Stepwise variations of strain (hereafter called the strain step) were observed by a couple of borehole strainmeters installed 30 meters apart in a fracture zone of the southern Fossa Magna. Strain changes for a year after the installation carried at December, 1986—January, 1987 were apparent expansion indicating the cooling of grout and the relaxation of stress concentrated around the strainmeter. For two years after the installation the number of the strain steps decreased gradually. No clear relationships are seen between the occurrence of strain steps and the environmental conditions such as the rainfall, the atmospheric pressure, the groundwater temperature, and the groundwater level. The strain steps that occurred in the two boreholes are independent, which means that the spatial extent of the phenomenon is less than the distance between the strainmeters. The amplitude-frequency relationship for the strain steps is analogous to those of fracture processes such as earthquakes and acoustic emission in rocks. The decrease in the number of strain steps with time is represented by a formula similar to Ohmori's for the aftershock sequences of earthquakes. A laboratory experiment showed that the frequency of strain step occurrence depends on the stress levels in the surrounding media. Consequently it is inferred that the strain steps observed by our strainmeters are caused by local fracturing processes around the boreholes, the number of which indicates relaxation of the initial stress induced by the installation of the strainmeters.
The dome emission effect of Eppley pyrgeometers, PIRs, was quantitatively investigated by laboratory experiments with a newly developed calibration apparatus and by field measurements under clear and cloudy conditions. The calibration apparatus contained a blackbody to provide homogeneous thermal radiation at a variable temperature between -25°C and room temperature and also a device to change the dome temperature. The dome coefficients of PIRs were obtained by laboratory experiments with the apparatus. Hysteresis of PIR outputs disappeared successfully after correction of the dome emission effect, and the linear regression for the PIR output versus the radiation input had a straight line giving calibration constants for the PIRs. Errors involved in the laboratory experiment and the field measurement were estimated. When the downward infrared flux was measured with PIRs, ground-based measurements under the clear sky overestimated by more than 20W/m2 unless the dome emission effect was eliminated. By the contrary effect, airborne measurements underestimated by 10-20W/m2. It is shown that PIRs can measure infrared radiative flux with an error around 10W/m2 if the dome temperature is measured in order to correct the dome emission effect.