Seasonal variability of daily evapotranspiration and energy fluxes in the Central Andes of Peru using eddy covariance techniques and empirical methods

In this study, we analyze the mechanisms associated to evapotranspiration over the high central Peruvian Andes, a place where evapotranspiration has been poorly characterized. We made use of the eddy covariance system (sonic anemometer and a krypton hygrometer) installed at the Huancayo Observatory (12.04° S, 75.32°, 3330 m.a.s.l.) to document for the first time the hourly, daily and monthly variability of surface energy fluxes from July 2016 to June 2017. The relationship between evapotranspiration and meteorological variables is also examined. Furthermore, we evaluated the performance of three empirical equations that estimate the potential evapotranspiration to explore their adequacy in the central Peruvian Andes. These are the FAO Penman-Monteith (PM), Priestley-Taylor (PT) and Hargreaves. Likewise, the accuracy of the MODIS16A2 evapotranspiration product was also examined. We show that evapotranspiration over the high central Peruvian Andes is modulated by the water- and energy-limited states during the dry and wet season, respectively. During the wet season (January–March), latent heat flux (LE) is greater than sensible heat flux (H), and the daily evapotranspiration variability is mainly related to incoming solar radiation (SW↓; R² = 0.76, p-value < 0.01), having a daily mean of 3.45 mm. In contrast, the dry season (June–August) is characterized by a greater H with a daily mean evapotranspiration of 0.95 mm. Furthermore, evapotranspiration is significantly tied to the soil moisture variability on daily time scales (R² = 0.77, p-value < 0.01). Of the three equations, PT has a good performance in reproducing the daily evapotranspiration variability (R² > 0.72; p-value < 0.01). In contrast, strong biases are noticed during the dry season mainly because this empirical equation does not account for the soil water content. Thus, our results show that the inclusion of soil water content or a physiological-plant parameter would be necessary to estimate the real evapotranspiration during the dry season.