Masking

[znicholls]

Another question (sorry for having so many @miturbide!). I wanted to check that I have correctly understood how your masking works.

ATLAS/aggregated-datasets/scripts/calculate_regional_means.R

Line 33 in 8824f76

maskland <- binaryGrid(mask, condition = "GT", threshold = 0.999, values = c(NA, 1))

suggests that any region with a surface land fraction greater than 0.999 is classified as land.

ATLAS/aggregated-datasets/scripts/calculate_regional_means.R

Line 34 in 8824f76

masksea <- binaryGrid(mask, condition = "LT", threshold = 0.001, values = c(NA, 1))

suggests that any region with a surface land fraction less than 0.001 is classified as sea.

Does this mean that any cell with surface land fraction greater than 0.001 and less than 0.999 (about 5% of cells looking at https://github.com/SantanderMetGroup/ATLAS/blob/master/reference-grids/land_sea_mask_1degree.nc4) is not included in either the sea or land calculation? In other words, only cells which are clearly one or the other are included in the masked averages?

[gutierjm]

Yes. We compute land and ocean results from purely land or ocean griboxes and discard the mixed (coastal) ones.

[znicholls]

thanks @gutierjm, do you have any thoughts on #11 ?