In support of the Landsat8 data available through M.App Chest (and other sources, such as the AWS S3 Bucket (https://aws.amazon.com/public-datasets/landsat/), it's useful to have a Model/Recipe which enables you to easily create an MSI layer stack out of the appropriate TIFF files without needing to individually select all of them.
If a USGS Landsat8 .tar.gz archive file is imported using the ERDAS IMAGINE USGS Landsat7 / Landsat8 importer it will automatically create appropriate Pan and Multispectral image files. However some customers unzip the data manually, which creates a directory structure where the data is stored as separate TIFF files per band. This sort of structure is also created when you download Landsat 8 data from sources such as M.App Chest and AWS S3. This structure is inconvenient for using in processes which expect multispectral data.
This Model is designed as a utility to stack individual Landsat8 TIFF files into a single Multispectral image, based on the standard MSI composite (Bands 1 through 7). The user selects the name of the Band1 file, where the naming convention is assumed to be <root>_Bn.tif. The other layers will be automatically selected from the same directory based on this pattern. The output file is automatically named <root>_msi.img.
The model could be easily modified to stack more (or less) bands, such as adding the TIRS bands (10 and 11), or to apply to other sensors such as Sentinel-2.
Band1 In: Input the TIFF file which is the first band of an unzipped Landsat8 archive.