Dfsu 3D

Layered dfsu files comes in several different shapes:

• 3D

• 2D Vertical slice (transect)

• 1D Vertical profile

Two layer systems exist:

• sigma (terrain and surface following coordinates)

• sigma-z (sigma layers at the top and fixed z-layers at the bottom)

In sigma-layered files, all columns has the same number of layers. In sigma-z files, the number of z-layers can be different for different columns.

Layered dfsu files have a “hidden” first dynamic item called “zn” with the (dynamic) z-positions of the nodes.

Read the MIKE IO dfsu-3d documentation for more info.

import matplotlib.pyplot as plt
import mikeio

3D Sigma-z

filename = "data/oresund_sigma_z.dfsu"
dfs = mikeio.open(filename)
dfs

Dfsu3DSigmaZ
number of elements: 17118
number of nodes: 12042
projection: UTM-33
number of sigma layers: 4
max number of z layers: 5
items:
  0:  Temperature <Temperature> (degree Celsius)
  1:  Salinity <Salinity> (PSU)
time: 3 steps with dt=10800.0s
      1997-09-15 21:00:00 -- 1997-09-16 03:00:00

Apart from the normal dfsu properties, layered dfsu files have these properties:

print(f"Maximum number of layers: {dfs.n_layers}")
print(f"Number of sigma layers: {dfs.n_sigma_layers}")
print(f"Maximum number of z-layers: {dfs.n_z_layers}")
print(f"The layer number for each 3d element: {dfs.layer_ids}")
print(f"List of 3d element ids of surface layer: {dfs.top_elements}")
print(f"List of 3d element ids of bottom layer: {dfs.bottom_elements}")
print(f"List of number of layers for each column: {dfs.n_layers_per_column}")
print(f"The 2d-to-3d element connectivity table for a 3d object: {dfs.e2_e3_table[:3]} ...")
print(f"The associated 2d element id for each 3d element: {dfs.elem2d_ids}")

Maximum number of layers: 9
Number of sigma layers: 4
Maximum number of z-layers: 5
The layer number for each 3d element: [4 5 6 ... 6 7 8]
List of 3d element ids of surface layer: [    4     8    12 ... 17106 17111 17117]
List of 3d element ids of bottom layer: [    0     5     9 ... 17102 17107 17112]
List of number of layers for each column: [5 4 4 ... 5 5 6]
The 2d-to-3d element connectivity table for a 3d object: [array([0, 1, 2, 3, 4]) array([5, 6, 7, 8]) array([ 9, 10, 11, 12])] ...
The associated 2d element id for each 3d element: [   0    0    0 ... 3699 3699 3699]

The associated 2D geometry for a 3D file can be outputted in this way:

geom2d = dfs.geometry2d
geom2d

Flexible Mesh Geometry: Dfsu2D
number of nodes: 2090
number of elements: 3700
projection: UTM-33

geom2d.n_elements

3700

Layers in a 3D file

Get element ids for a specific layer with the get_layer_elements() method. Layer ids are 0-based (new in MIKE IO 0.10).

print("Number of elements per layer (5 z-layers, 4 sigma layers):")
for j in range(dfs.n_layers):
    print(f"Layer {j}: {len(dfs.get_layer_elements(j))}")

Number of elements per layer (5 z-layers, 4 sigma layers):
Layer 0: 10
Layer 1: 86
Layer 2: 187
Layer 3: 528
Layer 4: 1507
Layer 5: 3700
Layer 6: 3700
Layer 7: 3700
Layer 8: 3700

# the bottom layer is a list of the elements with the lowest id for each column
# For sigma-z files it is NOT the same as layer 1
print(f"Bottom layer: {len(dfs.bottom_elements)}")

Bottom layer: 3700

Surface layer of 3D file

ds = dfs.read(layers="top")
print(ds)

<mikeio.Dataset>
dims: (time:3, element:3700)
time: 1997-09-15 21:00:00 - 1997-09-16 03:00:00 (3 records)
geometry: Dfsu2D (3700 elements, 2090 nodes)
items:
  0:  Temperature <Temperature> (degree Celsius)
  1:  Salinity <Salinity> (PSU)

ds["Temperature"].plot();

max_t = ds['Temperature'].values.max()
print(f'Maximum temperature in top layer: {max_t:.1f}C')

Maximum temperature in top layer: 17.5C

Find position of max temperature and plot

Use numpy argmax() method to find the element with the largest value.

timestep = 0
max_elem_id = ds['Temperature'].isel(time=timestep).values.argmax()
top_element_coordinates = dfs.element_coordinates[dfs.top_elements]
max_x, max_y = top_element_coordinates[max_elem_id][:2]
max_x, max_y

(333934.08510239207, 6158101.508170851)

ax = ds['Temperature'].plot(figsize=(6,7))
ax.plot(max_x, max_y, marker='*', markersize=20);

Read 1D profile from 3D file

Find water column which has highest temperature and plot profile for all 3 time steps using dynamic z information.

dsp = dfs.read(x=max_x, y=max_y) # select vertical column from dfsu-3d 

dsp['Temperature'].shape   # 3 timesteps and 4 layers (no z-layers at this position)

(3, 4)

dsp['Temperature'].plot();

Read vertical slice

filename = "data/oresund_vertical_slice.dfsu"
dfs = mikeio.open(filename)
dfs

DfsuVerticalProfileSigmaZ
number of elements: 441
number of nodes: 550
projection: UTM-33
number of sigma layers: 4
max number of z layers: 5
items:
  0:  Temperature <Temperature> (degree Celsius)
  1:  Salinity <Salinity> (PSU)
time: 3 steps with dt=10800.0s
      1997-09-15 21:00:00 -- 1997-09-16 03:00:00

print(dfs.bottom_elements[:9])
print(dfs.n_layers_per_column[:9])
print(dfs.top_elements[:9])

[ 0  5 10 15 20 24 28 32 36]
[5 5 5 5 4 4 4 4 4]
[ 4  9 14 19 23 27 31 35 39]

da = dfs.read(items="Temperature")[0]

da.plot();