def DEM_creator(H, H_wt, seed, elev_range, max_level, gradient_values, DEMcreator_option):
"""
Generates a DEM map with specified parameters described below.
Args:
H : List containing H values --> [H1, H2, H3, ...]( List of floats )
H_wt : List containing H_wt --> [H1_wt, H2_wt, H3_wt, ...] ( List of floats )
seed : seed for random no generator --> [seed1, seed2 seed3, ...] ( list of ints)
elev_range: List containing elev bounds --> [elev_min, elev_max] (int, int)
max_level : size if grid 2^(max_level) + 1 ( int )
gradient_values: list containing gradient values (List of float values)
DEMcreator_option: Specify the method used to generate DEM (fm2D or SS)
Result:
Output : [ DEM_arr, [List of DEM input grids], [ List of DEM input grid names ], [list of parameters for DEM input grid]]
where DEM_arr :Final Output DEM grid
"""
Input_DEMs = []
name = []
parameter = []
if DEMcreator_option == 'fm2D':
#Generate first DEM with gradient = 1 (i.e. TRUE) using FM2D method
DEM_arr = MapGeneration_pure_python.midPointFm2d(max_level = max_level, sigma = 1, H = H[0], addition = True,\
wrap = False, gradient = 1, seed = seed[0], normalise=True, bounds = elev_range, gradient_values = gradient_values)
Input_DEMs.append(DEM_arr)
file_name = "InputDEM_arr%d" % (1)
name.append(file_name)
parameter.append((H[0], H_wt[0]))
DEM_arr = DEM_arr*H_wt[0]
for i in range(1,len(H)):
#Generate other DEM's with gradient = 0 (i.e. FLASE) and method specified by DEMcreator_option
temp_arr = MapGeneration_pure_python.midPointFm2d(max_level = max_level, sigma = 1, H = H[i], addition = True,\
wrap = False, gradient = 0, seed = seed[i], normalise = True, bounds = elev_range, gradient_values = gradient_values)
Input_DEMs.append(temp_arr)
file_name = "InputDEM_arr%d" % (i+1)
name.append(file_name)
parameter.append((H[i], H_wt[i]))
DEM_arr = DEM_arr + H_wt[i]*temp_arr
else:
#Generate first DEM with gradient = 1 (i.e. TRUE) using FM2D method
DEM_arr = MapGeneration_pure_python.midPointFm2d(max_level = max_level, sigma = 1, H = H[0], addition = True,\
wrap = False, gradient = 1, seed = seed[0], normalise=True, bounds = elev_range, gradient_values = gradient_values)
DEM_arr = DEM_arr[:-1,:-1] # omit last row and last col to maintain consistency with Spectral method output grid
Input_DEMs.append(DEM_arr)
file_name = "InputDEM_arr%d" % (1)
name.append(file_name)
parameter.append((H[0], H_wt[0]))
DEM_arr = H_wt[0]*DEM_arr
for i in range(1,len(H)):
#Generate other DEM's with gradient = 0 (i.e. FLASE) and method specified by DEMcreator_option i.e SS in this case
temp_arr = SpectralSynthesisFM2D.SpectralSynthesisFM2D(max_level = max_level, sigma = 1, H = H[i],\
seed = seed[i], normalise = True, bounds = elev_range)
Input_DEMs.append(temp_arr)
file_name = "InputDEM_arr%d" % (i+1)
name.append(file_name)
parameter.append((H[i], H_wt[i]))
DEM_arr = DEM_arr + H_wt[i]*temp_arr
Output = [DEM_arr, Input_DEMs, name, parameter]
return Output
def RiverNetwork(H1, H1wt, H2, H2wt, H3, H3wt, elev_min, elev_max):
print "Generating Digital Elevation Maps using FM2D algorithm"
#Generate first DEM with gradient = 1 (i.e. TRUE) and high H value
DEM_arr1 = MapGeneration_pure_python.midPointFm2d(max_level = 9, sigma = 1, H = H1, addition = True,\
wrap = False, gradient = 1,seed = 0, normalise=True,lbound=elev_min, ubound=elev_max)
pylab.imsave("Output/DigitalElevationModel1",DEM_arr1)
#Generate second DEM with gradient = 0 (i.e. FLASE) and medium H value
DEM_arr2 = MapGeneration_pure_python.midPointFm2d(max_level = 9, sigma = 1, H = H2, addition = True,\
wrap = False, gradient = 0,seed = 65, normalise = True,lbound=elev_min, ubound=elev_max)
pylab.imsave("Output/DigitalElevationModel2",DEM_arr2)
#Generate third DEM with gradient = 0 (i.e. FLASE) and medium H value
