We need to be able to aggregate multiple bluesky models into one model. In other words, given a list of N model runs over a time period p, we need to be able to select x hours of every Nth model to combine into a new raster brick. This utilizes the "best" forecasts for each N model over the time period p, assuming that predictions get worse as p increases.
We need to be able to aggregate multiple bluesky models into one model. In other words, given a list of N model runs over a time period p, we need to be able to select x hours of every Nth model to combine into a new raster brick. This utilizes the "best" forecasts for each N model over the time period p, assuming that predictions get worse as p increases.