Laboratory calibration for multidirectional spectroradiometers

Abstract

A method for the calibration of multidirectional spectroradiometers (MUDISs) capable of the simultaneous measurement of spectral radiance at different wavelengths is presented. The calibration of the spectroradiometer is challenging and crucial for high quality measurements of the angular dependence of the radiance. The calibration device consists of an integrating sphere (also known as Ulbrichtkugel), with a diameter of 100 cm, equipped with three 100 W lamps positioned in the lower hemisphere, with the input optics of the MUDIS directed towards the upper hemisphere. The MUDIS detects radiation from 113 different directions simultaneously in a wavelength range from 300 nm to 550 nm. Due to multiple reflections within the sphere, the radiance from the upper hemisphere is nearly homogeneous with deviations of less than 3% on average. Disregarding the 3% variability and assuming a homogeneous radiance inside the upper hemisphere of the integrating sphere, the spectral responsivities of all the MUDIS channels were determined based on the measured zenith radiance, which was detected by a pre-calibrated Network for the Detection of Atmospheric Composition Change reference spectroradiometer containing a scanning double monochromator with a unidirectional input optics. The input optics of the MUDIS contains thin fibers that should not be moved to avoid changes in the instrument’s responsivity. The proposed method is therefore suited to determine the absolute responsivity of the MUDIS for all directions.

A novel process control method for remote calibration based on improved multiscale permutation entropy and dynamic Hotelling’s T2 control chart

“…1) In-control dataset X train : µ = [0, 0], Σ = I 2×2 , number of samples is 1000; 2) On-site dataset X test,1 : add 1% abnormal samples to X train , and the abnormal samples obey the normal distribution µ = [3,3], Σ = I 2×2 ; 3) On-site dataset X test,2 : add 1% abnormal samples to X train , and the abnormal samples obey the normal distribution µ = [5,5], Σ = I 2×2 ; 4) On-site dataset X test,3 : add 5% abnormal samples to X train , and the abnormal samples obey the normal distribution µ = [3, 3], Σ = I 2×2 ; 5) On-site dataset X test,4 : add 5% abnormal samples to X train , and the abnormal samples obey the normal distributionµ = [5,5], Σ = I 2×2 .…”

“…There are two working modes of traditional calibration: (1) the instruments are sent to a superior metrology center for calibration; (2) the standard instruments are brought to the customer site by professional personnel of the superior metrology center for calibration. The above two modes are also called laboratory calibration [3] and field calibration [4], respectively. No matter which mode is adopted, a lot of manpower, material and time resources will be wasted, causing a low calibration efficiency.…”

Absolute radiance calibration in the UV and visible spectral range using atmospheric observations during twilight

“…Usually, measurements of the spectral radiance are calibrated in the laboratory using calibrated light sources (e.g. Pissulla et al, 2009;Yu et al, 2014, Niedzwiedz et al, 2021. Uncertainties of the calibration procedures typically range from 3 to 10 % (Wuttke et al, 2006;Pissulla et al, 2009).…”

Ground-based measurements of the weather-driven sky radiance distribution in the Southern Hemisphere

“…Measuring the radiance distribution requires sampling the downwelling solar radiance across zenith and azimuthal viewing angles. These measurements can be conducted sequentially using sky-scanners [11][12][13][14][15] or they can be conducted at once using non-scanning motionless multidirectional systems [16][17][18]. The latter are fitted with a set of multidirectional radiance input optics able to sample the radiance from several directions at the same time.…”