The folks at UCAR had a few comments on the dashboard:
The climate projections begin in 1850. Do we want to show that entire time series? If not, at what time point should the projections start?
"I think 1990 would be a good starting point."
How should we handle the uncertainty? Currently, the standard deviation is shown as a shaded region around the time series.
"Standard deviation works for me.. but maybe adding a dashed line for 2 standard deviations around the mean would be good to get a sense of how big the variability could get.."
"Should we think about merging the current time series and the projections? Or perhaps show the last few years of satellite data and then transition to the projections? Something like this?"
Yes I like this idea. If we keep the separate tabs then I think we should rename "Forecast" to "Projection", and perhaps rename "Climatological" to "Historical"?
"One idea that might be nice is to show a monthly climatology for each decade starting in the 1990s and going to the 2090s (perhaps a different color line for each decade or let the user decide which decades they want to see on the plot)... this way we can see, e.g., how hot the summers get, or the minimum pH during winter, etc..."
"Zooming in and out were not obvious, particularly the zoom-out
Not clear to me that the SST time series are from satellite. And also coinsuign that there are two SST panels (one for MUR, and one for CoralReefWatch), can they be overlaid instead?
Year 2100 at all sites seems wrong to me, there's always a big jump in SST."
The folks at UCAR had a few comments on the dashboard:
"I think 1990 would be a good starting point."
"Standard deviation works for me.. but maybe adding a dashed line for 2 standard deviations around the mean would be good to get a sense of how big the variability could get.."
Yes I like this idea. If we keep the separate tabs then I think we should rename "Forecast" to "Projection", and perhaps rename "Climatological" to "Historical"?
"One idea that might be nice is to show a monthly climatology for each decade starting in the 1990s and going to the 2090s (perhaps a different color line for each decade or let the user decide which decades they want to see on the plot)... this way we can see, e.g., how hot the summers get, or the minimum pH during winter, etc..."
"Zooming in and out were not obvious, particularly the zoom-out
Not clear to me that the SST time series are from satellite. And also coinsuign that there are two SST panels (one for MUR, and one for CoralReefWatch), can they be overlaid instead?
Year 2100 at all sites seems wrong to me, there's always a big jump in SST."