Each WRF domain should be at least 100x100 grid points (e_we and e_sn).
When running a simulation, the input data for the lateral boundaries, every so many intervals (e.g., every 3 hours), are much more coarse than the simulation you’re trying to run.
Synoptic-scale systems typically move through a large area in a short amount of time. If a domains is too small, there isn’t enough space for the model to develop anything of interest before the large-scale parcels have propagated through the domain and exited the other side. This is sometimes referred to as “sweeping,” meaning the input essentially wipes out what WRF is trying to simulate at high resolution, with different physics and higher-res terrain, landuse, the dynamics, and all the other features you want. Even if you’re only interested in microscale events, the large-scale factors are important for forcing.
Essentially, if your domain is too small, you’re basically going to get the output as if the domain was still a coarse-resolution. You want your domain to have enough space so that you can actually see the effect.
This is not a computational limitation. It’s more an issue of reasonable and beneficial output.
Also note that the model will use the rows and columns around the edges of the domain for communication. This is set, by default, to 5 rows and 5 columns (it can be changed with the &bdy_control namelist parameter ‘spec_bdy_width,' but shouldn't be smaller than 5). This means that if e_we = 100 and e_sn = 100, at a minimum, the domain is already cut down to 90x90.
When running a simulation, the input data for the lateral boundaries, every so many intervals (e.g., every 3 hours), are much more coarse than the simulation you’re trying to run.
Synoptic-scale systems typically move through a large area in a short amount of time. If a domains is too small, there isn’t enough space for the model to develop anything of interest before the large-scale parcels have propagated through the domain and exited the other side. This is sometimes referred to as “sweeping,” meaning the input essentially wipes out what WRF is trying to simulate at high resolution, with different physics and higher-res terrain, landuse, the dynamics, and all the other features you want. Even if you’re only interested in microscale events, the large-scale factors are important for forcing.
Essentially, if your domain is too small, you’re basically going to get the output as if the domain was still a coarse-resolution. You want your domain to have enough space so that you can actually see the effect.
This is not a computational limitation. It’s more an issue of reasonable and beneficial output.
Also note that the model will use the rows and columns around the edges of the domain for communication. This is set, by default, to 5 rows and 5 columns (it can be changed with the &bdy_control namelist parameter ‘spec_bdy_width,' but shouldn't be smaller than 5). This means that if e_we = 100 and e_sn = 100, at a minimum, the domain is already cut down to 90x90.