.
Main / About My Weather Station

My Weather Station

Area Map with My Station at 41.1338N, 85.0307W
Fort Wayne, Indiana
Station located at 41.1338N, 85.0307W, elevation 792' (241m) ASL
Instruments

In August 2005 I installed a Davis Vantage Pro2™ weather station at our home in Fort Wayne, Indiana. In December 2005, I added the WeatherLink® datalogger and computer software. Thus, the AWC Weather Station was born and online!

The outdoor sensors consist of the basic ISS (Integrated Sensor Suite): a tipping bucket rain gauge, solar-shielded temperature and humidity sensors, and an anemometer and directional vane. The indoor console has temperature and humidity sensors and a barometer. The ISS was connected to the console by a long cable fished through a window. The ISS was mounted on a 4"x4" mailbox post at the northwest corner of our deck. The main part of the ISS (with the rain bucket) was mounted on the northerly side of the post (facing away from the deck), and the anemometer/wind vane were mounted on the southerly side. The CoCoRaHS rain gauge (when I started doing that) was mounted on the easterly side of the post, and a decorative birdhouse was mounted on the westerly-pointing crossbeam.

In May 2016, I began noticing occasional periods (sometimes hours) of data dropout from the ISS. The, at 00:05 on 21 June 2016, the last data arrived from the ISS. So, I ordered a new Vantage Pro2; but this time got a wireless one because I wanted to relocate it, since a nearby tree was growing tall enough to block some rain when the wind had a northerly component. The new ISS was installed in a temporary location off the northeast corner of the deck, new computer software was installed (see below for details), and we were back in business on 05 July 2016.

On 29 July 2016, I moved the ISS to its permanent location. It was mounted on a new (taller) 4"x4" mailbox post. This time the post was installed so that the sides of the post faced cardinal directions. The main part of the ISS (rain bucket, temperature/humidity sensors, transmitter) is mounted about 6' 1" (1.85m) above ground level on the south-facing side with the middle of the solar-shield (so the solar panels on the transmitter are facing the sun). The anemometer and wind vane are mounted on the north-facing side, with the bottom of the anemometer cups about 8' 9" (2.67m) above ground level. My CoCoRaHS rain gauge is mounted on the east-facing side, and a decorative birdhouse is mounted on the west-pointing crossbeam.

After a month, I built a two-tier planting box around the mounting post, primarily to stabilize it as it had a tendency to sway in stronger winds. The top tier is filled to 1' above ground level, making the main ISS 5' 1" (1.55m) above the dirt, and the anemometer 7' 9" (2.36m) above it. The corners of the top tier point in cardinal directions. In the photo at right, the east-pointing corner is closest.

Looking west
Looking west toward mounted station
ISS encased in ice

Ice

The anemometer can become encased in ice during freezing rain events and, thus, cease to record observations of wind speed or direction (until I can break the ice loose or it thaws). Also, snow can become caked in the anemometer cups and affect wind speed readings.


Location

The current location (since 29 July 2016) of the ISS is the best available on my property. It is in the largest clearing on our property, positioned about halfway between our house and the neighbor's house, and at the point along the line between the two backyard trees where the elevation angles from the mounting post to the treetops are equal (so slightly closer to the shorter, westerly, tree). The elevation angles from the mounting post to the roof lines of our house and the neighbors is about 40°; the elevation angle to the treetops is closer to 50°.

Wind

Our house blocks wind from the southeast to the west-southwest. The neighbor's house blocks wind from the north to northeast. So, the wind vane almost always points in an easterly or westerly direction.

In addition to the houses, there are trees to the west, northwest, and east. Further to the west and northwest are other houses and more trees. Since only two trees are evergreen, the trees have less of an effect on winds speed in winter. Based on comparisons with an accurate hand-held anemometer along the front sidewalk (by the street) and official reporting stations in the county, I apply a scale factor of 1.6 to the Davis Vantage Pro2 wind speed to approximate what the "open field" wind speed would be.

Original Location

The original location (up to June 2016) was limited by the cable connecting the outdoor ISS to the indoor console. The mounting post was located about 2' from the northwest corner of the deck. So, it was much closer to our house and way too close to a tree. In fact, at least on two occasions over the years, I had to trim out a branch reaching nearly over the top of the rain bucket.

© 2016 Google Earth
WX1 marks the original location, WX2 marks the current location

CoCoRaHS

CoCoRaHS logo
IN-AL-39
In January 2010, I became part of the Community Collaborative Rain, Hail, and Snow (CoCoRaHS) Network and obtained the officially approved manually operated 4" rain gauge. It has always been attached to the same mounting post as the Davis Vantage Pro2, and both times on the easterly side of the post. The complete gauge sticks about 3" above the top of the green post (slightly above the top of the rain bucket). My CoCoRaHS station id is IN-AL-39 (i.e. the 39th CoCoRaHS rain gauge installed in Allen County, Indiana). In addition to rain, I also report snowfall, snow depth, and hail information to CoCoRaHS; all of which is archived on the CoCoRaHS website. Simply click on the View Data link near the top center of the webpage, pick the data set, and enter my station id to browse through all my reports.

