Weather Bug - What data is available?

[JMC]

IM,

 

Ok, I'll try CTRSH for a while and see what happens. I've been reluctant to change away from KAUN simply because most of the others in a reasonable distance are about 1,000' lower in elevation and that distorts their relevance quite a bit. Not as bad as zeroing out the day's rainfall though, so it may be an improvement. Fingers crossed.

 

mike

[IndyMike]

Hello Mike,

 

I've been monitoring 3 airports and 2 schools in your area. The feeds from the schools (CTRSH in particular) look good. The airport data is too screwed up to even comment on. High and low temps are being reset multiple times during the day. Using these values to calculate ET will be totally nonsensical.

 

I've also monitored stations in my own area. Same result - the school feeds look good while the airports are totally out to lunch (one actually reverses it's time feed).

 

I do not know when when this started. I was using my airport feed at times over the summer and this was not occurring.

 

I captured the following data using an Excel program to call the Weatherbug API. It was totally independent of the ISY. Data from Weatherunderground shows the Airport sites as reliable.

 

Good feed:

1) Low temperature hits a minimum and maintains until the midnight reset.

2) High temperature hits a maximum and maintains until the midnight reset.

 

 

Bad Feed

1) Low temperature reset at 16:00

2) High and low temp track current temperature from 16:00 to midnight.

3) Midnight data captured by the ISY: High temp = Low Temp; Average temp is actually > High temp.

[JMC]

Thanks, greatly appreciate your efforts. I'll stick with CTRSH for the time being but it's too bad that I can't find a closer school to monitor, there are two schools in my area that are no more than 50 feet different in elevation from me.

 

I never noticed the change in the data from KAUN until the rains started and I noticed the rainfall total being reset. It does make me wonder though, if some of my previous issues with the irrigation system wanting to start applying water way too soon after a rain might not be connected. I'll monitor the actual soil moisture as soon as the rains stop and see how long it says the system should wait before starting irrigation, and compare that with what the system thinks it wants to do.

 

mike

[TJF1960]

IndyMike,

 

You mentioned monitoring 3 airports, by any chance was KMYV one of them?

So far today, as I intermittently checked on the rain today, WB has been reporting rain from that site. Just wondered if it was one of the three.

 

Thanks,

Tim

[IndyMike]

Hi Tim,

 

I started monitoring KMYV after I saw your post. The Airports have been going haywire just after 16:00 your time daily. Today was no exception.

 

The temperatures and rainfall on all three of the following stations reset at 16:15 your time. I actually checked ~ 15 NOAA stations from San Diego to Seattle. All of them reset after 16:00.

 

I'm 2,100 miles away in Indiana. My Airport station (KSBN) is due to go down at 19:54 tonight (assuming the pattern is consistent).

 

 

 

 

 

[Panda88]

I wonder if WB uses GMT time (8 PST is midnight GMT)

Do we see rain picking up after 4pm, but before midnight?

I really wonder if one could simple make the daily summary at midnight GMT and then use that for calculating the irrigation - and temp etc probably needs to be done differently -

 

Does anybody have experience with a private weather station feeding WB? will that also reset @ 4pm?

[TJF1960]

Thanks so much Mike, that's what I was afraid of. I wish I had a school close by.

 

Tim

[GuitarBuilder]

Don't put too much faith in Weatherbug. I found my local stations to have enough errors or gaps in data to throw off the evapotranspiration calculations. Case in point: it's been raining for 4 days and the local school station is reporting 0 inches of rain. Never mind the lack of light data.

 

This approach is fine from a hobbyist standpoint, but is not reliable enough for professional or home automation use. The latter requires 99+% reliability without user intervention. I have to constantly second-guess my ISY program to make sure the yard is watered properly (right now it's trying to water my plants in the middle of a thunderstorm!). It gets very old after a while!

 

We need a 100% reliable source of information to base our evapotranspiration calculations on, and Weatherbug is not it.

[IndyMike]

Unfortunately, the backyard stations appear to be offline as well.

 

Weatherbug is in the process of migrating them to the cloud: http://www.wxforum.net/index.php?topic=21904.0

[JMC]

IndyMike:

 

I attached the current data from CTRSH. The Rain Today is shown as 0.98" which is wrong. In fact, CTRSH hasn't changed that value for at least 3 days. It seems as though at least one school station is also failing to deliver correct data, only this time instead of resetting the rainfall when it shouldn't, this station appears to not be resetting the rainfall when it should.

 

I haven't checked the data at exactly midnight so it is possible it is resetting to 0 and then restoring the 0.98" value some time later. In any event, my Irrigation Requirement has not gone down at all even though we have had consistent and significant rainfall for the last several days. Also, the Water Deficit Yesterday value has been equal to the Eto value each day and that is something I understand should never happen when there is significant rainfall.

 

Right now I have the irrigation system completely shut down and I plan to leave it that way until I start seeing some realistic values.

 

mike

201403032145.xml

[larryllix]

IndyMike:

Right now I have the irrigation system completely shut down and I plan to leave it that way until I start seeing some realistic values.

