Elsevier

Renewable Energy

Volume 36, Issue 1, January 2011, Pages 1-8
Renewable Energy

Review
Experimental solar radiation measurements and their effectiveness in setting up a real-sky irradiance model

https://doi.org/10.1016/j.renene.2010.04.039Get rights and content

Abstract

The present work investigates the effectiveness of an innovative procedure to calculate the global real sky irradiance of a mountain urban region, the city of Trento (Italy). The proposed methodology improves the predictive Bird’s real-sky model by introducing in it both atmospheric parameters, specifically defined for the analyzed site, and a local cloud cover factor, based on experimental data, to calculate the global real sky irradiance. The experimental data have been measured at the meteorological station of the University of Trento located in the city center. At first, a selection of the global irradiance measurements, representative of daily clear-sky conditions of each season, is presented and compared with the corresponding values obtained by the improved Bird’s clear-sky model. Making use of the improved procedure, the monthly mean daily irradiation is then calculated and compared both with experimental measurements covering the years from 2003 to 2006 and available models as well as data banks. The results, presented in terms of statistical functions, demonstrate that the generalized calculation procedures usually adopted, also available from commercial software tools, reach a satisfactory accuracy if compared with an experimental methodology approach as the one proposed in this work.

Introduction

The availability of improved predictive models for the calculation of the real global irradiance is of great importance to support engineering solutions and decisions for advanced solar energy applications. A great diversity of solar radiation models has been developed considering different time averages (hourly, daily, monthly, etc.) or surface position (horizontal or inclined) as well as the different components of solar radiation (global, direct and diffuse). In literature the differences between global irradiance models are basically amenable to the evaluation of diffuse contribution. The simplest model considers diffuse irradiance to be isotropically distributed in the sky [1], while the more realistic models consider diffuse irradiation to be anisotropic. The latter, as they involve a relevant number of parameters, are complicated by complex fitting procedures. The most successful ones are those of Gueymard [2] and Perez et al. [3] that present advanced models based on the concept of “sky categories” improving in some way the idea of two specific sky zones models earlier proposed by Klucher [4] and Hay [5]. Recently, Li et al. [6] proposed a new model for estimating the global vertical irradiance based on beam and ground-reflected irradiance, while Badescu [7] has introduced a new formulation for the isotropic model, based on a three-dimensional model. Gomez et al. [8] present an innovative approach based on fuzzy logic procedures utilized to define clustering zones improving then the concept of the sky zone. The adopted optimization procedure makes use of a competitive learning Artificial Neural Network (ANN) not so easy to be applied and to be implemented in end-users' programs. A recent review on sky irradiance modelling, proposed by Torres et al. [9], can be helpful to deeply investigate and to model sky diffuse irradiance. The present study focuses on proposing a calculation procedure that evaluates the global irradiance of a defined site through a simple procedure directly involving local experimental data. Basically it improves Bird’s clear sky model by implementing in it an hourly cloud cover factor specifically calculated on experimental data covering, for the analyzed case, the period from 2003 to 2006. The originality of the model is that, though it requires experimental data as the recent cited ones, it avoids complicated fitting procedures. At the same time the investigated methodology allows to test the potentiality to utilize local experimental data in setting up a dedicated global irradiance calculation procedure. This appears to be of significant interest as it performs very well in evaluating the monthly daily mean irradiation if compared both with experimental data and results of advanced available models.

Section snippets

Bird’s model

Bird’s et al. [10] model is a diffuse radiation models widely used to calculate the hourly clear sky irradiation on a defined horizontal surface. Knowing the sun position at a defined time of a selected day and with respect to a defined location on the earth, this model extends the basic equations, reported as Eqs. (1), (2) in this section, usually adopted to determine the extraterrestrial irradiance I0. In the present paper the hourly sun position calculation is based on NOAA’s functions [11].

Real sky conditions

The real sky irradiance can be simulated if the cloud cover factor of the sky is known. In the cloudy sky version of Bird’s model [21] the total radiation IT,real is obtain by multiplying the total clear sky radiation IT,c, available from Eq. (5), with the term Tcl, corresponding to the mean relative global transmittance of clouds evaluated at the zenith angle of 45°:IT,real=IT,c·TclThe term Tcl can be determined following different approaches. A raw empirical methodology, based on the

Experimental data and atmospheric parameters definition for the investigated site

Measurements reported in the present section refer to the site of Trento, a mountain city located in Northern Italy whose main geographical coordinates are reported in Table 2. Table 3 reports the values adopted here for the atmospheric attenuators previously introduced and determined following the procedures described in Section 2.4. The solar constant Ics is assumed to be 1367 W/m2, while the atmospheric pressure P = 991 mbar refers to the mean value covering the years 2005–2006. The

Clear sky conditions

This section compares a selection of hourly global horizontal irradiation values obtained from the Bird clear sky model and the corresponding experimental values obtained at the “Molino Vittoria” station. Four days representative of clear sky conditions have been selected for each season of the year 2006. It is important to remark that the adopted Bird’s model is a free horizon model and, therefore, cannot take into account the shadowing effects of the mountains that have to be considered for

Conclusions

The present work investigates the capability prediction of the Bird model to calculate the clear-sky irradiance when applied to a defined site for which specific values of the atmospheric attenuators can be determined. In terms of AAD, the improved Bird clear-sky model reaches an accuracy prediction of about 2.16%. To determine the real-sky irradiance, an innovative calculation procedure has been introduced which extends the Bird clear-sky model by defining a specific cloud cover factor on the

References (35)

  • R.E. Bird et al.

