Qiu Jinhuan, Xu Xiaofeng, Yang Jingmei. An accuracy estimation of global solar radiation measurements at meteorological observatories in China. J Appl Meteor Sci, 2008, 19(3): 287-296. .
Citation: Qiu Jinhuan, Xu Xiaofeng, Yang Jingmei. An accuracy estimation of global solar radiation measurements at meteorological observatories in China. J Appl Meteor Sci, 2008, 19(3): 287-296. .

An Accuracy Estimation of Global Solar Radiation Measurements at Meteorological Observatories in China

  • When routine observations are performed by the pyranometer in the outside field, measurement errors will be caused by some factors, such as the polluted camera lens by which the measured value will be made less than the reality, as well as the temperature error, the cosine errorand the tilt error etc. Reliable accuracy estimation of the pyranometer-received global solar radiation (GSR) is significant to improve the use of the measurements. In the case of thin aerosol optical thickness (AOT) and small solar zenith angle, three characteristics of pyrheliometer-received GSR are presented. Compared with the diffuse solar radiation, the direct solar radiation contributes dominatingly to GSR; owing to the opposite shifts of direct and diffuse solar radiation with increasing AOT (the former decreases while the latter increases), GSR shows a weak sensitivity to AOT, molecular optical thickness, aerosol size distribution and refractive index; there is a small hemispherical reflectance and hence a weak sensitivity of GSR to the surface albedo. According to these properties, a method for accuracy estimation of GSR measurements at Chinese meteorological observatories is proposed. In the condition of cloud-free, thin AOT and small solar zenith angle, the solar constant (marked as E0, P) is retrieved from the GSR measurements and the broadband transmittance which is calculated using AOT retrieved from direct solar radiation. The deviation of the E0, P to the GRR (global radiation reference) is treated as an indicator of the uncertainty of GSR data. The simulations show that the uncertainties of the imaginary part of aerosol refractive index and the water vapor content are the two main error factors. Effects of random errors in input parameters such as aerosol optical properties, water vapor content and so on can be weakened by more statistical examples for the estimation, then more reasonable estimation results are produced. This method is applied to evaluating accuracies of GSR data at seven observatories which are Shenyang, Ejin Qi, Beijing, Urumqi, Golmud, Shanghai and Guangzhou in China. There are totally 1161 E0, P retrievals from GSR data in these seven observatories during 2000—2004, which are all measured in such conditions as clear (cloud-free) days; the solar zenith angle cosine (μ0) being larger than 0.7; τa(750 nm AOT) /μ0 < 0.3; the atmospheric column water vapor content being less than 2.0 cm. Among all these E0, P retrievals, the maximum deviation to GRR is 7.33 %, deviations of 97.78 %, E0, P retrievals are less than 5%, and the standard deviation is within 4.05% for every observatory. It is estimated that the uncertainty of GSR data in these observatories is usually better than 5% in the clear days and for μ0≥0.7.
  • loading

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return