Bad dew point readings

[swyman18]

Hi Steve,

Wanted to run something by you. This morning I got a bad dew point reading in my log file, as seen below, at 07:23. I suspect that this is from one of of the occasional bad temp and dew point readings I get due to interference. Now, in this case, I believe the reason why the temp reading looks ok is because I am using the EWTempdiff setting to filter out bad temp readings. I also have the Calculate Dew Point setting active in the station configuration settings. So, I believe what is happening is that Cumulus is calculating the dewpoint based on the bad "filtered" temperature recieved from the console. I guess what I'm wondering is if there is a way to make Cumulus calculate the dew point based on the latest "good" temp value? Or perhaps have an EWDewPointdiff setting implemented to allow filtering of bad dew point values?

Let me know your thoughts... thanks!

-Scott



27/11/09,07:21,66.0,82,60.3,3.6,7.6,155,0.00,0.26,30.05,24.15,76.6,65,3.8,66.0,66.0
27/11/09,07:22,66.0,82,60.3,3.5,7.6,154,0.00,0.26,30.06,24.15,76.6,65,0.0,66.0,66.0
27/11/09,07:23,66.0,82,40.3,3.5,7.6,153,0.00,0.26,30.06,24.15,76.6,65,3.8,66.0,66.0
27/11/09,07:24,66.2,82,60.5,3.5,7.6,153,0.00,0.26,30.06,24.15,76.5,66,3.1,66.2,66.2
27/11/09,07:25,66.2,82,60.5,3.5,7.6,149,0.00,0.26,30.06,24.15,76.5,66,3.8,66.2,66.2

[steve]

The intention was that it wouldn't calculate the dew point using a rejected temperature. That's what I did for the Fine Offset code (which is where I did this hack originally); it's possible that I forgot to do the same in the La Crosse code - I'll check.

[steve]

I've looked at the Cumulus-calculated dew point code, and it's in the 'common' code, so it's the same for all station types. What it does is that every time it gets a humidity reading, it calculates the dew point based on that reading and the last good temperature reading. So I can't explain how you got that dew point with those figures. But just because those figures are in the log entry doesn't mean that those are the values that were used to calculate the dew point, if you see what I mean? There may have been odd readings between log entries.

Have a look in the diags file for around that time, and see if there is any mention of bad readings being ignored.

[swyman18]

OK, I understnd what you're saying... I checked the diags and there is no indication of anything odd around 07:23. I'm guessing there must have been an odd humidity reading at some point between the 1 minute log entries that caused the bad dew point calculation.

Thanks anyway for checking.

-Scott