Getting Started

Creating a model

We create a grid on which the blobs are discretized with using the Model class. Here, we specify the geometry of the model by the number of grid points, the lengths of the domain, the time step, the time length and the initial time of our realization.

In addition, we can specify the blob shape, drainage time and the number of blobs when creating a Model object. For more details about the geometry, take a look at the Model Geometry section.

    from blobmodel import Model

    bm = Model(
        Nx=10,
        Ny=10,
        Lx=10,
        Ly=10,
        dt=0.1,
        T=20,
        periodic_y=True,
        num_blobs=100,
        t_drain=1e10,
        t_init=10,
    )

Superposing blobs

We can now call the make_realization() method to sum up the individual blobs. The blob parameters are sampled from the according distribution functions (see parameter-distributions for further details). If we provide a file_name to the make_realization method, it will store the realization under this name on your machine.

    ds = bm.make_realization(file_name="example.nc")

The make_realization method can take two more arguments, speed_up and error, which can be helpful for integrating very large datasets. By setting spee_up to True, the code will truncate the blobs when the blob values fall under the given error value. The code assumes an exponential shape for the blobs when calculating the truncation position (see Blob Shapes for further details). Increasing the spatial resolution (the Nx and Ny arguments in the `Model) will lead to something like this:

Such animation can be plotted with the following code:

from blobmodel import show_model
show_model(dataset=ds, interval=100, gif_name="2d_animation.gif", fps=10)

Here, all the blobs move with the same speed and have the same shape and size. We will learn how to set distribution functions for the blobs parameters in parameter-distributions

The make_realization method returns the realization in the form of an xarray dataset. The superposed pulses are stored in the n variable of the dataset. The dimension coordinates are x, y and t for the horizontal, vertical and time coordinate, respectively.

We can now analyse the data using the convenient xarray syntax, for example:

    ds["n"].isel(y=0).mean(dim=("t")).plot()