Açık Ve Bulutlu Atmosfer Koşullarında Saatlik Toplam Işınım Öngörüsü İçin Bir Model

Abstract

Bu çalışmada, yeryüzeyine gelebilecek saatlik toplam ışınım öngörüsü, açık ve bulutlu atmosfer şartla rında olmak üzere iki kısımda incelenmiştir. Öncelikle açık bir atmosferde yeryüzeyine gelebilecek saatlik top lam ışınımın hesaplanması için parametrik bir model su nulmuştur. Bu modelde, atmosferde güneş ışınımını azaltı cı faktörler olarak ozon, kuru hava molekülleri, subuharı ve aerosoller gözönüne alınmıştır. Saatlik toplam ışınım, aynı periyottaki direkt ve yaygın ışınımın toplamı olarak hesaplanmıştır. Yeryüzeyine ulaşan saatlik direkt ışınım, ozon tarafından absorbsiyon, kuru hava molekülleri tara fından saçılma, subuharı ve aerosoller tarafından absorb siyon ve saçılma etkileriyle ilgili geçirgenlik fonksi yonları yardımıyla elde edilmiştir. Saatlik yaygın ışınım direkt ışınımın ve yerden yansıyan toplam ışınımın atmos fer bileşenleri tarafından saçılması şeklinde iki bileşe ne ayrılarak incelenmiştir. İstanbul için uygulanan modelde, atmosferdeki subuharı miktarı, radyosonde verilerinden yararlanılarak, aerosol parametresi ise toplam ışınım ölçümleri ve model sonuçları karşılaştırılarak aylık ortalama değerler ha linde elde edilmiştir. Ayrıca ozon absorbsiyonu, molekü- ler saçılma, subuharı ve aerosoller tarafından absorbsi yon ve saçılmayla ilgili geçirgenlikler zenit açısına bağlı olarak verilmiştir. Aerosollerin, güneş ışınımını en çok etkileyen atmosfer bileşeni olduğu görülmüştür. Bulutlu atmosferde, yeryüzeyine gelebilecek sa atlik toplam ışınım değerleri hesaplamak için, bulutlu atmosferde ölçülen saatlik toplam ışınımın, açık günde gelebilecek saatlik toplam ışınıma oranı ile, aynı peri yottaki bulut kapalılığı ve ışının zenit açısı arasında bir çoklu regresyon bağıntısı ileri sürülmüştür. Sözkonu- su bağıntı, alçak, orta ve yüksek bulut tipleri ve çeşit li zenit açısı değerleri için incelenmiştir. Her üç bulut tipi için, bulut geçirgenlikleri zenit açısına bağlı ola rak hesaplanmıştır. Yüksek bulutların güneş ışınımını or talama %50 sini geçirdikleri, alçak bulutların ise atmos fer geçirgenliğini önemli derecede etkiledikleri görül müştür.

