1r.sim.water(1) Grass User's Manual r.sim.water(1)
2
6 r.sim.water - Overland flow hydrologic model based on duality parti‐
7 cle-field concept (SIMWE)
8
10 raster
11
13 r.sim.water
14 r.sim.water help
15 r.sim.water [-mt] elevin=string dxin=string dyin=string rain=string
16 infil=string [traps=string] manin=string [sites=string]
17 [depth=string] [disch=string] [err=string] [outwalk=string]
18 [nwalk=integer] [niter=integer] [outiter=integer] [density=inte‐
19 ger] [diffc=float] [hmax=float] [halpha=float] [hbeta=float]
20 [--overwrite]
21 Flags:
23 -m Multiscale simulation
24 -t Time-series (dynamic) output
26 --overwrite
28 Parameters:
30 elevin=string
31 Name of the elevation raster file
32 dxin=string
34 Name of the x-derivatives raster file
35 dyin=string
37 Name of the y-derivatives raster file
38 rain=string
40 Name of the rainfall excess raster file
41 infil=string
43 Name of the infiltration excess raster file
44 traps=string
46 Name of the flow control raster file
47 manin=string
49 Name of the Mannings n raster file
50 sites=string
52 Name of the site file with x,y locations
53 depth=string
55 Output water depth raster file
56 disch=string
58 Output water discharge raster file
59 err=string
61 Output simulation error raster file
62 outwalk=string
64 Name of the output walkers site file
65 nwalk=integer
67 Number of walkers Default: 2000000
68 niter=integer
70 Number of time iterations (sec.) Default: 1200
71 outiter=integer
73 Time step for saving output maps (sec.) Default: 300
74 density=integer
76 Density of output walkers Default: 200
77 diffc=float
79 Water diffusion constant Default: 0.8
80 hmax=float
82 Threshold water depth (diffusion increases after this water depth
83 is reached) Default: 0.4
84 halpha=float
86 Diffusion increase constant Default: 4.0
87 hbeta=float
89 Weighting factor for water flow velocity vector Default: 0.5
90
92 r.sim.water is a landscape scale, simulation model of overland flow
93 designed for spatially variable terrain, soil, cover and rainfall
94 excess conditions. A 2D shallow water flow is described by the bivari‐
95 ate form of Saint Venant equations. The numerical solution is based on
96 the concept of duality between the field and particle representation of
97 the modeled quantity. Green's function Monte Carlo method, used to
98 solve the equation, provides robustness necessary for spatially vari‐
99 able conditions and high resolutions (Mitas and Mitasova 1998). The
100 key inputs of the model include elevation (elevin raster file), flow
101 gradient vector given by first-order partial derivatives of elevation
102 field (dxin and dyin raster files), rainfall excess rate (rain raster
103 file) and a surface roughness coefficient given by Manning's n (manin
104 raster file). Partial derivatives raster files can be computed along
105 with the interpolation of a DEM using the -d option in v.surf.rst mod‐
106 ule. If elevation raster is already provided, partial derivatives can
107 be computed using r.slope.aspect module. Partial derivatives determine
108 the direction and magnitude of water flow. Partial derivatives com‐
109 puted from terrain can be modified to include pre-defined water flow
110 (e.g. channels).
111 The module automatically converts data from feet to metric system using
113 database/projection information. Rainfall excess is defined as rainfall
114 intensity - infiltration rate. Rainfall intensities are usually avail‐
115 able from meteorological stations. Infiltration rate depends on soil
116 properties and land cover. It varies in space and time. For satu‐
117 rated soil and steady-state water flow it can be estimated using
118 saturated hydraulic conductivity rates based on field measurements or
119 using reference values which can be found in literature.
120 Output includes water depth raster file depth in [m], water dis‐
122 charge raster file disch in [m3/s]. Error of the numerical solution can
123 be analyzed using err raster file (the resulting water depth is an
124 average, and err is its RMSE). Output site file outwalk can be used to
125 analyze and visualize spatial distribution of walkers at different sim‐
126 ulation times (note that the resulting water depth is based on the den‐
127 sity of these walkers). Number of theese output walkers is controled by
128 density parameter, which says how many walkers used in simalution
129 should be used in the output Duration of simulation is controled by
130 niter parameter. The default value is 1000 iterations, to reach the
131 steady-state may require, depending on the time step, complexity of
132 terrain and land cover and size of the area, several thousand itera‐
133 tions. Output files can be saved during simulation using outiter
134 parameter defining the time step for writing output files. This option
135 requires the time series flag -t. Files are saved with suffix contain‐
136 ing iteration number (e.g. name.500, name.1000, etc.).