    A simplified clear sky model for direct and diffuse insolation on horizontal surfaces

    (1981)
  • NOAA. General Solar Position Calculations. Available from:...
  • WMO. World Meteorological Organization. Available from:...
  • Woudc. World Ozone and Ultraviolet Radiation Data Centre. Available from:...
  • WDCA. World Data Centre for Aerosols, Institute for Environment and Sustainability, Joint Research Centre, Ispra (Va),...
  • ECMWF. European Centre for Medium-Range Weather Forecasts. Available from:...
  • Soda. Integration and exploitation of networked SOlar radiation DAtabases for environment Monitoring. Available from:...
  • Cited by (17)

    • Modelling solar potential in the urban environment: State-of-the-art review

      2015, Renewable and Sustainable Energy Reviews
      Citation Excerpt :

      On the other hand, for ideal input and all-sky conditions it was Reindl [29] which performs the best and gives noteworthy results for vertical surfaces, which are very sensitive to any inaccuracies in the ground reflected radiation due to its anisotropic features. A methodology to calculate the global horizontal real sky irradiance was introduced by Grigiante et al. [32] by improving an existing model, the predictive Bird׳s clear sky model [33], through the introduction of atmospheric parameters such as the cloud cover factor, specifically defined for the location. The definition of octas sky covered conditions is proposed as an empirical evaluation of the sky brightness based on experimental data.

    • The HelioMont method for assessing solar irradiance over complex terrain: Validation and improvements

      2014, Remote Sensing of Environment
      Citation Excerpt :

      Yet another promising approach are aerosol re-analysis projects like MACC (Inness et al., 2013) and GOCART (Chin, Rood, Lin, Müller, & Thompson, 2000; Chin et al., 2002), which assimilate satellite-based aerosol states, integrate them with known aerosol sources, and project their global distribution by use of atmospheric transport models. The proposed integration of surface aerosol measurements with satellite measurements enhances the applicability of satellite data and is more valuable than the analysis of single isolated stations or station networks (Grigiante, Mottes, Zardi, & De Franceschi, 2011). The resulting method improves the reliability and precision of solar radiation estimates over complex terrain, which is a key requirement for applications pertaining both to short-term weather forecasting, and to long-term climatological assessment of available radiation.

    • Simple models to compute solar global irradiance from the CMSAF product Cloud Fractional Coverage

      2014, Renewable Energy
      Citation Excerpt :

      For instance, three sky classes were used in Ref. [20]: skies without clouds (N = 0), partly cloudy (N = 1 ÷ 7) and overcast skies (N = 8). Four classes of skies were considered in Ref. [50] (see their Table 1): fine skies (N = 0 ÷ 2), partly cloudy (N = 3 ÷ 5), cloudy (N = 6 ÷ 7) and overcast skies (N = 8). Three classes of skies (i.e. C = 0 ÷ 0.3, 0.3 ÷ 0.7 and 0.7 ÷ 1, respectively) have been considered in Refs. [51–53].

    • Effects of changes in observational sites position and surrounding urbanisation on the temperature time series of the city of Trento

      2014, Urban Climate
      Citation Excerpt :

      A particular feature typical for Trento is the arrival of a lake breeze blowing from a nearby valley (de Franceschi et al., 2002, 2003; Laiti et al., 2013a,b). During wintertime ground-based temperature inversions frequently occur (de Franceschi and Zardi, 2009), as a consequence of the weaker solar radiation input and the reduced sky-view factor (Grigiante et al., 2010). All the above processes may significantly affect the spatial distribution of temperature in the urban area of Trento, adding additional complexity to the reconstruction of the time series.

    • Atmospheric boundary layer structures associated with the Ora del Garda wind in the Alps as revealed from airborne and surface measurements

      2013, Atmospheric Research
      Citation Excerpt :

      Notice also that RK method can be applied to any kind of atmospheric quantity, such as moisture, concentration of chemical species, aerosols, clouds and radiation (Baumgardner et al., 2011). Accordingly, RK analyses of airborne measurements could help achieving a more complete characterization of typical ABL processes over complex terrain, such as turbulence (de Franceschi and Zardi, 2003; de Franceschi et al., 2009), urban weather and climate (Giovannini et al., 2011, 2013, in press; Grigiante et al., 2011), and thermally-driven flows triggering atmospheric convection and eventually leading to precipitation events (Barthlott et al., 2006; Bertò et al., 2004; Gladich et al., 2011; Kalthoff et al., 2009; Pucillo et al., 2009). The understanding of the typical patterns of the above processes will be beneficial for various applications, in particular for air quality management.

    View all citing articles on Scopus
    View full text