OpenSprinkler Forums OpenSprinkler Unified Firmware Incorrect historic Zimmerman calculation using Weather Underground PWS

Viewing 5 posts - 1 through 5 (of 5 total)
  • Author
    Posts
  • #83783

    franzstein
    Participant

    I have checked the scales received from the weather service for the past days, if using Weather Underground PWS. Only the scale received for Yesterday corresponds to the calculation procedure defined by Zimmerman. Please look at the details below:

    Weather Service Request

    weather.opensprinkler.com/weather1.py?loc=49.47432,10.94365&wto=”h”:100,”t”:100,”r”:100,”bh”:65,”bt”:70,”br”:0,”provider”:”WU”,”pws”:”IFRTH239″,”key”:”my_api_key”

    results in:

    &scale=4&tz=56&sunrise=423&sunset=1156&eip=1469121255&rawData={“wp”:”WU”,”h”:83.53,”p”:0.34,”t”:67.7}&errCode=0&scales=[4,40,49,53,48,46]&ttl=81451159
    The Weather Underground Weather History for IFRTH239 lists:

    Date Temperature Average Humidity Average Precipitation Scale received -> calculated
    September 21 19.8 °C 84 % 8,64 mm 4 -> 4 %
    September 20 20,2 °C 82 % 0.00 mm 40 -> 77 %
    September 19 18.4 °C 78 % 0.00 mm 49 -> 68 %
    September 18 18.3 °C 80 % 0.00 mm 53 -> 66 %
    September 17 14.3 °C 88 % 0.00 mm 48 -> 29 %
    September 16 14.8 °C 80 % 0.00 mm 46 -> 40 %

    The scales received for the days before yesterday seemed to be far too low calculated.
    I have done a quick check for other PWS weather locations, and it looks like it is a general issue. I’m not sure anymore, but I have seen the wrong calculation for the first time having precipitation reported. I may be wrong in my findings, but please have a look at this issue.

    #83784

    Ray
    Keymaster

    Please note that the scales array contains the multi-day average, which is a cumulative average. For example, the second number (40) is NOT Sep 20’s number, rather, it is the average of Sep 20’s and Sep 21’s (so (77+4)/2 = 40%). It is for programs that run every 2 days (or 3, 4, 5 days etc.) so that they should be using the average watering scale that accounts for multiple days of watering data.

    #83785

    franzstein
    Participant

    Ok. Thanks for the explanation.

    #83789

    franzstein
    Participant

    One more question:

    After the heavy rain on Sep 22, the request

    weather.opensprinkler.com/weather1.py?loc=49.47432,10.94365&wto=”h”:100,”t”:100,”r”:100,”bh”:65,”bt”:70,”br”:0,”provider”:”WU”,”pws”:”IFRTH239″,”key”:”my_api-key”

    results on Sep 23 in

    &scale=0&tz=56&sunrise=425&sunset=1154&eip=1339803529&rawData={“wp”:”WU”,”h”:99,”p”:1.48,”t”:51.4}&errCode=0&scales=[0,0,0,0,0,0]&ttl=65634576

    Are these multi-day average values correct?

    #83791

    Ray
    Keymaster

    The values are correct. A large amount of rainfall on a day can offset the watering scales over multiple days, making them 0.

Viewing 5 posts - 1 through 5 (of 5 total)
  • You must be logged in to reply to this topic.

OpenSprinkler Forums OpenSprinkler Unified Firmware Incorrect historic Zimmerman calculation using Weather Underground PWS