To estimate the quality of long-term measurements of irradiance sensors, values were recorded over a period of 8 days with changing weather and sky conditions. The calculated sums of irradiation were compared to those of the CM21. During the period or recording, the CM21 measured an irradiance sum of an average 3,39 kWh per m2 and day, a quantity that is well representing the daily average of irradiation during a whole year in central Europe.

The observed relative deviation of all sensors did not exceed 10 % and for 6 sensors it was even less than 4 %. Table 4 shows the results of the analysis. The lower sums of irradiation of the sensors 9 and 10 are due to a shorter period of recording.

Off-Set

During the night-time the measured values of the devices were not absolutely zero. When this factor is ignored in a counter and the values are straight away added up, depending on the value of this off-set, the long-term measurements of irradiation might be more or less falsified. To quantify the influence of the off-set on the annual irradiance sum, the recorded values when G(device) < 0 W/m2 during the 8-days period were accumulated and extrapolated to one full year.

The results of the analysis for all sensors can be seen in Table 5. Surprisingly, the reference-pyranometer had the second highest overnight off-set of all tested devices with an annual sum of -9,8 kWh/a. For one sensor the calculation resulted in a sum of -24 kWh/a (a quantity in the range of 2 % of the annually irradiation in central Europe), while 7 sensors showed practically no off-set with an extrapolated sum of less than -1 kWh/a. To on the other hand rule out a significant positive off-set, the values of all sensors when G(CM21) < 0 W/tF were added and extrapolated to a whole year. The difference to the projected sums of the negative off-set were less than 1 KWh/a for all sensors.