DEM_arr3 = MapGeneration_pure_python.midPointFm2d(max_level = 9, sigma = 1, H = H3, addition = True,\
wrap = False, gradient = 0,seed = 6, normalise = True,lbound=elev_min, ubound=elev_max)
pylab.imsave("Output/DigitalElevationModel3",DEM_arr3)
DEM_arr = DEM_arr1
(x_len,y_len) = DEM_arr.shape
#Get the co-ordinates having highest elev , required for catchment extraction
(max_x, max_y) = ndimage.maximum_position(DEM_arr)
for i in range(0,x_len):
for j in range(0,y_len):
#Combine 3 DEM's
DEM_arr[i][j] = (H1wt * DEM_arr1[i][j]) + (H2wt * DEM_arr2[i][j]) + (H3wt * DEM_arr3[i][j])
print "Iteratively removing sink using 3x3 window"
for p in range(0,6):
for i in range(1,x_len-1):
for j in range(1,y_len-1):
#Remove pits by 3 x 3 window
A = min(DEM_arr[i-1][j-1],DEM_arr[i-1][j],DEM_arr[i-1][j+1],\
DEM_arr[i][j-1],DEM_arr[i][j+1],DEM_arr[i+1][j-1],
DEM_arr[i+1][j],DEM_arr[i+1][j+1])
if DEM_arr[i][j] < A:
DEM_arr[i][j] = A + 1
#Initialize various arrays to hold flow direction, Flow accumulation,catchment info etc
Flow_arr = numpy.zeros((x_len,y_len) , dtype = "uint8" )
# River_arr will hold the River_accumulation matrix
River_arr = numpy.ones((x_len,y_len) , dtype = "int" )
# Catchment_boundary_arr will hold the Catchment boundaries
Catchment_boundary_arr = numpy.zeros((x_len,y_len) , dtype = "uint8" )
# Found_arr will hold the catchment with different labels
Found_arr = numpy.zeros((x_len,y_len),dtype = "uint8")
# Pour_point_arr keeps track of pour point of a catchment on the map
Pour_point_arr = numpy.zeros((x_len,y_len),dtype = "uint8")
Pour_point_list = [] #keep track of Pour_point in a list
pit_list = [] #contains all the pit in DEM
print "Assigning Flow Directions"
for i in range(1,x_len-1):
for j in range(1,y_len-1):
#Assign Flow direction
(value,dirn) =max(((DEM_arr[i][j] - DEM_arr[i-1][j-1])/1.41,3),\
(DEM_arr[i][j]-DEM_arr[i-1][j],2),((DEM_arr[i][j]-DEM_arr[i-1][j+1])/1.41,1),\
(DEM_arr[i][j] - DEM_arr[i][j-1],4),(0,8),(DEM_arr[i][j] - DEM_arr[i][j+1],0),\
((DEM_arr[i][j] - DEM_arr[i+1][j-1])/1.41,5),(DEM_arr[i][j] - DEM_arr[i+1][j],6),\
((DEM_arr[i][j] - DEM_arr[i+1][j+1])/1.41,7))
Flow_arr[i][j] = dirn
if dirn == 8:
# If there is a pit append it to the pit_list
pit_list.append((i,j))
label = 0 # will be used to assign labels to differnet catchments
#_____________Catchment Extraction_____________________________________________
print "Extracting Catchment and filling Depressions"
while len(pit_list) >= 1:
#_______________For each and every pit in the DEM do _________________________
stack = []
pit = pit_list.pop(0)
stack.append(pit)
label = label + 1 #increase the label being assigned to the catchment
#_______________________Identify catchment for each and every pit
catchment_pixels = []
catchment_pixels.append((DEM_arr[pit[0],pit[1]],pit[0],pit[1]))
while len(stack) > 0:
(p,q) = stack.pop(0)
Found_arr[p][q] = label
#Pop an element from stack check if its adjacent pixels exist and contribute
# its flow to the central pixel(pixel popped) then append it into list, continue
# this while stack gets empty
if pixel_exist(p-1,q-1,x_len,y_len):
if Flow_arr[p-1][q-1] == 7:
stack.append((p-1,q-1))
catchment_pixels.append((DEM_arr[p-1,q-1],p-1,q-1))
if pixel_exist(p-1,q,x_len,y_len):
if Flow_arr[p-1][q] == 6 :
catchment_pixels.append((DEM_arr[p-1,q],p-1,q))
stack.append((p-1,q))
if pixel_exist(p-1,q+1,x_len,y_len):
if Flow_arr[p-1][q+1] == 5:
catchment_pixels.append((DEM_arr[p-1,q+1],p-1,q+1))
stack.append((p-1,q+1))
if pixel_exist(p,q-1,x_len,y_len):
if Flow_arr[p][q-1] == 0 :
catchment_pixels.append((DEM_arr[p,q-1],p,q-1))
stack.append((p,q-1))
if pixel_exist(p,q+1,x_len,y_len):
if Flow_arr[p][q+1] == 4:
catchment_pixels.append((DEM_arr[p,q+1],p,q+1))
stack.append((p,q+1))
if pixel_exist(p+1,q-1,x_len,y_len):
if Flow_arr[p+1][q-1] == 1:
catchment_pixels.append((DEM_arr[p+1,q-1],p+1,q-1))