In recent years, mainly due to its renewable and nonpollutant character, Solar energy (utilization) has been a subject of utmost importance. In this context, Turkey is known to have a considerable amount of potential. Owing to its location within the world sun zone, Turkey has annually mean value of sunshine duration of 26.40 hours and annually mean solar intensity of 308 cal/m.day. Arising from this fact, studies on solar energy utilization in Turkey have been receiving a growing interest. Materialization of solar energy-based projects requires an accurate determination of source, intensity and characteristics of solar energy. In order to determine the areas with solar energy potential and to obtain a detailed solar radiation climatology for a particular area, a long term extensive solar radiation data set is essential. However, several calculation procedures have been developed for regions where the data cannot be obtained or tor augmenting short term solar data. Presently, there are many statistical, numerical and parametric models for estimation of direct, diffuse and global radiation. An extensive review of these models is given in the first chapter of this thesis. In this study, we first develop a parametric model to estimate hourly total solar radiation under cloudless sky in Istanbul (41.1 N). Then we proceed to incorporate the effect of cloud and formulate necessarry relationships to calculate the total solar radiation at the surface with cloud present. Concequently, a multiple regression relationship has been set up between the ratio of the total solar radiation measured during cloudy hours to the total solar radiation calculated for clear hours, hourly cloud cover and the zenith angle of the sun. Also, an extensive investigation of the coefficients of this relationship for low, middle and high cloud types and various zenith angles has been carried out. - XI The first chapter of this thesis is devoted to a state of the art review of the related studies. In doingso, complemantary to the models to estimate direct, diffuse and total radiation special attention has been given to the studies on solar constants, absorbtion and scattering of solar radiation. The theory on incident solar radiation at the earth-atmosphere system has been given in the second chapter. Absorbtion of solar radiation by ozone, water vapour and aerosols, and scattering of it by dry air molecules and aerosols have been extensively reviewed. Absorbtion, reflection and transmission of solar radiation by clouds of different types have been outlined. In the third chapter, general aspects of the theoretical, parametric and statistical models for clear- and cloudy-sky conditions have been presented. In addition, some experimental studies on cloud transmittance have been discussed. A parametric model for estimating hourly total surface radiation for cloudless sky of Istanbul (41.1 N) is given in the fourth chapter. Whereas, multiple regression equations to estimate hourly total radiation for cloudy atmosphere in Istanbul are developed in the fifth chapter. Correlation and error analyses between measured and calculated hourly total radiation for both cloud-free and cloudy conditions are carried out in the same chapter. Following a comparative evaluation of the results obtained from the models developed in this study with those obtained from others, conclusions and comments are outlined in the sixth chapter. To arrive at a model to the hourly global radiation at the horizontal surface, it was considered as the sum of direct and diffuse radiation components. In this model, ozone, dry air molecules, water vapour and aerosols were assumed as reducing factors of irradiance. This led to the following expression in which the direct irradiance appears as multiplication of extraterrestrial irradiance and transmittances of atmospheric attenuation components I = I. T. T. T-.T.T o oz R wa ws a where, I0 is extraterrestrial irradiance, whereas Toz, T-R T'wa., Tws and Ta are transmissivities after absorbtion by ozone, Rayleigh scattering, absorbtion by water vapour, scattering by water vapour and extinction by aerosols, respectively. - x 11 - On the other hand, diffuse irradiance on a horizontal surface was calculated as the sum of two diffuse irradiance components. One of these components is the primary diffuse irradiance (D,) and is due to the scattering of the direct beam by dry and water vapour molecules and aerosols in the atmosphere. The other diffuse component is a secondary one and is also due to the multiple scattering between the ground and the atmosphere. However, the primary component represents the forward scattered radiation from-and the secondary component represents the backscattered radiation towards the ground. In calculating these components, the following equations were used D, = 0.60 I T, T. T. (1-T T T ) 1 o oz wa aa ws R as D0 = a(I +D, ) [0.40 T.. T. (1-T ? T_. T.)] 2 1 ws aa ws R as In these equations, 0.60 and 0.40 define the forward scattering and backscattering factors. Taa and Tas stand for transmissivities after absorbtion and scattering by aerosols. Taa- ', ^wa-'» Tws', Tr ' and Tas ' represent transmissivities of backscattered radiation The transmissivitv terms employed in the above expressions of the direct and the diffuse radiations are functions of amount of ozone, aerosols and water vapour in the atmosphere and the optical air mass. Monthly mean values of precipitable water for Istanbul were calculated from radiosonde data of 10 years. In