Caveats, details, excuses and information:
The equipment: An Oregon Scientific WM-918 Weather Station feeds data via serial cable to a Windows 2000 server running “Virtual Weather Station” 9.07 by Ambient, LLC. The data is uploaded to the Internet via modem at a blistering 28.8K maximum; no DSL is available here on the San Juan Ridge. Most of the images are updated every 15 minutes. The Temperature/Humidity and Barometer/Rainfall graphs are updated every half hour, and the Temperature pie chart is updated each hour. Data uploads to Weather Underground , Northern California Weather and Weather for You occur every 15 minutes.
The location: Near the intersection of Tyler-Foote Crossing road and Murphy road on the San Juan Ridge in Nevada City, CA at an altitude of 2050 feet. This is roughly 500 feet lower than, and 6 linear miles northeast of, the National Hotel on Broad Street in downtown Nevada City.
Temperature/Humidity data: While I strive to make the data posted as accurate as possible there are various glitches, the most notable of which is occasional hanging of the outside temperature/humidity sensor, causing the dreaded “flatline” shown below. This is a well known problem with the WM-918, which I have investigated and deemed unsolvable, short of purchasing a completely different weather station. As of 7/30/03 this sensor is located in a shaded location under the eave on the north side of the house and should provide good accuracy.
Rainfall data: The rain collector is located about 16 feet above ground level at one end of my roof peak. It contains a bi-stable plastic element consisting of two tiny “buckets” with a vertical vane between them, mounted in a pivot arrangement such that the vane ducts water from the collection funnel above into one of the buckets until the weight of collected water is sufficient to cause the element to flip to the opposite position, emptying the first bucket and allowing the vane to duct water into the other bucket, and so on. This flipping action triggers a signal which informs the base unit that .04” of rain has fallen. This arrangement creates several issues: If it is raining very lightly, that rainfall will not be accurately portrayed over time due to the .04” granularity. The hourly rainfall graph, which displays the rate of rainfall over the preceding hour, appears as a series of .04” steps, rather than a smooth curve. If it is snowing, the snow will pile up in the (unheated) collection funnel, melt at some later time and show precipitation when it might be clear and sunny outside.
Wind data: The wind vane and anemometer are located about 20 feet above ground level or 3 feet above my roof peak, at one end. While the house is in a clearing, it is surrounded by trees, including a number of 80’ to 100’ Ponderosa pines. When the wind is blowing, it is clear from the motion of the trees that its speed is significantly higher at treetop level than it is at the anemometer. The trees also produce eddy currents, causing the wind direction data to be more variable than it probably is in the higher air.