Input data

Input data#

Pressure#

Atmospheric pressure comes in different flavors.

Surface pressure (sp) is the pressure at ground level, i.e. following the terrain.

In the following plot you can clearly see the decrease in surface pressure with altitude in areas with mountains.

This is NOT the variable used by MIKE 21 FM HD.

ds0.sp.plot()

<matplotlib.collections.QuadMesh at 0x7a5150e0cee0>

_images/15cf7ea31ca940d965afb513f52ccda0cac0f4fc7ea3f654152b2c1f8b06af51.png

MIKE 21 FM HD expects the atmospheric pressure reduced to Mean Sea Level. (msl)

In the following plot the pressure field has a much smaller range, since the variation with altitude have been removed.

ds0.msl.plot()

<matplotlib.collections.QuadMesh at 0x7a5150cf80a0>

_images/4779e55179dec5f949776868fdb47ffbe010b370db294a88ac2e524f386a20cd.png

Wind#

The wind velocity components U,V at 10m above the ground, are found in the two data variables, u10 and v10

ds0.u10.long_name

'10 metre U wind component'

ds0.v10.long_name

'10 metre V wind component'

Unstructured grid forcing#

E.g. roughness map

msh = mikeio.Mesh("data/FakeLake.dfsu")
msh.plot();

_images/83c3869907cddf400f5571890d44b00ca9773c27425e596f35c237095a593732.png

import numpy as np
import matplotlib.pyplot as plt

z = np.linspace(0,-30)
manning = 32 + (40 - 32)*(1 - np.exp(z/10))

plt.plot(z, manning)
plt.xlim(0,-30)
plt.ylabel("Manning")
plt.xlabel("Depth (m)")
plt.title("This is only an example...")

Text(0.5, 1.0, 'This is only an example...')

_images/4360e4766b2db81147fbaaef0e52ffe45f6a7b976f8698cb581dde6768b35db8.png

Extract the Z values in each element.

ze = msh.element_coordinates[:,2]

Create a new array of manning numbers based on the depth. This is only a hypothetical example of a data transformation based on the data in the file, not a best practise for bed friction.

me = 32 + (40 - 32)*(1 - np.exp(ze/10))

The primary use case of DFS files is for time-varying data, so in order to use this time-independent data, we have to introduce a new axis, a time dimension.

One simple way to do this is to use the numpy function atleast_2d.

me2d = np.atleast_2d(me) # Introduce time axis
me2d.shape

(1, 1011)

Now the array is a 2d array

from mikeio.eum import ItemInfo

my_ds = mikeio.Dataset(data=[me2d],
                time="2000-1-1", # For static data, this is not used by MIKE 21
                items=[ItemInfo("Roughness", EUMType.Mannings_M)],
                geometry=msh.geometry)
my_ds

<mikeio.Dataset>
dims: (time:1, element:1011)
time: 2000-01-01 00:00:00 (time-invariant)
geometry: Dfsu2D (1011 elements, 798 nodes)
items:
  0:  Roughness <Mannings M> (meter pow  1 per 3  per sec)

my_ds.Roughness.plot();

_images/c74970987f5d4ae5b8adb71107d767495240bf03ea84fa1b7d657a52477d6c4d.png

my_ds.to_dfs("manning.dfsu")

dfs = mikeio.open("manning.dfsu")
dfs

Dfsu2D
number of elements: 1011
number of nodes: 798
projection: PROJCS["UTM-17",GEOGCS["Unused",DATUM["UTM Projections",SPHEROID["WGS 1984",6378137,298.257223563]],PRIMEM["Greenwich",0],UNIT["Degree",0.0174532925199433]],PROJECTION["Transverse_Mercator"],PARAMETER["False_Easting",500000],PARAMETER["False_Northing",0],PARAMETER["Central_Meridian",-81],PARAMETER["Scale_Factor",0.9996],PARAMETER["Latitude_Of_Origin",0],UNIT["Meter",1]]
items:
  0:  Roughness <Mannings M> (meter pow  1 per 3  per sec)
time: time-invariant file (1 step) at 2000-01-01 00:00:00

ds = dfs.read()
ds["Roughness"].plot(title="Manning");

_images/06a2246564467b2d1c80acfa6b601b6a01998fceda29e7b54f0ba288719f1816.png

Input data

Contents

Input data#

Weather forcing#

Pressure#

Wind#

Convert to dfs2#

Sub-region#

Time#

Unstructured grid forcing#