Create a CI subset of the HSC data for image differencing

Yusra AlSayyad generated an HSC dataset as part of the June '19 False Positive Sprint. This ticket is to identify a small subset to regularly process with CI and to create an ap_verify-compatible repository from it.

At the moment it will be necessary to use a template from the same time period; a future ticket will replace the output of this ticket with a dataset where PDR1 data is used for templates and PDR2 for science.