Closed hdrake closed 4 years ago
Currently, the performance metrics implemented in xskillscore (https://github.com/raybellwaves/xskillscore/blob/master/xskillscore/core/deterministic.py) do not allow for weighting by grid cell area when calculating a global-mean skill metric (as is customary in the literature, e.g. Reichler et al. 2008 https://journals.ametsoc.org/doi/abs/10.1175/BAMS-89-3-303).
xskillscore
Step 1 (a simple hack): multiply fields by a grid-cell area field, calculate the skill metric, and divide by the global-mean grid cell area.
Step 2: submit a PR to xskillscore (https://github.com/raybellwaves/xskillscore/) to make this a supported feature.
Now implemented in master branch https://github.com/hdrake/cmip6hack-multigen/pull/12
Will wait to close this issue until area-weighting has been merged into the master branch of xskillscore.
Currently, the performance metrics implemented in
xskillscore
(https://github.com/raybellwaves/xskillscore/blob/master/xskillscore/core/deterministic.py) do not allow for weighting by grid cell area when calculating a global-mean skill metric (as is customary in the literature, e.g. Reichler et al. 2008 https://journals.ametsoc.org/doi/abs/10.1175/BAMS-89-3-303).Step 1 (a simple hack): multiply fields by a grid-cell area field, calculate the skill metric, and divide by the global-mean grid cell area.
Step 2: submit a PR to
xskillscore
(https://github.com/raybellwaves/xskillscore/) to make this a supported feature.