In 2004 I purchased a Davis Weather Wizard III weather station and installed it on the roof of the house in Snowville, Virginia. The Wizard III monitors outside temperature, wind speed and direction, and rainfall; however, it doesn’t collect dewpoint or pressure readings (that version now costs over $600). Over the next several years it provided excellent observations and ran without a hiccup until 2009 when I then decided to take the station offline until I moved to go to graduate school. In May of 2011 I uninstalled the station to bring back to Indiana.
The anemometer which provides wind speed and direction measurements was inoperable, so instead of investing in a new station, I contacted Davis Instruments and they were kind enough to send a replacement. If this was our permanent residence, I would have upgraded to a newer station with dewpoint and pressure readings, but the current setup would have to do for now. The issue was where and how to install the station and not violate the rental agreement, which explicitly forbid any antenna installations. Not to mention that I’d enjoy installing several antennas to expand my amateur radio hobby, but that will have to wait. As for the weather station, I decided to keep the installation footprint as small as possible, so the height was not to exceed 10 feet and the station would be installed in a relatively isolated location which would not be in the way of any maintenance work.
I started with a 10′ section of 3/4″ conduit from Lowe’s. I also purchased three 3/4″ T-connections to attach to the main, central beam. These would be used to add smaller sections of conduit to attach the rain gauge and allow for wiring. Additionally I needed two 3/4″ conduit caps to close off the end of the exposed pipes, so I used 3/4″ plumbing pipe end caps. All of this would be painted white to reflect radiation from the sun.
Capturing only weather information was just not enough, so I decided to add a webcam to the station. I had originally installed a Logitech Webcam Pro 9000 in the extra bedroom at the apartment, but glare from interior lighting and my computers always seemed to interfere with the images. I took a tip from David Savery and Cowboy Frank, both of which have installed webcams outside. Cowboy Frank actually had the same Logitech webcam, and come to find out, had the same light fixture that I purchased at Lowe’s.
The light fixture is the kind that you would install outdoors for security or home safety purposes. Since they are made for outdoors, they are pretty much waterproof as long as you install them properly. I removed the halogen lightbulb and all the internal wiring and grounding parts, then had to figure out a way to run the USB cable through the fixture.
Cowboy Frank originally cut his USB cable to feed it through the fixture housing, but I found a much easier solution from Gary Honis. Gary’s page shows step-by-step how to disassemble the Webcam Pro and remove the wiring without cutting the USB cable, but you need a soldering gun. I also had to consider how to connect the webcam to my computer, as the standard USB 2.0 only allows for a 15′ length of cable before degradation. I had two options, either extend the USB connection over Cat5 Ethernet or buy a dedicated USB booster extension cable. The former option is cheapest but the later allows for USB 2.0 speeds. I decided to buy both and try them out to see which would work best. The Cat5e option did work, but the image was grainy due to the limited USB 1.1 transfer speeds over 50′ of cable. The booster cable worked best and came in a 39′ length of cable, which was perfect.
After figuring out the webcam situation, I had to purchase a temperature sensor shelter to minimize the effects of direct radiation on the probe. I also had to purchase a shelf to hold the tipping bucket rain gauge. Both these items were bought direct from Davis Instruments. After some ingenuity, I was able to attach both to the main beam of the station, drilled wiring holes, and attached the pipe end caps with JB Weld. Next, I installed the webcam fixture, webcam, temperature sensor, rain bucket, and anemometer. Each of these had cables that had to be ran through the main support beam and out the very bottom T-bracket. In total, three instrument cables and one USB cable had to be ran from the main support beam, under the roof lip of our apartment shed, and then along the apartment wall to the upstairs window where the weather server machine is located.
After everything was mounted to the main support beam, the weather station was mounted, leveled, and attached to the wooden privacy wall behind our apartment. Leveling the station is extremely important due to the tipping bucket rain gauge. If the gauge isn’t level, it takes more or less rain to cause the internal bucket to tip, resulting in inaccurate rain measurements. Once the station was leveled, 3/4″ conduit brackets were used to attach the station firmly to the fence. A basic grounding wire was attached to the main beam to hopefully protect it in the event of a direct lightning strike. I positioned the webcam and inserted some silicon desiccant packets to absorb any moisture should it accumulate inside the webcam housing. I also attached a bird deterrent of spikes to the top of the tipping bucket to keep birds from landing and nesting on the gauge.
The location is not the most ideal site for a weather station. The current recognized WMO standards for installation include an anemometer height of 10 m, rain gauge height of 1 m, and temperature height between 1.0 and 2.0 meters. Only my temperature sensor meets these requirements. Additionally, it’s preferable to have some sort of fan-aspirated radiation shield to fully minimize the effects of radiation; my station only has the basic radiation shield as Davis Instruments no longer manufactures the fan-aspirated model for my station. The full details can be found here and through the CWOP here. The weather station held up through some very strong thunderstorms in July 2011 that brought torrential rain and high winds; however, due to the station’s location the station show definite signs of under-reporting of winds and rainfall. This is ONLY a problem when winds and rain are from a westerly (SW, W, or NW) direction as the station has reported fairly consistent measurements as that of the AWOS at the Muncie regional airport.
The weather station and webcam became operational on 2 July 2011 and soon after data was streaming online to several websites, including the Weather Underground. Below are some pictures of the weather station, links to the brand new WX4SNO weather station webpage, and to some commercial websites that collect my station’s data.