As discovered in https://stackoverflow.com/questions/68477310/issue-using-mpcalc-advection-to-calculate-advection-of-a-scalar-field, the [QG analysis notebook](https://unidata.github.io/python-training/workshop/MetPy_Advanced/qg-analysis/) appears to use pre-MetPy 1.0 API for advection. So, this looks like it was another notebook unfortunately missed in the updates of https://github.com/Unidata/python-training/pull/113.