Accuracy of Vantage Pro2 rain gauge

In use through June 2016, when compared to the co-located CoCoRaHS rain gauge, my original Vantage Pro2 rain gauge fairly consistently under-reported liquid precipitation by 10-15%, when water was able to flow freely. The discrepancy seemed to be largest during heavier rain events, so I believe there was some water loss as the tipping mechanism couldn't keep up. Additionally, in use until August 2014, the original rain bucket (which funnels rain into the tipping mechanism) would at times become clogged by leaves or bird droppings so that only a small amount of water would seep through into the tipping mechanism; and occasionally pine needles would lodge in the bottom hole in such a way as to redirect the stream of water through the bottom hole away from the tipping mechanism inside and cause the Vantage Pro2 to record zero precipitation when the CoCoRaHS rain gauge caught 0.25 or more inches.

Davis came out with an improved rain bucket in 2014 which has a leaf guard within to mitigate clogs/redirects and bird spikes to reduce droppings. Only on a couple of occasions since I installed that (in August 2014) did a narrow small leaf (from the neighbor's twisted willow tree) make it thru the leaf guard and block the bottom hole. I never again had bird droppings within the bucket.

With the current Vantage Pro2 (installed in July 2016), for the first couple of months the discrepancy appeared to be generally less, and was frequently spot on or even a 0.01" over the CoCoRaHS rain gauge catch. However, over the winter of 2016-2017, I have noticed occasions where the new Vantage Pro2 under-reports by as much as 10%.

There was a potential for some rain shadowing at the original location (i.e. prior to 29 July 2016); but as the CoCoRaHS gauge and Vantage Pro2 instruments were mounted on the same post, it would've affected both equally. The current location is as far from trees and buildings as possible on my property — the elevation angle to the two trees on either side (E/W) is closer to 50° than the ideal max of 45°, but the tops of the houses on either side (N/S) are about 40°.

None of the rain buckets have been heated, so precipitation amounts are unreliable during winter months. The first couple of years I had the Vantage Pro2, I even covered the rain bucket to keep everything out; but then decided some winter information was better than none and have left it uncovered since then. Still, as long as the temperature is below freezing, no precipitation amount will be recorded. Then, when it does warm up, the precipitation amount for that day will include the melted equivalent of whatever fell (and remained) in the bucket since the last time it was above freezing. Less than 4 inches of snow will fit in the rain bucket; and much of that often blows out or sublimates before it melts. Thus, even monthly precipitation amounts reported when frozen precipitation has occurred will usually be less than the melted equivalent of the actual precipitation for that month.

Bottom line, you will always get a more accurate picture of the precipitation that falls at my house by reviewing the CoCoRaHS data set.

CoCoRaHS Water Year Summary

These reports, generated by CoCoRaHS, provide summary data as well as every individual observation I reported during their Water Year (01 October through 30 September).

National Centers for Environmental Information

NOAA's National Centers for Environmental Information also maintains a history of the daily precipitation data I've reported through CoCoRaHS: IN-AL-39 Station Details. Unfortunately, they do not capture the multiday reports in any way.


Availability of Data

Data provided on wx.awcolley.com

  • Current Conditions, data on this page is updated every five minutes (except when my computer or my internet connection is down)
  • Monthly Summary, data on this page is updated shortly after midnight daily
  • Annual Summary, data on this page is updated shortly after midnight daily
  • Data Archive, access to every data record (i.e. every 5 minutes) from 07 Dec 2005 thru mid-June 2016

In addition to this website, data from my weather station is uploaded for personal and research use to two other websites.

CW4824

Citizen Weather Observer Program

The Citizen Weather Observer Program (CWOP) is a private-public partnership with three main goals: 1) to collect weather data contributed by citizens; 2) to make these data available for weather services and homeland security; and 3) to provide feedback to the data contributors so that they have the tools to check and improve their data quality. The CWOP collects data from over 4500 nongovernmental weather stations and transmits it to the Meteorological Assimilation Data Ingest System (MADIS) that NOAA provides for the purpose of improving weather forecasting, by providing support for data assimilation, numerical weather prediction, and other hydrometeorological applications.

My CWOP call sign is CW4824; current and historical data can be viewed at

findU is a database archiving weather, position, telemetry, and message data. The primary source of data is an amateur radio system called APRS, some weather data comes from an internet based system called the Citizen Weather Observer Program. This large (>50 GB) database is constantly updated (about 20 new reports come in every second), and is accessed via a number of dynamic web pages.

