Evaluation of Variation in Surface Solar Irradiance and Clustering of Observation Stations in Japan

Abstract

Variation in surface solar irradiance is investigated using ground-based observation data. The solar irradiance analyzed in this paper is scaled by the solar irradiance at the top of the atmosphere and is thus dimensionless. Three metrics are used to evaluate the variation in solar irradiance: the mean, standard deviation, and sample entropy. Sample entropy is a value representing the complexity of time series data, but it is not often used for investigation of solar irradiance. In analyses of solar irradiance, sample entropy represents the manner of its fluctuation; large sample entropy corresponds to rapid fluctuation and a high ramp rate, and small sample entropy suggests weak or slow fluctuations. The three metrics are used to cluster 47 ground-based observation stations in Japan into groups with similar features of variation in surface solar irradiance. This new approach clarifies regional features of variation in solar irradiance. The results of this study can be applied to renewable-energy engineering.

Spatial and Temporal Non-Linear Dynamics Analysis and Predictability of Solar Radiation Time Series for La Reunion Island (France)

“…As usual, complexity in a solar irradiation time series stands for randomness or disorder. In order to identify the underlying behaviour of these two concepts, there are various methods to study the behaviour of solar irradiation time series [ 16 , 23 , 26 ]. In the framework of Kolmogorov complexity, we have focused on the need to delineate the predictability of solar irradiation.…”

An uncertainty-based protocol for the setup and measurement of soot–black carbon emissions from gas flares using sky-LOSA

“…The prior probability distributions employed within the MC method inherently link light absorption measurements and computed mass column density and emissions using sky-LOSA; as such, in keeping with Andreae and Gelencsér (2006) and Petzold et al (2013), sky-LOSA-inferred soot-BC mass might therefore be called "equivalent BC" as is recommended for all light-absorption-based diagnostics, especially where absorption-enhancing non-BC material may be present. For the case of flare-generated soot-BC however, studies in the field (Schwarz et al, 2015;Weyant et al, 2016) have identified that the presence of non-BC aerosols in flare plumes is "not statistically different from zero" (Weyant et al, 2016), which is supported by laboratory observations (e.g., Kazemimanesh et al, 2019). This justifies use of soot property probability distributions derived from literature data of freshly emitted soot particulate by Johnson et al (2013).…”

“…For CIE sky models with an unobstructed sun (types 7-15), ξ ED (a) ∼ U (0.75, 1.25), and for models with an obstructed sun (types 1-6), ξ ED (a) ∼ U (0, 1.25). These prior distributions of ξ ED (a) were based on observations by Watanabe et al (2016), who studied the "clearness index" (ground-level horizontal normalized by extraterrestrial horizontal irradiance) over 5 years at 47 observation stations across Japan. They found that the relative variation in the clearness index was approximately 4.3 % for skies with unobscured suns and 35 % for skies with obscured suns; corresponding variance-equivalent uniform distributions would have a range of 15 % and 121 %, respectively.…”

“…At present, there are only two published methods for the quantitative measurement of soot-BC emissions from individual in-field flares. One technique employs aircraft-based sampling of a flare plume (Gvakharia et al, 2017;Weyant et al, 2016), where measurements of soot, methane, and CO 2 concentrations during transects through the plume are used to provide flare-specific estimates of soot yield, using assumed flare gas compositions. The second technique is a ground-based remote optical measurement called sky-LOSA (line-of-sight attenuation using skylight; Conrad and Johnson, 2017;Johnson et al, 2010Johnson et al, , 2011Johnson et al, , 2013.…”

Analyzing big time series data in solar engineering using features and PCA

“…Besides the geographical locations, the next best approach to generate characteristics is using statistics. Watanabe et al (2016) used sample mean, variance and entropy to evaluate the variation in solar irradiance. Woyte et al (2007) used a wavelet-based localized spectral analysis to identify and classify the fluctuations in time series of the instantaneous clearness index.…”