Broadband UV radiometers are widely used for measuring UV irradiance or radiant exposure in various areas of health,
industry, and science, such as in weather aging of materials, semiconductor lithography, non-destructive testing, UV
curing, catalytic process, environmental monitoring, atmospheric research, water purification, medical diagnostics and
therapy, space-based astrophysical observations and other applications. The accuracy of measurement is influenced by
many factors such as the characteristics of the radiometer, operating conditions, environmental conditions and the UV
sources to be measured. Due to the lack of spectral resolving power, significant measurement errors may occur if the
radiometer has poor quality.
Five different calibration and traceability methods of UV radiometers were realized and analyzed at National Institute of
Metrology (NIM), including absolute spectral irradiance method based on standard lamps, absolute spectral power
responsivity method of detectors, absolute thermoelectric radiometer method, electrically calibrated pyroelectric
radiometer method and method of comparing with the standard radiometer.
The above methods were used to realize the irradiance responsivity of a broadband UVA meter independently.
Compared to absolute spectral irradiance method based on standard lamps, relative deviation of these traceability
methods are 0%, 0.18%, 2.50%, -3.04% and 4.11% respectively. Absolute spectral power responsivity method of
detectors is adopted by the most national metrology institute to realize UV irradiance responsivity. The deviation of
absolute thermoelectric radiometer method results from poor signal-to-noise of the combination of radiation source and a
UV filter, near to the sensitivity limit of the absolute radiometer. The largest deviation of method of comparing with the
standard radiometer comes from different response wavelength, different receiving area between the standard radiometer
and the test radiometer. Spectral irradiance method based on standard lamps is the principal traceability method of NIM
tracing to the national spectral irradiance primary standard directly. This method was proved by international comparison
of "Irradiance Responsivity of UVA Detectors APMP PR-S1". The comparison results showed that the consistency
between the NIMs' value and the international reference one is the best among the seven participants.
On the basis of the above theories and methods, the uncertainty of broadband ultraviolet irradiance standard under the
defined conditions was cut down to 2.0% (k=1) from 10% (k=1).