search

NOAA-PMEL / LAS

Live Access Server
https://ferret.pmel.noaa.gov/LAS/
The Unlicense
13 stars 5 forks source link

Gray mooring symbols #1754

Open karlmsmith opened 6 years ago

karlmsmith commented 6 years ago

In the latest SOCAT of datasets current waiting for QC, some of the mooring show up as light gray (or white?) box outlines. (In the attached screenshot, South Alaska near land and much further into the Pacific, Hawaii, equator.) There are difficult to see, and not sure why they are this color.

screen shot 2018-01-23 at 9 57 23 am
karlmsmith commented 6 years ago

Marking with major severity because of the difficulty in detecting these datasets on the map.

AnsleyManke commented 6 years ago

If data is missing, the ribbon plots draw the symbol at the location colroed with gray. (Check this by switching to color by a variable like longitude or day-of-year which should not have any missing data.) We're over-plotting all the data at the mooring locations, so the color is for the last data in time. If the last data is missing, it'll look like this.

One solution would be to change the scripts so that they do Mooring plots with /MISSING=blank so any data that's there will be drawn. That would make them disappear if there's no data at all. Another slightly more complicated answer would be to plot the last-valid data. That'd also speed up the plotting,

AnsleyManke commented 6 years ago

A decent answer is to draw the location of the mooring on the plot with black not gray. If there's data then the square is colored as always, otherwise it'll show up drawn with a thin black line. The thicker one out in the middle of this plot is a place where there are a couple of moorings at slightly different locations.

5a410054479abebfb609e1da7b75d61a_plot_image

karlmsmith commented 6 years ago

Probably need to do this for only moorings? The change shows lots of black lines for actual cruises, which is sort of distracting and might have been why they were changed to gray. So might need to reconsider this change. (Especially in view of my next comment.)

karlmsmith commented 6 years ago

The actual problem here is why is there no fCO2_recommended values? Looking at the dataset 316420090613, all the fCO2 derived values are all-missing. But xCO2 dry at equi temp is given along with equi temperature, equi pressure, and sal. So seems like there should have been an fCO2_recommended from method 1.