1r.spread(1) Grass User's Manual r.spread(1)
2
3
4
6 r.spread - Simulates elliptically anisotropic spread.
7 Generates a raster map of the cumulative time of spread, given raster
8 maps containing the rates of spread (ROS), the ROS directions and the
9 spread origins. It optionally produces raster maps to contain backlink
10 UTM coordinates for tracing spread paths. Usable for fire spread simu‐
11 lations.
12
14 raster, fire, spread, hazard, model
15
17 r.spread
18 r.spread --help
19 r.spread [-si] base_ros=string max_ros=string direction_ros=string
20 start=string [spotting_distance=string] [wind_speed=string]
21 [fuel_moisture=string] [least_size=odd int] [comp_dens=decimal]
22 [init_time=int (>= 0)] [lag=int (>= 0)] [backdrop=string] out‐
23 put=string [x_output=string] [y_output=string] [--overwrite]
24 [--help] [--verbose] [--quiet] [--ui]
25
26 Flags:
27 -s
28 Consider spotting effect (for wildfires)
29
30 -i
31 Use start raster map values in output spread time raster map
32 Designed to be used with output of previous run of r.spread when
33 computing spread iteratively. The values in start raster map are
34 considered as time. Allowed values in raster map are from zero to
35 the value of init_time option. If not enabled, init_time is used in
36 the area of start raster map
37
38 --overwrite
39 Allow output files to overwrite existing files
40
41 --help
42 Print usage summary
43
44 --verbose
45 Verbose module output
46
47 --quiet
48 Quiet module output
49
50 --ui
51 Force launching GUI dialog
52
53 Parameters:
54 base_ros=string [required]
55 Raster map containing base ROS (cm/min)
56 Name of an existing raster map layer in the user’s current mapset
57 search path containing the ROS values in the directions perpendicu‐
58 lar to maximum ROSes’ (cm/minute). These ROSes are also the ones
59 without the effect of directional factors.
60
61 max_ros=string [required]
62 Raster map containing maximal ROS (cm/min)
63 Name of an existing raster map layer in the user’s current mapset
64 search path containing the maximum ROS values (cm/minute).
65
66 direction_ros=string [required]
67 Raster map containing directions of maximal ROS (degree)
68 Name of an existing raster map layer in the user’s current mapset
69 search path containing directions of the maximum ROSes, clockwise
70 from north (degree).
71
72 start=string [required]
73 Raster map containing starting sources
74 Name of an existing raster map layer in the user’s current mapset
75 search path containing starting locations of the spread phenomenon.
76 Any positive integers in this map are recognized as starting
77 sources (seeds).
78
79 spotting_distance=string
80 Raster map containing maximal spotting distance (m, required with
81 -s)
82 Name of an existing raster map layer in the user’s current mapset
83 search path containing the maximum potential spotting distances
84 (meters).
85
86 wind_speed=string
87 Raster map containing midflame wind speed (ft/min, required with
88 -s)
89 Name of an existing raster map layer in the user’s current mapset
90 search path containing wind velocities at half of the average flame
91 height (feet/minute).
92
93 fuel_moisture=string
94 Raster map containing fine fuel moisture of the cell receiving a
95 spotting firebrand (%, required with -s)
96 Name of an existing raster map layer in the user’s current mapset
97 search path containing the 1-hour (<.25") fuel moisture (percentage
98 content multiplied by 100).
99
100 least_size=odd int
101 Basic sampling window size needed to meet certain accuracy (3)
102 An odd integer ranging 3 - 15 indicating the basic sampling window
103 size within which all cells will be considered to see whether they
104 will be reached by the current spread cell. The default number is 3
105 which means a 3x3 window.
106 Options: 3, 5, 7, 9, 11, 13, 15
107
108 comp_dens=decimal
109 Sampling density for additional computing (range: 0.0 - 1.0 (0.5))
110 A decimal number ranging 0.0 - 1.0 indicating additional sampling
111 cells will be considered to see whether they will be reached by the
112 current spread cell. The closer to 1.0 the decimal number is, the
113 longer the program will run and the higher the simulation accuracy
114 will be. The default number is 0.5.
115
116 init_time=int (>= 0)
117 Initial time for current simulation (0) (min)
118 A non-negative number specifying the initial time for the current
119 spread simulation (minutes). This is useful when multiple phase
120 simulation is conducted. The default time is 0.
121 Default: 0
122
123 lag=int (>= 0)
124 Simulating time duration LAG (fill the region) (min)
125 A non-negative integer specifying the simulating duration time lag
126 (minutes). The default is infinite, but the program will terminate
127 when the current geographic region/mask has been filled. It also
128 controls the computational time, the shorter the time lag, the
129 faster the program will run.
130
131 backdrop=string
132 Name of raster map as a display backdrop
133 Name of an existing raster map layer in the user’s current mapset
134 search path to be used as the background on which the "live" move‐
135 ment will be shown.
136
137 output=string [required]
138 Raster map to contain output spread time (min)
139 Name of the new raster map layer to contain the results of the
140 cumulative spread time needed for a phenomenon to reach each cell
141 from the starting sources (minutes).
142
143 x_output=string
144 Name of raster map to contain X back coordinates
145 Name of the new raster map layer to contain the results of backlink
146 information in UTM easting coordinates for each cell.
147
148 y_output=string
149 Name of raster map to contain Y back coordinates
150 Name of the new raster map layer to contain the results of backlink
151 information in UTM northing coordinates for each cell.
