Weather observations—temperature, humidity, air pressure, wind direction and speed, and amount of rain (but not snow)—have been collected at the observatory location since 2005. This page contains notes on local Open2300 software changes. For actual weather at the Observatory, see the link under WEATHER in the side bar.
Open2300 would not initially open a serial port. The original code uses O_NO_CTTY and this causes the open to fail. Replacing this with O_NONBLOCK and adding code to exit if the port was locked got the port open and communicating with the weather station. This is very similar to a patch on Lavrsen’s Open2300 wiki site except that the serial patch on that site adds O_NONBLOCK but does not remove the O_NO_CTTY. This modified patch has not been communicated to Lavrsen yet.
Additional changes (dated 2006 April 24) to retry the weather station poll up to 5 times if the results are out of whack (>50??C, >100%) and to not report to Weather Underground when the final results are still out of whack.
Removed http://www.wunderground.com from URL build to fix problem with updates on wunderground. Added new pass file to Open2300.conf to handle web page. This file is open for less than a second, so downtime while the file is updated is minimal.
The offset in barometric pressure from Absolute to Relative was changed to 2.66 kPa at 1.20 am EST on 24 January 2006. Previously it was 1.84 kPa. The change brought the Relative Pressure into line with that of YYZ (Toronto International Airport) for the same time.
Upcoming: currently in late afternoon the outdoor temperature sensor is in full sunlight and is being heated above ambient. This behaviour starts sometime after March 21 when the northwest-facing wall becomes illuminated in late afternoon. Changes will be made but the details are not fixed at this point – some form of shielding will be emplaced, but the sensor may also be moved to a less visible location. Any readings taken after approximately 5 pm should be discounted – so far, it is obvious from the graphical display when the readings are elevated. Rainfall and wind measurements are not affected (but when it’s raining, the sensor is not being heated, so all readings are OK). This section will be updated when the site change is complete.
As an experiment in late fall 2007 I tried covering the sensor in aluminum foil. Temperatures stabilised and became more consistent, but I expect the foil to suffer damage during the winter. As of this writing [early spring 2008] things seem to have held up quite well, but the foil is no longer covering the sensor completely, and some problems are already apparent. The foil will be refurbished and the experiment will continue.
November 2006 – Added a webcamera view from an upstairs window so that the observatory is visible (though normally it’s too dark at night).