 

mike

 

Wouldn't a few soil moisture sensors with some averaging and outlier logic and alarms be a much more reliable source than attempting to use a lot of fudged factor math based on unreliable information from some school kids?

[JMC]

I use a handheld soil moisture sensor to oversee the irrigation operation. If you know of self-powered wireless moisture sensors that could be put in the ground and provide data to a control system I would greatly appreciate hearing about them.

 

mike

[IndyMike]

Mike,

 

We have two different problems here:

1) CTRSH is transmitting temperature reliably - it is not resetting during the day. The problem is that it does not transmit rain data. It provides "N/A" in the rain data field.

2) I think you have uncovered a bug in the handling of the "N/A" response. The ISY is expecting numeric data and can't convert the text provided. As a result, I believe it is using the last valid data it received (From KAUN?). If you switch to KAUN (currently at 0.05") and switch back to CTRSH the KAUN value will be retained.

 

Could you try switching to FRKSH? It does provide rain data and is currently registering 0.02 inches.

 

On the subject of a wireless moisture sensor - I purchased a Toro Xtra Smart Soil Sensor last year http://www.amazon.com/Toro-53812-Smart-Moisture-Sensor/dp/B007J0M93U

 

The sensor uses soil conductivity to "estimate" the water content. It uses two 4"(??) metal probes to measure the conductivity. In my mind, the probes should be at least 6". Battery powered, it transmits RF to a base unit and has decent range (I have it roughly 150' from the base). The base is basically a programmable switch that can be used to "inhibit" your irrigation system.

 

I had the switch connected to a IOlinc so I could log it's activity and compare it with the ISY ET calculations (I used the ISY to control my system). I can't give you a good or bad review on the unit. We had so much rain last year, I rarely had to run my irrigation system. I did get the impression that the unit's probes would sometimes become "isolated" from the soil causing low readings. I don't have definitive data on this, and it could be due to my very sandy soil.

 

Unfortunately, winter came quick and hard this year. I barely got my sprinkler system blown out before the first storm. The transmitter is still in the back yard buried under several feet of snow. If it survives, I'll give it some points for durability.

 

IndyMike:

 

I attached the current data from CTRSH. The Rain Today is shown as 0.98" which is wrong. In fact, CTRSH hasn't changed that value for at least 3 days. It seems as though at least one school station is also failing to deliver correct data, only this time instead of resetting the rainfall when it shouldn't, this station appears to not be resetting the rainfall when it should.

 

I haven't checked the data at exactly midnight so it is possible it is resetting to 0 and then restoring the 0.98" value some time later. In any event, my Irrigation Requirement has not gone down at all even though we have had consistent and significant rainfall for the last several days. Also, the Water Deficit Yesterday value has been equal to the Eto value each day and that is something I understand should never happen when there is significant rainfall.

 

Right now I have the irrigation system completely shut down and I plan to leave it that way until I start seeing some realistic values.

 

mike

[JMC]

IndyMike,

 

I made the change to FRKSH and we'll see what happens. If all works right, ETo and Yesterday's Water Deficit should be different tomorrow. This is not a long term fix though, summer temps in Fair Oaks can be significantly different from those here.

 

mike

[JMC]

Last night's calculation produced a slightly smaller change than I expected but at least the numbers were not frozen from the previous day. We're supposed to get some rain tonight so I should be able to observe again tomorrow.

 

mike

[bpwwer]

Mike,

 

I just looked at this thread for the first time. It sounds like you might be close to my location. I'm running a backyard station now that may provide better data for you depending on the distance between us.

 

I was having a lot of problems with other stations too mostly related to temperature because all the surrounding stations were at quite different altitudes.

 

My station id is p17648 It's at about 1300ft in EDH. I've had some issues with rain data but I think I have most of those worked out. The station just came back up on the weatherbug cloud so I should know soon as rain is on the way.

 

And if the station is useful to you, at least you have someone to complain to if it is having problems

[JMC]

Bob,

 

Assuming EDH is Eldorado Hills we should be very similar, I'm in Auburn. I would like to try your station, how can I get my ISY to find it? I haven't tried locating a private station yet.

 

BTW, my email is jmcarron@starstream.net

 

mike

[bpwwer]

Bob,

 

Assuming EDH is Eldorado Hills we should be very similar, I'm in Auburn. I would like to try your station, how can I get my ISY to find it? I haven't tried locating a private station yet.

 

Yes, El Dorado Hills.

 

Before getting my station and getting it to report to weatherbug, I would bounce between Placerville airport and the CTRSH station. Mostly using Placerville. But the temperature difference between Placerville and here would vary between 3 to 10 degrees. Not real useful when used to control when heat or air should be on.

 

You should be able to enter the "p17648" for the station ID and it will pick it up. You can also look at it on-line at http://weather.weatherbug.com/CA/El%20D ... tat=p17648

 

I haven't been tracking the rain stats, but I'm sending the same values I send to weather underground and my rainfall data is matching another wunderground station just down the street. While looking at it today, I noticed that that weatherbug wants monthly and yearly rain totals so I just added that to the data I'm sending them. It hasn't started raining yet, but it's supposed to tonight.

