search

MFEh2o / db

**Contains the main issue tracker for the MFE DB!** Functions for interacting with the MFE database, in script format. (See also MFEUtilities, which is an R package that includes many/most of the same functions).
1 stars 0 forks source link

sensorDB meteorological data QC #41

Open joneslabND opened 4 years ago

joneslabND commented 4 years ago

We should QC the meteorological data a bit more. For example, 2016 EL PAR is a lot higher than usual and saturates for the mid-day on most days.

kaijagahm commented 3 years ago

Put together some figures etc. PAR is amount of light, showing pattern throughout the day. Should be highest midsummer etc. Seems like there are mistakes interrupting this pattern--try to find them, visualize them.

kaijagahm commented 3 years ago

For 2016 at EL, we only see this problem in May. Screen Shot 2020-09-02 at 12 29 28

In 2015 at EL, all months also show this leveling-off pattern. Screen Shot 2020-09-02 at 12 30 00

Of course, it's possible that this isn't a strictly month-to-month pattern--e.g. some of the other months in 2016 may have the problem for a few days, just not enough to create a huge bar in the histogram. I can investigate that in more depth if we want.

kaijagahm commented 3 years ago

I started looking at other lakes and years, but I'm a little stumped because I don't know what the values are supposed to be. Is there a value that's necessarily too high, so I can filter using it? Is that kind of maxing out always a bad thing? I checked the metadata files, and they don't say much about what we'd actually expect here, data-wise, just about how the data were processed. Maybe I need to look at B3?

kaijagahm commented 2 years ago

This issue has a script started. See 'gh041_met_QC_sensor.Rmd' and 'gh041_metstation_QC_functions.R' in my archive.

ctsolomon commented 2 years ago

Here's another look at the 2015 EL PAR problem, which I ran into separately before remembering that there was an issue already started.

First, here's PAR plotted for EL and WL for a couple days at the beginning of June image

And here's a plot showing this over all of 2015 image