137 Overland flow is routed based on partial derivatives of elevation
138 field or other landscape features influencing water flow. Simulation
139 equations include a diffusion term (diffc parameter) which enables
140 water flow to overcome elevation depressions or obstacles when water
141 depth exceeds a threshold water depth value (hmax). When it is reached,
142 diffusion term increases as given by halpha and advection term (direc‐
143 tion of flow) is given as "prevailing" direction of flow computed as
144 average of flow directions from the previous hbetanumber of grid cells.
145
147 A 2D shallow water flow is described by the bivariate form of Saint
148 Venant equations (e.g., Julien et al., 1995). The continuity of water
149 flow relation is coupled with the momentum conservation equation and
150 for a shallow water overland flow, the hydraulic radius is approximated
151 by the normal flow depth. The system of equations is closed using the
152 Manning's relation. Model assumes that the flow is close to the kine‐
153 matic wave approximation, but we include a diffusion-like term to
154 incorporate the impact of diffusive wave effects. Such an incorporation
155 of diffusion in the water flow simulation is not new and a similar
156 term has been obtained in derivations of diffusion-advection equations
157 for overland flow, e.g., by Lettenmeier and Wood, (1992). In our
158 reformulation, we simplify the diffusion coefficient to a constant and
159 we use a modified diffusion term. The diffusion constant which we have
160 used is rather small (approximately one order of magnitude smaller than
161 the reciprocal Manning's coefficient) and therefore the resulting flow
162 is close to the kinematic regime. However, the diffusion term improves
163 the kinematic solution, by overcoming small shallow pits common in
164 digital elevation models (DEM) and by smoothing out the flow over slope
165 discontinuities or abrupt changes in Manning's coefficient (e.g., due
166 to a road, or other anthropogenic changes in elevations or cover).
167 Green's function stochastic method of solution. The Saint Venant equa‐
169 tions are solved by a stochastic method called Monte Carlo (very simi‐
170 lar to Monte Carlo methods in computational fluid dynamics or to quan‐
171 tum Monte Carlo approaches for solving the Schrodinger equation
172 (Schmidt and Ceperley, 1992, Hammond et al., 1994; Mitas, 1996)). It
173 is assumed that these equations are a representation of stochastic
174 processes with diffusion and drift components (Fokker-Planck equa‐
175 tions). The Monte Carlo technique has several unique advantages which
176 are becoming even more important due to new developments in computer
177 technology. Perhaps one of the most significant Monte Carlo proper‐
178 ties is robustness which enables us to solve the equations for com‐
179 plex cases, such as discontinuities in the coefficients of differen‐
180 tial operators (in our case, abrupt slope or cover changes, etc).
181 Also, rough solutions can be estimated rather quickly, which allows
182 us to carry out preliminary quantitative studies or to rapidly
183 extract qualitative trends by parameter scans. In addition, the sto‐
184 chastic methods are tailored to the new generation of computers as
185 they provide scalability from a single workstation to large parallel
186 machines due to the independence of sampling points. Therefore, the
187 methods are useful both for everyday exploratory work using a desktop
188 computer and for large, cutting-edge applications using high perfor‐
189 mance computing.
190
192 v.surf.rst r.slope.aspect r.sim.sediment
193
195 Helena Mitasova, Lubos Mitas
196 North Carolina State University
197 hmitaso@unity.ncsu.edu
198 Jaroslav Hofierka
199 GeoModel, s.r.o. Bratislava, Slovakia
200 hofierka@geomodel.sk
201 Chris Thaxton
202 North Carolina State University
203 csthaxto@unity.ncsu.edu
204 csthaxto@unity.ncsu.edu
205
207 Mitasova, H., Thaxton, C., Hofierka, J., McLaughlin, R., Moore, A.,
208 Mitas L., 2004, Path sampling method for modeling overland water flow,
209 sediment transport and short term terrain evolution in Open Source GIS.
210 In: C.T. Miller, M.W. Farthing, V.G. Gray, G.F. Pinder eds., Computa‐
211 tional Methods in Water Resources, Elsevier.
212 Mitasova H, Mitas, L., 2000, Modeling spatial processes in multiscale
214 framework: exploring duality between particles and fields, plenary talk
215 at GIScience2000 conference, Savannah, GA.
216 Mitas, L., and Mitasova, H., 1998, Distributed soil erosion simulation
218 for effective erosion prevention. Water Resources Research, 34(3),
219 505-516.
220 Mitasova, H., Mitas, L., 2001, Multiscale soil erosion simulations for
222 land use management, In: Landscape erosion and landscape evolution mod‐
223 eling, Harmon R. and Doe W. eds., Kluwer Academic/Plenum Publishers,
224 pp. 321-347.
225 Neteler, M. and Mitasova, H., 2004, Open Source GIS: A GRASS GIS
227 Approach, Second Edition, Kluwer International Series in Engineering
228 and Computer Science, 773, Kluwer Academic Press / Springer, Boston,
229 Dordrecht, 424 pages.
230 Last changed: Date: 2003/11/01 15:55:10 $
232GRASS 6.2.2 r.sim.water(1)