Redo stressor standardization

[david-beauchesne]

At the moment, I standardized the drivers only with their own data, meaning for example that demersal fisheries for 2010-2015 was transformed only using the data from that period. However, this makers the results for ther change in cumulative effects over time less reliable than if standardization was made from the joint distribution of demersal fishries from 2010-2015 and 2016-2021. This should be changed accordingly.