stack.append((p+1,q-1))
if pixel_exist(p+1,q,x_len,y_len):
if Flow_arr[p+1][q] == 2 :
catchment_pixels.append((DEM_arr[p+1,q],p+1,q))
stack.append((p+1,q))
if pixel_exist(p+1,q+1,x_len,y_len):
if Flow_arr[p+1][q+1] == 3 :
catchment_pixels.append((DEM_arr[p+1,q+1],p+1,q+1))
stack.append((p+1,q+1))
# Find catchment Outlet
pour_point = (max_x, max_y)
flag = 0
for i in range(0,len(catchment_pixels)):
(p,q) = ( catchment_pixels[i][1],catchment_pixels[i][2] )
label = Found_arr[p][q]
# Catchment Outlet will be the minimum catchment boundary pixel
if (Found_arr[p-1][q-1] != label or Found_arr[p-1][q] != label or Found_arr[p-1][q+1] != label or
Found_arr[p][q-1] != label or Found_arr[p][q+1] != label or Found_arr[p+1][q-1] != label or
Found_arr[p+1][q] != label or Found_arr[p+1][q+1] != label):# if pixel lie on boundary of catchment
Catchment_boundary_arr[p][q] = 255
if DEM_arr[ pour_point[0] ][ pour_point[1] ] > DEM_arr[p][q]:#if height of boundary is less then update pour point
pour_point = (p,q)
flag = 1
if flag == 1:
Pour_point_list.append((DEM_arr[pour_point],pour_point[0],pour_point[1]))
Pour_point_arr[pour_point] = 255
for i in range(0,len(catchment_pixels)):
if catchment_pixels[i][0] < DEM_arr[pour_point]:
#fill the depression in the catchment
DEM_arr[catchment_pixels[i][1],catchment_pixels[i][2]] = DEM_arr[pour_point]
print "Assignnig flow dirnection again after Depression filling"
for i in range(1,x_len-1):
for j in range(1,y_len-1):
# Again assign Flow direction again after filling the depressions
(value, dirn ) = max( ((DEM_arr[i][j] - DEM_arr[i-1][j-1])/1.41,3),(DEM_arr[i][j] - DEM_arr[i-1][j],2),((DEM_arr[i][j] - DEM_arr[i-1][j+1])/1.41,1),\
(DEM_arr[i][j] - DEM_arr[i][j-1],4),(0,8),(DEM_arr[i][j] - DEM_arr[i][j+1],0),\
((DEM_arr[i][j] - DEM_arr[i+1][j-1])/1.41,5),(DEM_arr[i][j] - DEM_arr[i+1][j],6),((DEM_arr[i][j] - DEM_arr[i+1][j+1])/1.41,7))
Flow_arr[i][j] = dirn
if value <= 0:
Flow_arr[i][j] = 8
# Calculate flow accumulation by calling Generate_River function
print "Performing Flow accumulation"
River_arr = Flow_accum.Generate_River( Flow_arr,River_arr,DEM_arr)
Distance_arr = city_block_dist_erosion.CityBlock(River_arr)
# Create a mask for differnet distances used for DEM erosion
print "Eroding DEM"
mask4 = [ Distance_arr <= 15 ]
mask5 = [ Distance_arr > 3 ]
mask3 = [Distance_arr == 3]
mask2 = [Distance_arr == 2]
mask1 = [Distance_arr == 1]
mask0 = [Distance_arr == 0]
max_flow_accum = numpy.max(River_arr)
# TODO - maybe change the block below - have already combined the DEMs. Weight erosion more simply
for i in range(0,x_len):
for j in range(0,y_len):
#Erode the landscape using diffent weighing factor for different distances from
#river while combining 3 DEM's
if mask0[0][i][j] == True:
DEM_arr[i][j] = 0.3*DEM_arr[i][j] + 0.45*DEM_arr2[i][j] + 0.19*DEM_arr3[i][j]
elif mask1[0][i][j] == True:
DEM_arr[i][j] = 0.3*DEM_arr[i][j] + 0.46*DEM_arr2[i][j] + 0.20*DEM_arr3[i][j]
elif mask2[0][i][j] == True:
DEM_arr[i][j] = 0.3*DEM_arr[i][j] + 0.46*DEM_arr2[i][j] + 0.21*DEM_arr3[i][j]
elif mask3[0][i][j] == True:
DEM_arr[i][j] = 0.3*DEM_arr[i][j] + 0.46*DEM_arr2[i][j] + 0.23*DEM_arr3[i][j]
elif mask4[0][i][j] == True and mask5[0][i][j] == True:
DEM_arr[i][j] = 0.3*DEM_arr[i][j] + 0.47*DEM_arr2[i][j] + 0.23*DEM_arr3[i][j]
else:
DEM_arr[i][j] = 0.3*DEM_arr[i][j] + 0.50*DEM_arr2[i][j] + 0.25*DEM_arr3[i][j]
#Output different statistics for display and further use
print "printing statistics ...see the Output Folder"
numpy.save("River.npy",River_arr)
numpy.save("DEM.npy",DEM_arr)
pylab.imsave("Output/River",River_arr)
pylab.imsave("Output/Catchment",Found_arr)
pylab.imsave("Output/CatchmentBoundary",Catchment_boundary_arr)
pylab.imsave("Output/Combined_eroded_DEM",DEM_arr)
pylab.imsave("Output/RiverDistance",Distance_arr)
return DEM_arr
