Category Modeling Solar Radiation at the Earth’s Surfac

Qualitative Assessment

Most qualitative results appear in the form of scatterplots, which visually indicate the bias (systematic error) and scatter (random error) of predicted vs measured val­ues. Results from a perfect model would align on the 1:1 diagonal when compared to their perfectly measured counterparts. As an example, consider the measurement of instantaneous clear-sky irradiance and its prediction from meteorological data. In recent years, progress has been made due to the convergence of significant improve­ments in radiometry (described in Chap. 1), time resolution (radiation data are now often measured at e. g., 1-6 minutes intervals rather than hourly intervals), measure­ment of key ancillary data (such as aerosol optical depth or precipitable water), and radiation modeling.

Examples of scatterpl...

Read More

Model Validation and Performance Assessment

An error analysis can only provide general information corresponding to ideal or “worst-case” scenarios. But what is the actual accuracy of a model in prac­tice? Answering this question requires a specialized study called “validation” or “performance assessment”. A model can be declared “validated” even if it does not perform very well or better than others. It is only a way of saying that “it works”. A true performance assessment consists of a series of tests whose findings are usually summarized by qualitative and/or quantitative results.

Read More

Model Sensitivity to Input Errors and Error Analysis

A radiation model is driven by input data that are directly or indirectly related to the optical characteristics of the atmosphere for the location and period considered. The most sophisticated radiative transfer models used in atmospheric physics require a wealth of information about various atmospheric constituents, such as gases and aerosols, and their vertical distribution. In engineering and other disciplines, the input requirements may be vastly different, and usually simpler.

For models whose Criteria 2 and 9 above call for meteorological inputs and clear-sky results, the main factor will be aerosol turbidity, particularly for the di­rect and diffuse radiation components (Gueymard 2005a), and the second factor will be water vapor (Fig. 20.1)...

Read More

Model Validation Principles

A perfect model does not exist. Even if it existed, this would be impossible to as­certain because the “true” solar irradiance cannot be determined theoretically or measured experimentally with perfect certainty. Chapter 1 described the various sources of uncertainty in experimental radiation measurement. These must always be bore in mind when evaluating the performance of any model against measured data (Gueymard and Myers 2007). Normally, when a new model is proposed in the literature, it should be also tested so that potential users can be sure of its validity un­der such or such conditions. This, however, is not done systematically by all model authors...

Read More

Types of Models

Different types of models have been developed to provide the community with pre­dictions of solar radiation when or where it is not measured appropriately or at all. An accepted typology of solar radiation models does not currently exist; hence, what is proposed below should be considered tentative. From an exhaustive review of the literature over the past four decades, it is clear that radiation models can be categorized in different ways. The previous chapters concentrated on a few specific types of model. More types do exist, so that nine classification criteria have been identified, as follows.

• Criterion #1—Type of output data

Outputs ideally consist of direct, diffuse and global irradiance, but frequently only one component is necessary (e. g...

Read More

Validation and Ranking Methodologies for Solar Radiation Models

Christian A. Gueymard and Daryl R. Myers

1 Introduction

This chapter provides an overview of the methodologies that can be used to validate different types of solar radiation models currently in use in various applications, with a focus on solar energy applications.

Read More

Results and Discussions

The monthly radiation maps and the yearly radiation maps obtained from the above- mentioned process are shown in Figs. 19.7 and 19.8, respectively. As expected, the

monthly maps demonstrate a seasonal variation of global radiation. In January, the solar radiation is relatively low in the north and the east of the country. This is likely due to the fact that the northeast monsoon still influences these parts of the country, bringing cloudy skies to these mountainous areas. In addition, the sun is still in southern celestial sphere, causing less solar radiation in the north. The high solar radiation areas expand from the south to the entire country from February to May as the sun moves from the south to the celestial equator with the highest solar radiation in April...

Read More

Mapping of Global Radiation Over Lao PDR

The monthly averaged values for all model parameters are calculated at each pixel of the satellite image covering Lao PDR. Then the model is used to compute the monthly averaged daily global radiation for every month over a period of 12 years (1995-2006). The calculation of the air mass in the absorption coefficients of the model also accounts for the surface elevation. For each month, the monthly average of daily global radiation was again averaged over the 12 years to obtain the long­term average global radiation. The results are displayed as monthly solar radiation maps and a yearly map.

Read More