3Di supports to methods for computing infiltration. The so-called simple infiltration and the Horton based infiltration. The simple infiltration uses an infiltration rate that is constant in time, while the Horton based infiltration is based on the concept of infiltration that reaches in time an equilibrium rate. In 3Di, the infiltration computed with the first method removes the water from the model. For the second method, a ground water layer is defined from which water can also exfiltrate in case the groundwater level reaches the surface level.
Infiltration is the process of water slowly sinking into the soil. The infiltration rate depends on the type of soil and the land coverage. The infiltration rate is constant in time throughout the simulation. There are two exception for this: 1) the availability of water and when a user defines a maximum infiltration capacity.
In 3Di, the infiltration rate is defined for subgrid cell and defined in mm/day. The infiltration is computed per computational cell. Therefore, the infiltration rate per pixel is translated to values per computational cell and can depend on the water level.
When using the subgrid method, water start filling a cell from the lowest subgrid cell. When simulating rainfall run-off scenarios, the overall water depths are relatively small, therefore often only a small part of a computational cell is wet. If in this part of the cell the infiltration rate is low, the infiltration will be limited. However, in reality rain falls over a full spatial domain, and reaches the lowest areas only after a while, but might be infiltrated before it reaches that ares. Therefore, the total infiltration rate per computational can be made dependent of the rain. Besides defining the infiltration rates, a user defines also the infiltration_surface_option that can be found in the global settings table. In the Figure below, an overview is given of the various types.
The first option, defines the infiltration rate in a cell based on the assumption that the full cell is wet. In other words, the infiltration rate in a cell is independent of the water level (left panel). The second option, defines the infiltration rate on the cumulative rates in the wet areas (middle panel). In this case the infiltration rate is dependent of the water level. The third option is the default option and is a combination of the two previous options. In this case the infiltration rate of the full cell domain is used when it is raining and only the infiltration rate of the wet area is used when it is dry (right panel).
In 3Di, many raster input variables are stored in tables. This yields as well for the infiltration rate, as this allows the infiltration rate to be dependent of the water level and for an efficient storage of the information. The table increment for the infiltration tables equals the 2D increment.
Maximum infiltration capacity
The maximum infiltration capacity is the maximum volume of water that can infiltrate during one simulation at a certain location. The user defines a layer with a thickness per pixel [meters]. Based on this, the capacity per pixel is computed by multiplying the thickness with the pixel surface area. The maximum infiltration per computational cell is the sum of the capacity of all pixels within that area. It will stop infiltrating when the maximum infiltration is reached.
The user can define all the aspects for simple infiltration in the simple_infiltration_table.
Horton based infiltration
The second possibility to add infiltration to a system, is infiltration based on the Horton equation. This describes a infiltration rate that is initially higher and decays to an equilibrium infiltration. This type of infiltration is always used in combination with a groundwater level. In such, the infiltration is always limited. More information about the Horton based infiltration can be found here Horton based infiltration.
Infiltration extract water from the model. If less water is available in the node or cell then what is to be extracted, 3Di will limit the extraction. Just before the cell becomes dry, the infiltration rate will be less than the infiltration rate in the user input. This is done to guarantee stability and capture reality better.
A Surface-Subsurface Model; balancing speed, accuracy and reality. Stelling and Volp (to be published)