You can use WRF model for this reason like everyone else. But, I am just against using the current methods for down-scaling. My deeper understanding of the current methods is that the higher resolution we aim for, the more we are away from being scientific. You need to think beyond the state-of-the-art to know why I said this!
If you want to follow the state-of-the-art, just go ahead with WRF model. If you are able to think beyond the state-of-the-art, you never downscale the data available with the current methods.