152
154 r.spread is part of the wildfire simulation toolset. Preparational
155 steps for the fire simulation are the calculation of the rate of spread
156 (ROS) with r.ros, and the creating of spread map with r.spread. Even‐
157 tually, the fire path(s) based on starting point(s) are calculated with
158 r.spreadpath.
159
160 Spread phenomena usually show uneven movement over space. Such uneven‐
161 ness is due to two reasons:
162
163 1 the uneven conditions from location to location, which can be
164 called spatial heterogeneity, and
165
166 2 the uneven conditions in different directions, which can be
167 called anisotropy.
168
169 The anisotropy of spread occurs when any of the determining factors
170 have directional components. For example, wind and topography cause an‐
171 isotropic spread of wildfires.
172
173 One of the simplest spatial heterogeneous and anisotropic spread is
174 elliptical spread, in which, each local spread shape can be thought as
175 an ellipse. In a raster setting, cell centers are foci of the spread
176 ellipses, and the spread phenomenon moves fastest toward apogees and
177 slowest to perigees. The sizes and shapes of spread ellipses may vary
178 cell by cell. So the overall spread shape is commonly not an ellipse.
179
180 r.spreadsimulates elliptically anisotropic spread phenomena, given
181 three raster map layers about ROS (base ROS, maximum ROS and direction
182 of the maximum ROS) plus a raster map layer showing the starting
183 sources. These ROS layers define unique ellipses for all cell loca‐
184 tions in the current computational region as if each cell center was a
185 potential spread origin. For some wildfire spread, these ROS layers
186 can be generated by another GRASS raster program r.ros. The actual
187 locations reached by a spread event are constrained by the actual
188 spread origins and the elapsed spread time.
189
190 r.spreadoptionally produces raster maps to contain backlink UTM coordi‐
191 nates for each raster cell of the spread time map. The spread paths can
192 be accurately traced based on the backlink information by r.spreadpath
193 module.
194
195 Part of the spotting function in r.spread is based on Chase (1984) and
196 Rothermel (1983). More information on r.spread, r.ros and r.spreadpath
197 can be found in Xu (1994).
198
199 Options spot_dist, w_speed and f_mois must all be given if the -s
200 (spotting) flag is used.
201
203 Assume we have inputs, the following simulates a spotting- involved
204 wildfire and generates three raster maps to contain spread time, back‐
205 link information in UTM northing and easting coordinates:
206 r.spread -s max=my_ros.max dir=my_ros.maxdir base=my_ros.base \
207 start=fire_origin spot_dist=my_ros.spotdist w_speed=wind_speed \
208 f_mois=1hour_moisture output=my_spread \
209 x_output=my_spread.x y_output=my_spread.y
210
212 1. r.spread is a specific implementation of the shortest path algo‐
213 rithm. r.cost module served as the starting point for the development
214 of r.spread. One of the major differences between the two programs is
215 that r.cost only simulates isotropic spread while r.spread can simulate
216 elliptically anisotropic spread, including isotropic spread as a spe‐
217 cial case.
218
219 2. Before running r.spread, the user should prepare the ROS (base, max
220 and direction) maps using appropriate models. For some wildfire spread,
221 the r.ros module based on Rothermel’s fire equation does such work.
222 The combination of the two forms a simulation of wildfire spread.
223
224 3. The relationship of the start map and ROS maps should be logically
225 correct, i.e. a starting source (a positive value in the start map)
226 should not be located in a spread barrier (zero value in the ROS maps).
227 Otherwise the program refuses to run.
228
229 4. r.spread uses the current computational region settings. The output
230 map layer will not go outside the boundaries set in the region, and
231 will not be influenced by starting sources outside. So any change of
232 the current region may influence the output. The recommendation is to
233 use slightly larger region than needed. Refer to g.region to set an
234 appropriate computational region.
235
236 5. The user should be sure that the inputs to r.spread are in proper
237 units.
238
239 6. r.spread is a computationally intensive program. The user may need
240 to choose appropriate size of the computational region and resolution.
241
242 7. A low and medium (i.e. <= 0.5) sampling density can improve accuracy
243 for elliptical simulation significantly, without adding significantly
244 extra running time. Further increasing the sample density will not gain
245 much accuracy while requiring greatly additional running time.
246
248 · Chase, Carolyn, H., 1984, Spotting distance from wind-driven
249 surface fires -- extensions of equations for pocket calcula‐
250 tors, US Forest Service, Res. Note INT-346, Ogden, Utah.
251
252 · Rothermel, R. C., 1983, How to predict the spread and intensity
253 of forest and range fires. US Forest Service, Gen. Tech. Rep.
254 INT-143. Ogden, Utah.
255
256 · Xu, Jianping, 1994, Simulating the spread of wildfires using a
257 geographic information system and remote sensing, Ph. D. Dis‐
258 sertation, Rutgers University, New Brunswick, New Jersey (ref).
259
261 r.cost, r.mask, r.ros, r.spreadpath Sample data download: firedemo.sh
262 (run this script within the "Fire simulation data set" location.
263
265 Jianping Xu and Richard G. Lathrop, Jr., Center for Remote Sensing and
266 Spatial Analysis, Rutgers University.
267
269 Available at: r.spread source code (history)
270
271 Main index | Raster index | Topics index | Keywords index | Graphical
272 index | Full index
273
274 © 2003-2019 GRASS Development Team, GRASS GIS 7.8.2 Reference Manual
275
276
277
278GRASS 7.8.2 r.spread(1)