determining monthly aerosol parameters, hourly values of total radiation at clear sky conditions were used. Under these circumstances, with the help of computer programs hourly values of total radiation were calculated for each day during one year in Istanbul. Error analysis of the calculated hourly total radiation with respect to the measured radiation reveals that the mean, the root mean square and the relative errors are 0.034 MJ/m.h, 0.106 MJ/m.h and 0.045. On the other hand, the correlation coefficient between the calculated and the measured values is 0.986. As to the comparison of the individual estimated values with the corresponding measured ones : The deviation increases when clear atmosphere conditions cannot be met due to haze and smoke influence which takes place especially around the time of sunrise and sunset. The forecasting model developed herein shows a better performance at noon time when the zenith angle is maximum. - xni - The effect of zenith angle on the transmittances in relation with the amount of ozone, water vapour, air molecules and aerosols and that on the total transmittancy for direct, radiation were investigated in detail. In doing so, the" annually mean values of zone, precipitable water vapour and aerosol parameter were respectively considered as 0.33 cm, 1.900 cm and 0.88. It is found that when the zenith angle is less than 70, the transmittance due to ozone absorbtion is approximately 99 %. Furthermore, the transmittancy due to absorbtion and scattering by water vapour is approximately 3 % larger than that due to molecular scattering. More importantly, the main attenuation factors affecting the atmospheric transmittance are absorbtion and scattering by aerosols. Calculations for 21 of June solar noon time in Istanbul illustrate these points. For this particular case, the zenith angle is 17.54 and the transmittances due to Rayleigh scattering, absorbtions and scatterings by water vapour and aerosols were respectively calculated as TR = 0.89, T =0.91 and T = 0.77. At the same value of zenith angle, the atmospheric transmittances for direct and total radiation were obtained as I/I = 0.62 and G/G = 0.64. ° o In evaluating the incoming radiation to the earthsurface, definition of the cloud effects on direct, diffuse and total radiation is important. The radiative effect of clouds is similar to that of aerosols. They absorbe and scatter incoming radiation. Of course, the clouds at various altitudes have different effects on the incoming radiation. Additionally, the total cloud cover, the form of cloud and the position of it with respect to the sun disc are also important parameters influencing the incoming radiation to the surface. Herein, unlike other studies, not only the attenuation effect of the cloud type and the cloud cover but also the effect of zenith angle were taken into account. After extensive testing considering low middle and high cloud types and various classes of zenith angle the following multiple regression equation was established : Gb/G = a l + a2 C CosZ + a3C + a^ CosZ this equation provided a relationship between the ratio the total hourly ration measured under cloudy sky condition to that determined from the model for clear atmosphere, the cloud cover and the zenith angle. Within - xiv - the category of low clouds Cumulus, Stratus, Stratocurnulus, and Cumulonimbus were considered. As middle clouds Altostratus and Altocumulus, as high clouds, cirrus, Cirrostratus were taken into consideration. In addition, it was assumed that there existend only one of these types of cloud layer in the atmosphere at a time. Also assuming that the total cloud cover remained constant for a period of one hour, corresponding to the total radiation in a one-hour-period the total cloud cover observed at the beginning of that period was taken xnto consideration. To arrive at the multiple regression equation, 668 hourly cloud cover and hourly total radiation data were used. Calculated values of hourly total radiation for various cloud types and zenith angles are given in Tables 5.1, 5.2 and 5.3. Regardless of the cloud conditions, the mean multiple correlation coefficient between the ratio of the hourly total radiation at cloudy atmosphere to the hourly total radiation predicted at clear atmosphere (G,/G), the terms of cloudness and CpsZ is 0.910 and the mean standard error is 0.008 MJ/m.h. Using the equations of transmittance determined for low, middle and high level clouds, the relationship between the atmospheric transmittance and the cloud cover was analysed for various values of zenith angle (Z = 25, 40, 60 and 75 ). In general, as the cloud cover increases the atmospheric transmittance decreases. Additionally, considering the cloud transmittance under the conditions of overcast sky, the mean transmittances for low, middle and high level clouds were calculated. These values, in corresponding order, are 0.25, 0.26 and 0.53. The high level clouds transmitte almost half of the solar radiation. For all the cloud types considered in this study there is a very strong correlation between the measured and the calculated values of the hourly total radiation. The correlation coeficients for low middle and high level clouds are respectively 93.0 %, 96.1 %, 97.8 % and the mean relative error is 13.3 %. Finally, a comparison for selected days was made of the models developed in this study with the relevant ones. This comparison reveals that the models presented herein perform better with respect to the others.