January: This was another cold month with little if any precipitation falling as rain. For the first part of the month there was a layer of about 9 inches of snow and most of the "rainfall" recorded was actually this snow melting.
April: For some reason, at 03:05 on the 6th of this month the atmospheric pressure has jumped up by 29.1hPa. At the same time, the wind speed dropped to zero from around 13mph with gusts to about 24mph and stayed at zero for over six hours. Judging by other data that I have seen on the internet, the wind does appear to have dropped quite suddenly possibly corresponding to the recorded data. While the pressure did rise steeply around this time, it didn't suddenly jump up, nor would I expect it too, but the levels either side of this anomaly are consistent with other recorded data. Conclusion must be that a glitch in the measurements led to the sudden rise in the pressure records while the wind speed data may in fact be correct.
Further, there have been other problems with the data storage and almost five days worth of data has not been saved in one big chunk from the 10th to the 14th and several smaller ones on the 15th, 17th, 21st and 26th.... all in all a complete shambles!
I suspect the problem has been with the PC downloads from the weather station in spite of it having been working sufficiently for the backup program to send an email on completion of it's tasks.
In order to get a graph showing the data I do have at the correct relative place on the graph I have filled in a lot of the blanks, mostly be interpolating a straight line between where the real data ended and began again. Winds have been set to zero with a nominal direction of North applied to them. I have also modified the web page to add a link to the original graphs.
October: A slight glitch occurred in the data for this month. On the 2nd at 15:44 the wind speed was recorded as 816 miles per hour! Surprisingly the matching gust was only 286.5 mph. There was also 8.2mm of rain recorded, quite a lot for one minute!
I have set the wind speeds and the rainfall to zero since at least 40 samples before and after this time were zero. I also adjusted the wind chill to match the surrounding values.
November: At the end of this month, I found that I was missing some data. The problem turned out to be that the data file had become too big and was taking so long to load up that the backup program was shutting the system down before enough time had passed to allow all the data to be downloaded from the weather station.
December: A cabling problem this month caused me to lose about eighteen days worth of data in two blocks. The first block was from the eighth until the twentieth and again from the 25th until the thirty-first. This happened during an unusually cold December which makes the data loss especially disappointing.
An improved method of verifying the data collection is really needed to help find out when there are problems in future. At present, all that is supposed to happen is that the PC storing the data runs a backup (Using Cobian 7) which sends an email on completion of the backup. There is, however, no check made that any new data was backed up. A job for a DOS batch file perhaps?
I am writing this in May 2011, having let the updates get very behind because of the amount of time I knew it was going to take to sort out the mess that the data was in.