Sensible heat flux assessment in a complex coastal-mountain urban area in the metropolitan area of Rio de Janeiro, Brazil

Abstract

Obtaining accurate estimates of surface-atmosphere energy exchanges and improved comprehension of the mechanisms generating turbulent fluxes over non-uniform landscapes are both challenging goals and essential for validating numerical weather forecasting modeling. In this work, a large-aperture scintillometer (LAS) was deployed to evaluate the effect of mesoscale sea and valley breezes with respect to diurnal sensible heat flux (H) variability over an urban area close to a coastal bay near the metropolitan area of Rio de Janeiro (MARJ) during the austral summer of 2017–2018 time span. The region is characterized by a densely urbanized environment, surrounded by mountainous relief and bounded by a large ocean bay. The transmitter and the receiver of the LAS system were installed 1955 m apart. Diurnal cycle was divided into four periods according to the local time: dawn (00 a.m. to 6 a.m.), morning (6 a.m. to 12 a.m.), afternoon (12 a.m. to 6 p.m.), and night (6 p.m. to 00 a.m.). The results demonstrated the sensitivity of the LAS to record variations in H over the course of the day, taking into account the corresponding sea and valley breezes and the relative humidity variability. During the morning, when valley breezes are predominant, footprint area shrinks to its smallest size, encompassing an area of approximately 2 km2. On the other side, during afternoon and night time periods, during afternoon and night time periods, the presence of sea breeze jointly with a densely urbanized continental area contribute to promote significant changes in the footprint area ranging from 2.05 to 5.38 km2. The effects of topography are also well captured in the diurnal H cycle once sunset and corresponding shading in the late afternoon abruptly reduces heat-sensible flux. Diurnal variations in footprint shape and area lead to modifications in terms of the mode each type of urban surface will contribute to H, increasing the influence of anthropogenic elements in afternoon and early evening. Complementarily, it is noteworthy to mention that a sensitivity test was performed to assess uncertainties in H estimates for the free convection method, which requires the estimation of an empirical constant b. Variations in this constant introduce larger dispersion in H estimates than potential errors in LAS effective height (Zeff) estimations or temperature measurements. These last results highlight the importance of conducting further evaluations as the one pursued in this study to more rigorously validate the outputs of numerical mesoscale and local hydrometeorological models and remotely sensed products.