WXforum.net
Weather Related Organizations => WeatherUnderground => Topic started by: greyhame on April 20, 2019, 03:17:46 PM
-
I know this has been mentioned several times buried in threads, but I wanted to surface it to a full topic. Losing decimal degrees Celsius for Temperature and Dew Point in the new PWS UI feels like a real loss. I am on the west coast of Canada where we hover within decimal degrees of fog and freezing throughout much of the winter. Whole degrees Celsius feels like a large margin of error in reported temperature and dew point. Would really love to see that return to decimals. Attached is a comparison of whole vs. decimal degrees between the old UI and new for April 18 at my weather station.
[ You are not allowed to view attachments ] [ You are not allowed to view attachments ]
-
0.1C resolution is a must for metric users. Considering that the vast majority of the world uses metric for weather observations, it should be a priority. Evidently WU will be going back to that at some point. They have heard the complaints.
The odd thing is WU has 0.1 km/h resolution for wind, and .01 mm resolution for precipitation, which is MORE than needed. Go figure..
-
I also couldn't agree more - bring back the 0.1C resolution for the temperature. Also in the interests of data quality, I am dumb founded as to why you do the conversion to F for data storage. This is really silly as any data from those weather stations that send them in as Celsius and then converted to Fahrenheit causes loss of data accuracy so in reality significantly degrades the quality of the data.
-
Guys this is apparently beyond IBM, Deep Blue, Cloud, etc. computing power to calculate decimals so we shouldn't expect that ;).
Enough has been said about the update in general but judging from the whole updating process which was an epic fail/snafu that should go into guides like "How not to update production environment" I'm not even surprised such bugs exist. And yes - things like this should have been rectified in 1 day. They probably have 1 or 2 people working on the project so it is what it is. And not answering support requests is another thing that describes the company.
-
I'm sure this a top priority. Certainly it is higher priority than the webcam issues.
-
The problem seems to be the inability of those in charge of the updating to correctly convert between Imperial & metric or appreciate the implications of their actions to date.
Apart from the obvious outcome of dropping the decimal degrees in Celcius, there is also the incorrect use of the metric units commonly used.
The symbol for metre is m not mt; rainfall is normally measured in mm not cm; no it was not 67°C at home this afternoon (18°C) and my station is not 1161mt ASL but 353m.
Their navigation is also a little awry as stations in NZ that I checked were all placed around 600 Km out to the East in the Pacific.
So hopefully they get the coding sorted out soon and reintroduce the decimal degrees in Celcius as well.
-
Following this topic. I also think this issue has to be a high priority one. For metric, the decimals "tenths" are a must.
Thank you WU guys
-
Still no progress on getting 0.1C resolution displayed (both data and graphing) (Not that I'm using WU anymore to track my data.. I have other software to do that) Anybody hear of anything from WU on when this may be resolved? Not sure why it's taking so long.. how hard can it be to make such a straightforward and much requested change?
-
Temperature is displayed as an integer yet wind speed is shown to the first decimal place. Who or what can measure wind speed with that accuracy and why would you want to?
It seems to me that the variables used to hold these values are incorrectly defined. Temperature should be defined as single precision and wind speed as an integer, Programming 101.
-
Unfortunately I don't see this being fixed for some time as it will need a complete rewrite of all the variables use for both data ingestion and display. Holy crap what a job.
Someone really should have had a look at the standards for meteorological data and they would have been able to see the following variables, well as specified for a metric system :lol:
ONE SECOND DATA FORMAT.
Ident Name in Full Units Format Value
Tsam Air Temperature sample deg. Celsius (°C) *##.# 99.9
Wsam Wet Bulb Temp sample deg. Celsius (°C) *##.# 99.9
Hsam Relative Humidity sample percent (%) ### 999
Ssam Wind Speed sample knots ### 999
Dsam Wind Direction sample degrees (°) ### 999
R Rainfall millimetres (mm) ###.# 999.9
Psam Barometric Pressure sample hectopascals (hPa) ####.# 9999.9
where "#" / one digit ie. 0 to 9.
"*" / one digit or a minus sign ie. 0 to 9 or "-"
Note: If the sensor is missing or the data is invalid then value is replaced by 9’s