Quoted from:Estimation of daily and monthly direct, diffuse and global solar radiation from sunshine duration measurements[J]. Solar Energy, 1984.
https://www.sci-hub.ren/10.1016/0038-092x(84)90267-6
Abstract
An accurate 200 W/m 2 threshold pyreheliometer instrument for measuring the duration of bright sunshine has been used to derive daily and monthly regressions for direct, diffuse, and global solar radiation component vs sunshine duration. Daily regression for diffuse/global are linear in sunshine duration, while quadratic regression forms are employed for direct normal, direct horizontal, and global/extraterrestrial components. Only the daily direct normal component had regression values which depend on season while all of the monthly regressions depend on season. Linear regression relations for monthly direct normal, diffuse/global and global/extraterrestrial are employed, with a quadratic form being used for direct horizontal. Effects of rainfall, especially in overcast conditions, and of atmospheric turbidity and precipitable water, especially under clear-sky conditions, are observed and documented.
INTRODUCTION
Thermal analyses for energy conservation or solar energy applications require knowledge of the solar radiation incident on buildings or collector surfaces which are typically inclined with respect to the horizontal. In the absence of solar radiation measurements on an appropriately tilted surface, this radiation may be estimated by one of several methods which usually require values of diffuse and direct components on a horizontal surface. Other applications, such as focusing or concentrating solar collectors, require values for the direct beam radiation, which, in the absence of direct beam measurements, may be estimated directly from the global component or indirectly from the difference between global and estimated diffuse components.
METHODS
Benson et al. have obtained regression constants in to two intervals of a year depending on the climatic parameters:
\( 𝐻/𝐻_0 = 0.18 + 0.60 𝑆/𝑆_0 \) for Jan–March and Oct–Dec
\( 𝐻/𝐻_0 = 0.24 + 0.53 𝑆/𝑆_0 \) for rest of the month