 

Hopefully the updates to the cloud will make things work a bit better. I'm sending updates once per minute and have seen a lot of errors where weatherbug refuses the connection. Nothing consistent, just burst of refused connections 5 to 10 times in an hour and then it's fine for hours.

[JMC]

Good, I'm using data from your station now. I just checked and your reported temp is about 1 degree off from what it is here. I'm not too concerned with temp differences since I removed my ISY control over my HVAC when I replaced the system. We are forecast to have heavy rain tonight so it will be interesting to see what my system report.

 

Later on I may try to pick your brain about the system you have and figure out if it is practical for me to install a similar arrangement. My local airport data is so flakey I don't dare use it for anything.

 

mike

[JMC]

This morning's numbers look good. Since the heavy rain last night was after midnight, the 0.73" Rain Today appears to be right on. Now we'll see what the Water Deficit Yesterday number looks like tomorrow.

 

mike

[bpwwer]

This morning's numbers look good. Since the heavy rain last night was after midnight, the 0.73" Rain Today appears to be right on. Now we'll see what the Water Deficit Yesterday number looks like tomorrow.

 

mike

 

Great!

 

I had to restart everything around midnight after a power glitch from the lightning so I think the daily total never reset. The .73 is likely what we got since about 10pm yesterday.

 

I'm still getting refused connections (43 since midnight), but none since about 5:30 this morning. I don't think this really effects anything since they're somewhat random and I don't see multiple refusals in a row.

 

I don't think I'm sending (or maybe not properly tracking) monthly & yearly rainfall. I'll look into that. I am sending hourly and daily rainfall totals and those seem to be working. Weatherbug doesn't seem to accept rainfall at any smaller granularities.

 

I don't know that my setup is something worth duplicating. When Acurite came out with a sensor package and internet bridge it seemed like a great solution. Works well if you want to use their cloud service, not so well if you want the data local. I do have the setup somewhat documented at http://www.bobshome.net/weather/

[TJF1960]

Bob,

 

Is this the acurite model you have? http://www.acurite.com/all/acurite-professional-weather-center-with-aculink-remote-monitoring-weather-alerts-temperature-humidity-wind-rain-01055.html

It looks like they bundled a bridge with it. If it will work it looks like a great deal. Does the bridge look like yours. How long have you had your system and are you happy with it?

I am thinking of getting one and running your script on a pogoplug.

 

Thanks,

Tim

[bpwwer]

Bob,

 

Is this the acurite model you have? http://www.acurite.com/all/acurite-professional-weather-center-with-aculink-remote-monitoring-weather-alerts-temperature-humidity-wind-rain-01055.html

It looks like they bundled a bridge with it. If it will work it looks like a great deal. Does the bridge look like yours. How long have you had your system and are you happy with it?

I have the costco version of that (01050C), but it's the same components. They only make one bridge and one 5-n-1 sensor. There's a number of different displays. I'm not real happy with that display but then I didn't buy the package for the display so never really used it. On sale, that is a pretty good deal.

 

Initially, they had a big problem with the humidity sensors, I've had 3 of the 5-n-1 sensors before getting one that works correctly. The bridge is supposed to be plug-n-play, but there can be some issues with getting it registered. It's been a while since I installed my but I think I had to wait a while after plugging it in because it spent the first 20+ minutes downloading new firmware. A lot of people seem to have problems getting the bridge to register, I think it's because it won't register while doing the firmware download so if you keep power cycling it and trying to register it fails. The key is to let it sit for while.

 

I've had mine for about 2.5 years now. They first released the bridge based product for sale at CostCo and I bought one. Then they pulled the bridge product for a about a year while they worked out some issues. My original bridge and last 5-n-1 sensor have been good.

 

Given the cost, I don't expect the same performance/quality as a $600 Davis weather station. I do consider it a pretty good value. I do wish they'd make the bridge a bit more hobbyist friendly, but I can work around that.

 

The support on the product has been beyond what I would expect for something at that price point. They seem to be willing to send out replacements when there is a problem. There's quite a few examples of that in their support forum (see link below).

 

I am thinking of getting one and running your script on a pogoplug.

 

I think a pogoplug would probably work. The key is that you need two network interfaces so you can sniff the packets being sent by the bridge. I've been working on the software to make it more modular and less specific to my environment. You should also look at the software George created (see below).

 

There's also a commercial product called MetroBridge (I think) that does basically the same.

 

The two places that have information on decoding data from the bridge are:

 

http://moderntoil.com/?p=794&cpage=1#comment-4918

https://support.acurite.com/acurite?topic-list[settings][type]=recent

 

George, http://nincehelser.com/ipwx, is really the one person responsible for most of the information on decoding the data from the c

[TJF1960]

Thanks very much for the info. Looks like I have some things to think about.

 

Tim

[JMC]

Bob,

 

This morning your system was showing 0.07 inches of rain. Right now it is still showing 0.07 inches even though it has been raining strongly all day. Wasn't it raining in EDH today?

 

mike