عنوان مقاله [English]
نویسندگان [English]چکیده [English]
In order to provide more realistic representation of processes governing the water and energy balances as well as water quality and plant physiological processes, weather data are needed at finer timescales than currently are available in most regions. Air temperature is one of the critical variables that drives all biological systems, and is a fundamental input parameter for climatological and agricultural models (e.g. crop growth). However, for the application of agricultural models and for climatological studies, many of the available data consist of only the daily minimum and maximum temperatures. For many applications, it is often useful to obtain an approximation of subdaily data. In the literature, a variety of models with varying degrees of complexity have been developed to approximate diurnal temperature curves. The diurnal temperature variation has been modeled in a variety of ways that vary from simple curve-fitting models based upon sine, cosine and/or linear curves to more sophisticated models utilizing Fourier analysis and complex energy balance models. Although modeling approaches are different, most models are based on minimum (Tmin) and maximum (Tmax) temperatures. It is useful to have accurate temperature trend models that are not site-specific. However, most empirical models show variable results. This study aimed to derive the daily air temperature disaggregation parameters for some regions of Iran and to compare the performance of some daily-to-subdaily air temperature disaggregation models. For this purpose, air temperature disaggregation models, including WAVE I, WAVE II, WCALC, ERBS, ESRA, and TM models were calibrated and their performance was compared, using long-term daily and three-hourly weather data obtained from 12 different synoptic weather stations in Iran. The WAVE I method uses a cosine function assuming that the Tmax occurs at 1500. The WAVE II model uses a cosine function for the period from the time of minimum temperature to the time of maximum temperature and another cosine function from the time of the maximum temperature to the time of minimum temperature of the next day. The method is fixed at 1400 hours the time of the maximum temperature, and at sunrise the time of the minimum temperature. In the WCALC model, the day is divided into 3 segments: (a) midnight to sunrise + 2 h; (b) daylight hours; (c) sunset to midnight. The method assumes a change from night to day temperature at sunrise + 2 h, and the night temperatures are linear with time. In addition to the current day's Tmax and Tmin, the method requires the Tmax and Tmin of the previous day and the Tmin of the following day. The empirical Erbs model is a Location- and month-independent to explain the diurnal variation in daily temperature data. In this study, the daily average temperature and thermal amplitude were used instead of their corresponding monthly values since they were available. The TM model divides the day into three segments: (a) from the sunrise hour (Hn) to the time of maximum temperature (Hx), (b) from Hx to the sunset hour (Ho), (c) and from Ho to the sunrise hour for the next day (Hp). The model uses two sine-wave functions in the daylight and a square-root decrease in temperature at night. Hn and Ho are determined as a function of the site latitude and the day of the year. The time of the maximum temperature is set 4 hours before sunset. Temperature at Hn, Hx and Hp are the observed minimum, maximum temperature for the current day and minimum temperature for the following day, respectively. For the facility to use the daily data, the minimum temperature for the current day is used for the temperature at Hp. The temperature at time Ho is calculated with an empirical equation. The capability of the studied models for retrieving the measured three-hourly air temperature was evaluated, using root mean square error (RMSE) (?C), the mean error (ME) (?C), the mean absolute error (ME) (?C), Pearson correlation coefficient (r), and Nash–Sutcliffe model efficiency coefficient (EF). Different contributions to the overall error were decomposed using a regression-based method. The results indicated that the inversely estimated air temperature disaggregation parameters were in a reasonable physical range. Compared to the WAVE I, WAVE II, WCALC, ERBS, and ESRA models, the calibrated TM model had the best performance to disaggregate daily air temperature with EF, RMSE, and r ranging between 0.977–0.988, 1.059–1.808 ?C, and 0.983–0.994, respectively. Compared to air temperature disaggregation models with an arbitrary value for the time of maximum and minimum air temperature, the models in which the above mentioned times are described as a function of sunrise and/or sunset had better performance in describing the diurnal variations in air temperature. The use of the WAVE II model (with EF, RMSE, and r ranging between 0.966–0.984, 1.126–1.815 ?C, and 0.983–0.992, respectively) can be recommended for the regions with no subdaily weather data needed for calibration of the disaggregation models.