An analysis of the accuracy and reliability of my weather station (based on other MADIS data in the area) is provided at weather.gladstonefamily.net.

KINFORTW9

Weather Underground

My data is also part of the Personal Weather Station network (currently including over 37000 personal stations worldwide) maintained by the Weather Underground. My Weather Underground station ID is KINFORTW9; current and historical data can be viewed at



Data Recording and Analysis Software

Since July 2016

Since development had stopped on wview in 2014 (and bits of the website, such as images, began disappearing), I decided that getting a new Vantage Pro2 was a good time to look for new software. When I found weewx and read that it used the same database setup as wview so that transferring my nearly 11 years of data (at 5 minute intervals!) would be as simple as copying a file, I decided to give it a shot.

So far, it has worked quite well on my Shuttle (see next section), and I think the graphs it produces are even nicer than wview's. It does not yet provide the daily archive report (listing the data at 5 minute intervals), but that is expected in the next release (hopefully sometime in March 2017).

April 2009 - June 2016

In April 2009, I transferred my Vantage Pro2 serial connection to a Linux (Fedora Core 10) computer running wview. That computer was 9 years old and suffered a hard disk crash in December 2009 (thankfully I back up my weather data nightly, and the console's data logger holds several days worth of observations, so no data was lost). I then began using wview in an Ubuntu virtual machine (using VMware Player) on my 64-bit Vista workstation, on which I first had to install a serial port card.

In November 2010, I purchased an old HP D510 Desktop P4 2.0 with 1GB RAM, an 80GB hard drive, and a serial port, from www.yugster.com for less than $120 shipped (without a monitor, which I already had). I installed the latest Ubuntu and wview, and it worked steadily, only going offline due to power outages. In 2013, I purchased a UPS (uninterruptible power supply) which resolved outages due to the not infrequent <5 second outages we get here; and longer outages rarely occur more than once a year.

In September 2015, I purchased a Shuttle XS36V4 ultra-slim fanless computer equipped with two serial ports (bought at newegg). I installed Xubuntu and wview, and migrated my archived data from the old HP. Not only does the Shuttle take up a fraction of the space of the old HP, it is noiseless and adds very little heat to the room (quite the opposite of the old HP in both respects).

The wview graphics are much nicer and are created much more quickly than WeatherLink’s. In fact, even though the original wview computer hardware was identical to the original computer running Windows and WeatherLink, all aspects of wview were quicker than WeatherLink. Also, wview natively handles uploading to CWOP and WeatherUnderground (i.e. without a third-party plugin). Even within a virtual machine, the performance was still much better and the load on the host computer much less than with WeatherLink.

December 2005 - April 2009

As mentioned earlier, I originally used the data recording, analysis, and uploading software that came with the Davis WeatherLink® datalogger.

The WeatherLink software had a tendency to hang or crash every so often. I have the serial port version of the datalogger at home and (in a previous job) the USB version at work. The serial port version worked much more reliably than the USB version; but I still wanted something that would automatically restart the WeatherLink software when it crashed. One of the most promising ones I found online is StartWatch, a shareware program that lets you coordinate the startup of your applications and monitor them after they're started.

At that same website (SoftWx.com), I noticed a program called VirtualVP, a shareware program that lets you connect up to 4 weather programs to a single Davis Vantage Pro 1 or 2 weather station (console or Envoy).

It turns out that using VirtualVP seems to alleviate the extra problems with the USB version of the datalogger. Couple that with StartWatch, and my weather uploading from work went flawlessly for months. However, shortly after I left that company, the data uploads ceased; and since I had no access to their computers, I could not determine the problem nor fix it.

It also improved the uptime from home; however, it seems my router, cable modem, and/or Comcast (my old ISP) were not quite as reliable and occasionally put WeatherLink into such a funk that not even StartWatch could get it to terminate cleanly so that StartWatch could restart it. Still, StartWatch and VirtualVP provided great value for the low registration fee.

Also from SoftWx.com is a freeware program called VPLive that connects to a Davis Vantage Pro 1 or 2 Weather Station console (either directly or through VirtualVP) and displays the live data. It also calculates the station pressure (i.e. actual pressure), altimeter pressure, and running averages needed to properly generate and send APRS/CWOP data. So, I used that (instead of WeatherLink) to upload my APRS/CWOP data. I'd have liked VPLive to also upload to WeatherUnderground; I once submitted a request for that to VPLive's author, but it was still on his to-do list a couple of years later.

CoCoRaHS
IN-AL-39
A2 Web Hosting
Creative Commons License
GoDaddy
Look for:

loaded 2017-05-26 05:10:33 • last modified 2017-03-28 03:43:01
Privacy PolicyDisclaimer
• wx.awcolley.com is powered by PmWiki v.2002016 •
• all content (unless otherwise noted) © 2017 A W Colley