I've been gathering data on and off for the past two years. I say on and off because several technical issues have come up over the years that causes my sensors to stop recording data. One summer while I was on vacation the usb device failed and I had to order a new one from Hong Kong. Once I have a little more free time, there will be some information pages on the weather of my backyard. For now, you can read about the instruments I've got running.
The aim of this weather station was to generate accurate weather data with a high temporal frequency on an automated basis. To keep the costs low, real-time reporting was not a requirement. So far, two pieces of equipment have been acquired: a one-wire thermometer system and an electronic barometer. Below the instrument specifications and operating procedure will be explained.
To gather accurate temperature measurements the thermometer should be housed in a shelter away from the influence of buildings and pavement. Ideally, the thermometer should also be shielded from direct sunlight and rain. For information on building considerations, visit globe.gov where you can view this document.
The shelter used here is 1.5 meters tall with a instrument platform approximately 45 centimeters wide.
The thermometer in my backyard is sheltered in a wooden box approximately 45cm wide, 1.5 meters above ground. Two coats of semi-gloss paint reflect direct sunlight. For airflow, all four sides have slats tilted at 45 degrees. The shelter itself is 10 meters away from any buildings or roads. Unfortunately, there isn't an ideal place on our property without tree cover. The picture above shows the surrounding area.
One-wire temperature sensor
A very simple and inexpensive temperature sensor was acquired from RDing Tech (PCsensor.com). This sensor is weather proof but requires Windows-based software to run (so I had to get an old PC and keep it running 24/7). This instrument can save temperature data every second with a maximum operating error of 1.5 degrees Celsius and a resolution of .06 degrees C; the actual error is much smaller depending on the length of the data cable (in this case approximately 15 meters of a possible 300 meters) and the amount of environmental electromagnetic interference (in this case relatively low). Four sensors are currently in operation: one at 1.5 meters above ground (in the shelter), one approximately 2cm below the surface, one 0.5 meters below ground, and one indoors near the computer.
The computer under my desk that runs the temperature recording software.
Each sensor was calibrated in ice water and then compared to other measurement devices at ambient temperatures to assure correct operation. The software was set to collect observations roughly every 8 seconds and then standardized to every 15 seconds.
A very accurate electronic barometer was acquired from Gulf Coast Data Concepts (gcdataconcepts.com). This device is much more reliable than the thermometer used. For example, if the power goes out, this barometer will continue to operate for days on battery power. Because atmospheric pressure differences relevant for weather reporting are negligible over short distances, this device is housed indoors and attached to a permanent power supply. This instrument can save pressure data up to 10 times per second with a maximum operating error of 2.5 hPa and resolution of 0.01 hPa; the actual error will be significantly lower since the operating temperature is maintained near room temperature. Full specifications for this device can be found here.
The Barometric pressure USB data logger from Gulf Coast Data Concepts
This device was configured to log pressure readings every 5 seconds and then standardized to 15 seconds. Pressure recording began 18 March, 2011. For the record, I love these guys and highly recommend their product.