The adaptability of empirical equations to calculate potential evapotranspiration and trend analysis of hydroclimatological parameters for agricultural areas in Newfoundland

Abstract

Calculation of potential evapotranspiration (PET) has been problematic in Newfoundland (NL) due to the lack of measured data. Therefore, PET data obtained from the Pacific Field Corn Association for St John’s, NL was compared against five empirical PET calculation equations (i.e. (i) radiation-based Priestley-Taylor (PT), and Makkink (M), (ii) temperature-based Hargreaves-Samani (HS), and Turc (T), and (iii) location-based Hamon (H)). Evaluation based on the results concluded that the HS equation would be appropriate to calculate PET in NL. Further calibrations and validations were done to modify the HS to better calculate PET for the growing season (May-October) in NL. The modifications improved the Root Mean Square Error (RMSE), Nash-Sutcliffe Efficiency (NSE) and co-efficient of determination (R2) of the validated data. Trend assessment carried out using Innovative Trend Analysis (ITA) and Mann-Kendal (MK) tests indicated that both methods were in par with each other. Most of the significant positive trends of monthly total precipitation (0.375-2.210 mm/month/year) were available for September and October. Positive trends for minimum and maximum temperatures were found mostly concentrated within August and September with increments ranging from 0.015 to 0.062 ºC/month/year. PET trends of magnitudes up to 0.011 mm/month/year were observed mostly within September and October. Total water balance did not show as many positive trends as other parameters considered. However, the available positive trends (ranging from 0.018 to 0.076 mm/month/year) were also focused mostly within September. As a conclusion, the HS equation with modifications and error margins (where necessary) can be used to calculate PET accurately for the growing season in NL, and positive trends are observed mostly within the later periods of the growing season. The results of this study could be used in consideration of agricultural expansion, selecting cropping systems and water management systems of NL in future.