Assessing the large-scale variation of heat budget in poorly gauged watershed-shallow lake system using a novel integrated modeling approach

Assessment of large-scale temperature dynamics and heat fluxes in shallow lakes requires comprehensive modeling capabilities to simulate water interactions, the processes related to the heat budget, and their driving forces. In this study, we investigated the spatial and temporal variations in water-surface temperature (WST) and surface heat fluxes for the period between January 2001 and December 2010 in Lake Mirim, a large (surface area ca. 4000 km(2)) shallow lake on the Brazil-Uruguay border. The aquatic ecosystem was examined using in-situ WST measurements, MODIS land surface temperature (M*D11A1 LST) products, and a coupling between a large-scale hydrologic model (MGB-IPH) and a hydrodynamic/water quality model (IPH-ECO). The WST values estimated by the hydrodynamic model were consistent with measured in-situ data and MODIS-derived WST, using both daytime and nighttime time series (R-2 = 0.88; Bias = 1.45 degrees C; RMSE = 2.16 degrees C; ENS = 0.77). Our findings also revealed that the dominant heat fluxes in the lake are the atmospheric longwave radiation (annual mean = 303.38 W m(-2)), water longwave radiation (annual mean = 336.01 W m(-2)), and the latent heat flux (annual mean = 51.14 W m(-2)). Additionally, the heat flux from the river inflows contribute minimally to the overall seasonal heat budget in the lake, being the Cebollati river the main contributor (annual average = 4.63 W m(-2)). Coupled large-scale hydrological and hydrodynamic models will benefit future modeling efforts by providing an effective management tool to study the effect of case scenarios (e.g., climate change) on temperature dynamics, heat fluxes, and physical and ecological processes in these ecosystems.