Notice on the Accuracy of the Lake Biwa Museum's Meteorological Data

Announcement concerning missing or invalid meteorological data observed on the ground of the Lake Biwa Museum.

There was a loose connection on the data path from the sensors of air temperature, ground temperature, humidity, and air pressure, and the recorded values have not been valid since about September 5, 2011. The records for wind velocity and rainfall are still valid because their data paths are independent. The connection was repaired at about noon on September 22.
There was an insulation failure on the data path from the ground temperature sensor, and the recorded values for ground temperature itself as well as air temperature, humidity, and air pressure have not been valid since about January 9, 2011. The records for wind velocity and rainfall are still valid because their data paths are independent from that for ground temperature. The insulator was repaired at about noon on January 25.
The rainfall sensor became clogged with leaves and the heavy rainfall on September 7 and 8, 2010 was not recorded as it fell. Because there seems to have been no overflow of rainwater from the sensor, the total rainfall for these dates should be valid, albeit recorded all at once when the clog was cleared (about 08:00, September 8).
The rainfall sensor became clogged with leaves and the heavy rainfall on November 10 and 11, 2009, was not recorded as it fell. Because there seems to have been no overflow of rainwater from the sensor, the total rainfall for these dates should be valid, albeit recorded all at once when the clog was cleared (about 15:00, November 11).
The path from the observing system to the recording system was damaged in the morning of July 10, 2004, and the data after then were lost. The damage may caused by a lightning strick near the museum (not ensured). We repaired the interface on the recording system in the morning of July 15, and the recording restarts.
The observing system's clock is designed to be self-correcting, but it did not work correctly and the clock was discovered to be about 27 minutes behind on October 15, 2002. The problem was once resolved on August 28, 2001, and the recoreded time was accurate then. It is uncertain when the clock became unaccurate. The observation stopped because of the maintainance of the museum's power supply system on October 15, 2002, and the data before the maintainance have delayed time information, but the time records are correct after that. There was no rainfall in the duration of the maintainance.
In the early morning on October 23, 2001, an electric leakage in the air conditioning system damaged various instruments in the museum. The data converter of the meteorological observation system was also damaged severely and the data recording stopped through 15:00 on November 30.
From 7:13 through 16:05 on October 22, 2001, the observation stopped because of the maintainance of the museum's power supply system. In that duration, there was rainfall that is a little less than 5 mm per hour, which is not included in the daily-integrated record.
The observing system's clock is designed to be self-correcting, but it did not work correctly from May of 1999 through August of 2001. So, the recorded times of observations are not accurate for this period. The clock was discovered to be about 10 minutes behind on August 28, 2001, so, it appears to have lost about 1 minute every three months. provided that the trend was monotonical. Another trouble occured when restarting the correction system, and the observations stopped for about one day. Therefore, all the data for August 28 have delayed time information, but starting from the next day (August 29), the time records are correct.
Damage to the rainfall sensor was detected on September 12, 2000, when apparently mistaken data for September 10 to 12 were noticed. The damage was repaired on September 13, and the data in that day include both a part of the past rainfall and artificial rainfall used to test the system.
The data recorded before the damage by lightning strick in July seem to be acuurate because they coincide well with those of Japan Meteorological Agency's AMeDAS. After the repair in August through September 9, there was almost no rainfall in fact, but some light rain may have been dropped from the record.
On July 2, 2000, the data converter was damaged by a lightning strick near the museum, and it took more than one month to repair. Therfore, records from that day through August 23 are absent.
Snowfall on the night of January 20, 2000 was recorded on the 10 o'clock data of January 21 because the heater of the rainfall sensor was switched off. For the same reason, all the snowfall records from late 1999 may have been delayed.
On August 16, 1999, the data converter was damaged by a lightning strick near the museum, and it took more than one month to repair. Therfore, records from that day through September 29 are absent.

Back to the Meteorological Data top page