing an entry from the droplist labeled "Closest standard soil type" will change the values in the dialog to tabulated values for the selected soil type. However, these values were determined from plot-scale measurements and are unlikely to be appropriate for large grid cells. For large grid cells, some type of upscaling is typically required.

Note on the equations

t_p = time of ponding [minutes] = the time when the soil becomes saturated at the surface, after which v₀=f_c or v₀=0 (after surface inputs stop). If (P + M) < K_s, then ponding cannot occur.

The equation for v0 implies that v₀ = 0 whenever (P + M) = 0, since f_c > 0.

If (P + M) > K_s, then after a sufficiently long time F will become large, the term with the exponential in the denominator will approach zero and f_c will decrease asymptotically to K_s.

The definition of F implies that dF/dt = v₀. Here, F is the quantity that Smith (2002) refers to as I', but that doesn't display well in HTML.

The current implementation is meant for single events only since F is only reset to 0 at the start of each model run.

In the case where (P + M) is uniform in time, it is possible to compute the surface soil moisture using a relationship of the form, K = K(θ). This type of relationship, found empirically, is called a soil characteristic function and is also used in conjunction with the 1D Richards' equation method of infiltration. See Smith (2002, pp. 81-85) for details.

The Green-Ampt and Smith-Parlange methods for modeling infiltration are based on the infiltrability-depth approximation or IDA, which uses the cumulative infiltrated depth as a "replacement" for time. For details, see Smith (2002, pp. 71-73). These methods are not well-suited to modeling redistribution between events or drying of surface layers by evaporation. They are best used for single events.

If there is standing water of depth, d, at the surface, then G can/should be replaced by (G + d) in the equation for J. This isn't done in the current version, since channels typically only occupy a small fraction of a grid cell and water depth on hillslopes is typically very small.

As x approaches zero, a Taylor series shows that exp(x) approaches (1 + x). Using this fact it can be shown that as γ approaches 0, this model approaches the Green-Ampt model . Experimental and numerical results suggest that γ values around 0.8 to 0.85 often give the best fit for normal soils (Smith, 2002, p. 81).

For the case γ = 1, the equation for fc can be simplified to:

   f_c = [K_s * exp(F / J) - K_i] / [exp(F / J) - 1]

and this case was analyzed by Smith and Parlange (1978) and Woolhiser et al. (1996).

More information on how soil is modeled in TopoFlow along with published soil property tables can be found on this soil properties page .

Examples

An example run with input parameters, BLD files, as well as a figure / movie of the output

Follow the next steps to include images / movies of simulations:

Upload file: http://csdms.colorado.edu/wiki/Special:Upload
Create link to the file on your page: [[Image:<file name>]].

Developer(s)

Scott Peckham

References

Smith, R.E. (2002) Infiltration Theory for Hydrologic Applications, Water Resources Monograph 15, AGU.

Smith, R.E. and J.-Y. Parlange (1978) A parameter-efficient hydrologic infiltration model, Water Resources Research, 14(3), 533-538.

Woolhiser, D.A., R.E. Smith and J-V. Giraldez (1996) Effects of spatial variability of saturated hydraulic conductivity on Hortonian overland flow, Water Resources Research, 32(3), 671-678.

Links