تجزیه و تحلیل حساسیت از عملکرد رشد محصول مدل های شبیه سازی برای روش های برآورد تابش خورشیدی روزانه در ایران
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|26082||2009||11 صفحه PDF||سفارش دهید||6023 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Energy Conversion and Management, Volume 50, Issue 11, November 2009, Pages 2826–2836
Solar radiation is the single most important environmental factor driving canopy photosynthesis and transpiration. This weather characteristic is measured only in a limited number of weather stations. Hence, in many situations it has to be estimated from other weather characteristics such as sunshine duration and temperature using empirical relations. In this study, the Ångstrom and Hargreaves formulas have been used for solar radiation estimation, based on monthly and annual weather data for three weather stations in Esfahan province, Iran. Deviations of estimated solar radiation from measured values (both absolute and relative) varied with month of the year and with estimation method. Estimated and measured radiation values were used in a crop growth simulation model to explore sensitivity of simulated production with respect to radiation estimation method. Maximum deviation for winter barley and silage maize was around 9%.
Crop growth models are computer programs that integrate information on daily weather, crop characteristics, soil characteristics, and management to calculate crop growth and yield . They are used, among others, to support agricultural decision-making and learning processes  and . Solar radiation provides energy for photosynthesis and (evapo) transpiration of crops and soils ,  and . Daily solar radiation is therefore one of the major inputs in crop growth models ,  and , required for calculation of daily gross CO2 assimilation, the basis for dry matter production and yield. In spite of the importance of solar radiation for crop growth, it is not routinely measured at all meteorological stations, probably because of the cost and the maintenance and calibration requirements of the measuring equipment ,  and . A number of empirical relations of varying complexity have been developed to estimate solar radiation at a given location from other climatic characteristics that are measured more frequently . Relations have been established with sunshine duration ,  and , air temperature ,  and , cloud cover  and  and combinations of different weather characteristics  and . Results of these studies have shown that sunshine-based models are more accurate than temperature-based models for estimation of daily solar radiation ,  and . Crop growth models have been tested for sensitivity of simulated crop yields to inaccuracies in solar radiation estimates ,  and . These analyses have included use of (10-day, monthly and seasonal) average weather data instead of daily data , using radiation from nearby stations , estimation of daily solar radiation from monthly means , exploration of climate change effects on solar radiation , and filling missing data of radiation by values generated by different methods of solar radiation estimation . Fodor and Kovacs , using crop growth model 4M  to analyze the sensitivity of yield to inaccuracies in measurements of weather characteristics, found that 2% error in solar radiation caused 3.7% and 2.3% error in simulated grain yield and biomass of maize, respectively. They also showed significantly larger inaccuracies in calculated yield due to errors in measured radiation in years with low yields than in years with high yields. Nonhebel  and  showed deviations of 5–10% in the simulated yield of spring wheat in both, the water-limited and potential production situations  with 10% inaccuracy in solar radiation. Water-limited production appeared less sensitive to inaccuracies in solar radiation than potential production. The analysis also showed that randomly replacing 10% of the solar radiation data by average values did not significantly affect simulated water-limited and potential yields of spring wheat. Xie et al. , in analyzing the sensitivity of sorghum and maize yields to solar radiation found changes of less than 8% at 10% changes in solar radiation. Trnka et al.  analyzed the effects of different estimation methods for daily global solar radiation on simulated yields of winter wheat and spring barley in the Czech Republic and Austria, using the WOFOST  and CERES  and  crop growth models. Simulated yields based on solar radiation estimated by the Ångstrom–Prescott  and Hargreaves  formulas deviated more than 10% from simulated yields based on measured radiation in 6% and 48% of the cases, respectively (in 1.4% and 16.3% of the cases deviations exceeded 25%). Moreover, the sensitivity varied with soil type. As a result of the model structure, diverging values of solar radiation resulted in many cases in disproportionate diversions in actual transpiration. Soltani et al.  investigated the sensitivity of simulated yields of wheat, maize and soybean to daily radiation generated by linear interpolation from monthly means. Their results showed around 23% difference from yields simulated on the basis of measured radiation. They concluded that only in specific situations, monthly average radiation data can be used as input in crop growth simulation models. Pohlert  randomly replaced 4.8% of the measured solar radiation data in two temperate (Wageningen, The Netherlands and Cordoba, Spain) and one tropical location (Los Baños, Philippines) by values estimated by different methods. Yields of maize, simulated with WOFOST  with these different sets of radiation data, were not significantly different. Nonhebel  found that use of average weather data in crop growth simulation models in the water-limited production situation resulted in overestimation of spring wheat yields in wet conditions and underestimation in dry conditions. In most studies reviewed here, a fixed percentage of error (over- or under-estimate), uncertainty or change in solar radiation has been considered. Radiation estimates based on various methods produced different results with different magnitudes of error. Moreover, in most cases, sensitivity analyses of crop yields were based on empirical formulas with annual coefficients. However, it has been shown that these coefficients may show strong temporal variation . Moreover, daily solar radiation estimated with empirical formulas may show much larger deviations from measured values than have been established in earlier work. The objective of this study was to examine the sensitivity of potential yields and evapotranspiration of a winter crop and a summer crop grown in Iran, simulated by the WOFOST model  to radiation estimates by sunshine and temperature-based models with different sets of annual and monthly coefficients for three weather stations.