While picking through the existing implementation, I've had some thoughts as to how I would like to interact with the power/storms workflow.
I appreciate the point of open-gira is to do global analyses, but I think we should prioritise an ability to execute a small unit of work, because it will make it easier to:
a) Further develop the workflow
b) Test and verify the correctness of the program (in a reasonable amount of time)
c) Compare scenarios
d) Use for finer grained analysis in future
Location
This is one of the following:
Storm basin
Country
Convenience classifier (e.g. "Caribbean" which maps either directly to a polygon, or to a set of countries)
Finer level admin region name (down to admin-1 or 2)?
Processing this argument generates a (multi-)polygon to select storms by intersection with. Perhaps we can then drop the concept of a box (the coarser of the two current grids). Instead, work on the current unit level, where wind speed is assumed to be constant within a unit.
Storm
One of the following:
Single storm name (e.g. synthetic storm ID, or historic storm name/ID)
Class of storm set (e.g. all HadGEM-3, or all historic). This is similar to the current STORM_MODEL variable.
Processing this argument acquires track data for storms matching the specification.
While picking through the existing implementation, I've had some thoughts as to how I would like to interact with the power/storms workflow.
I appreciate the point of
open-girais to do global analyses, but I think we should prioritise an ability to execute a small unit of work, because it will make it easier to: a) Further develop the workflow b) Test and verify the correctness of the program (in a reasonable amount of time) c) Compare scenarios d) Use for finer grained analysis in futureLocation
This is one of the following:
Processing this argument generates a (multi-)polygon to select storms by intersection with. Perhaps we can then drop the concept of a
box(the coarser of the two current grids). Instead, work on the currentunitlevel, where wind speed is assumed to be constant within a unit.Storm
One of the following:
STORM_MODELvariable.Processing this argument acquires track data for storms matching the specification.