Replies: 1 comment
-
If the heads are above the ground surface, you need something to remove water. A common approach is to use DRNs at the elevation of the ground surface. These represent water that would be lost to overland flow/surface water. Ideally the flux that ends up in these should make sense based on what you know about the regional water balance. When solving a steady-state simulation, the initial heads should not matter if you're model is converging well. In a SS simulation, all the initial heads do is provide an initial condition to start trying to iteratively solve the GWF equation. It is common to set them to ground surface, as this is easy and often leads to decently quick convergence. |
Beta Was this translation helpful? Give feedback.
-
Hi all,
I'm using FloPy to do a 40year steady-state simulation but got the result in which the simulated groundwater head is much higher than the land surface over the lowlands (mountainous areas are okay). I got some feedbacks that there might be some bugs in my model set-up, but since I'm still very new to this model, I want to ask some general question about how the steady-state simulation works in modflow.
The most likely reason for this water-head-higher-than-land-surface problem could be the absence of water-removal terms in my model. I already included a drain package in my model but I wonder does the drain package can only simulate the groundwater flux to river? Theoretically the fate of groundwater flux should be ocean or the terminal rivers/lakes. I considered rivers and wells in my model, but the water head over ocean are set as constant 0m throughout the simulation. Could this cause limited amount of water-removal, and eventually, water head higher than land surface?
Another possible reason could be that the initial conditions can influence the water head. Currently I set the initial water head to 0m at all grid cells. Does this mean the water fluxes will accumulate based on this level, and therefore the final water head will be higher than the land surface over the lowlands?
My next question is not directly related to the problem, but since the simulation was conducted under the steady-state mode and there is only one time step in total, I wonder how does MODFLOW conduct the equilibrium simulation? Because if the initial water head is 0 all over the grids but the final water head is heterogenous in my study domain, there should be several time steps conducted as the so called "spin-up" period for the model to reach the equilibrium state. Should I change to the transient mode to see the results though?
I also attached the FloPy code below:
Beta Was this translation helpful? Give feedback.
All reactions