Weather at Mouser's HouseI am in the middle of coding a custom interface for this project, as I found the existing shareware offerings to be unacceptable. It'll be some time before it is fully armed and operational, but in the meantime here are some plots: This is a combination of several temperature sensors and some derived values. The "Anemometer Temp" is located inside the non-vented anemometer enclosure on my roof. It shows unrealisticly high temperatures when the sun is shining on the enclosure. The "Inside Temp" sensor is located inside my house, but in a room that is poorly ventillated and is generally a little warmer tha the rest of the house during the summer and a little cooler during the winter. The "Outside Temp" sensor is in a vented pagoda structure that will eventually find its way onto my roof, but for now is sitting next to the Inside Temp sensor. The difference between these two sensors I have not yet accounted for.  The rainfall gauge is a tipping bucket type. One bucket tip is equal to 0.1mm of rain accumulation and the sensor is polled once per minute. Therefore, the minimum rain accumulation that can be measured is 0.1mm in a one minute integration window. The rain rate is therefore always some integer multiple of 6mm/hr.  The anemometer counter is sampled at about 0.5Hz and an effective windspeed is calculated for each sample based on the number of accumulated counts and the time of the sample interval (accurate to microseconds). These rates are averaged over one minute to produce the Avg line, above. The Gust value is the highest single rate estimate made during the one minute integration time. The anemometer counter increments once every half-revolution of the cup assembly.  Thanks to Brian Fiedler for the use of his vplot and windrose plotting scripts, this is now much less ugly. I still need to customize the script some, but the data here is real data for the wind direction and speed at my house for the last day. Up is North, down is South, right is East, etc. The radius of a sector indicates how often the wind was blowing in that direction. The area of a given color sector is related to how often the wind was blowing in that direction at the speed indicated by the color. Similarly, the area of the central circle is proportional to the time for which the wind speed was zero. Right now all of the plots are just for the past 24 hours and are updated once a minute. The computer is not configured to restart the weather daemon automagically yet, so there may be data outages whenever we lose power at the house (which, during monsoon season here in Los Alamos, which is now, happens a lot more than it should). There are additional sensors and hardware en route which will be added to the weather station as soon as I get back from Nashville. In case you're wondering, the sensors here are all using the Dallas Semiconductor 1Wire protocol. The sensors are located on the roof of my house, attached to the mast of my ham radio antenna. A homebrew perl script running on a linux box harvests data from the sensors via the One Wire File System and FUSE kernel module. The data are placed in a MySQL database on the same box. Another script periodically grabs data from the database and makes calls to gnuplot to generate the charts seen here. This'll get more customizable and user-friendly when I have more time. :) -mouser, 2006.07.11 |