Experimental Evaluation of Regression Relations of Instantaneous Photosynthetically Active Radiation (PAR) and Total Solar Radiation (TSR)

Exposure of human, animal and plants to sunlight has a major role for their growth. One of the most important applications of solar radiation is the agricultural sector. Photosynthesis is a photobiological phenomenon that depicts the ability of plants to convert light energy into chemical energy. In fact, if we provide suitable water and temperature, plant growth and consequently crop yields are directly dependent on photosynthetic active radiation (PAR). Regarding the importance of monitoring of PAR flux in agriculture, unfortunately, in most meteorological stations, this parameter is not routinely measured, as its determination is acostly process.
In this study, the radiation parameters were measured in a meteorological station located at the faculty of Agriculture, Bu-Ali Sina University in Hamedan. The station has a geographical position of 34 degrees and 47.91 minutes North latitude, 48 degrees and 28.98 minutes Eastern longitude and 1851 meters above sea level in an open space land inside the university campus (Hamedan, Iran). The climate of Hamedan is cold and semi-arid.Geonica Data Logger (GDL) and the PAR detector asradiation devices wereused in this study. The scientific name Pyronometer light sensor which is connected to GDL is LPO2 (Huksellux). The sensitivity of the sensor is between zero to 2000 watts per square meter, and its spectroral response rangesfrom 305 to 2800 nm. The intensity of the irradiance measured by the PAR device is from zero to 2000 µmols/(m2.sec) and its spectroral response covers 380 nm to 750 nm. The method used inthis workwas to measure daily PAR data from a PAR device (ELE) at least four times a day at a local time from April 2016 to February 2017. At the same time, the TSR data was also recorded bythe Geonica Loggerin nearby meteorological site. In this study, simple linear regression and exponential regression wereemployedto investigate the relationship between the TSR data(independent variable, predectors) and the PAR variable (a dependent variable). Using SPSS software, 70%of the data was used to construct the regression relationships and the remainder for evaluating the accuracy of the obtained relationships. Due to the different weather conditions, the measured data are divided into four groups: Clear Sky, Partly Cloudy Sky, Overcast, and All Condition (All sky). To report the cloudiness, Okta unit is used parts (e.g. each Okta corresponds to about 12.5% cloud coverage).
The analysis of regression relationshipbetween TSR and PAR in the clear sky, partly cloudy sky, overcast and all sky wasperformed for monthly, seasonal and annual scales. There wasa linear relationship between TSR and PAR fluxes. This linear relationship decreasedwith increasing cloudiness for both monthly and seasonal scales. These results were compared with those ofEscobedo et al. (2009) whomodelled hourly and daily fractions of UV, PAR and NIR to global solar radiation under various sky conditions at Botucatu, Brazil. Our findings were also in good agreement with their results, as they also observed a linear correlation between PAR and TSR fluxes at Botucatu. Moreover, the ratio between PAR and TSR was determined for all time scales. Our results showed that the highest and lowest ratio of PAR /TSR occurs in July (0.448) and February (0.407), respectively. Onseasonal and annual scales, the ratio PAR /TSR increasedas the sky conditions changed from the clear sky to the cloudy sky, mainly because of the effect of cloudiness. Cloudy sky absorbs longer wavelength radiation of solar spectrum (such as infrared radiation) ascompared withshort wavelengths (such as PAR and UV). This increases the radiation proportions from the clear sky to the cloudy Sky. Our results are in good agreement with the results of Alados et al. (southeast Spain), Papaioannou et al. (Athens), Jacovides and et al. (Eastern Mediterranean basin) and Udo and Aro in central Nigeriawhoexamined the PAR/TSR ratio.
In the present study, the following results were achieved:In monthly, seasonal and annual time scales, there wasa linear regression relationship between PAR and TSR varying with the change in clouds cover. The best correlations were observed in June and July, but the correlation coefficients decreased from October to February (autumn and winter) due to the increased cloudiness.
The PAR/TSRratio in the seasonal time scale showed an increment as the cloud cover increased. On annual scale, the ratio of photosynthetic active radiation (PAR) to total global irradiance (TSR) increased from 0.430 in clear sky to 0.489 in overcast condition.