Este artículo presenta los resultados de una investigación que busca establecer la relación entre las variables de calidad de agua y las variables hidrometeorológicas en un embalse tropical. La hidroeléctrica "La Miel I", localizada en Colombia, fue utilizada como caso de estudio. Se emplearon análisis estadísticos aplicando técnicas univariadas y multivariadas así como también, herramientas de modelación estocásticas basadas en vectores auto-regresivos. Para el análisis se usaron cinco puntos de monitoreo de calidad de agua y dos estaciones hidro-meteorológicas ubicadas dentro del área de estudio. Los resultados obtenidos de los análisis permitieron establecer que las series de tiempo de calidad de agua no presentan una distribución normal contrario a las series de tiempo hidro-meteorológicas. Los análisis de correlación mostraron una cercana relación entre la calidad del agua y la hidro-meteorología, específicamente en variables como la precipitación, el caudal, la temperatura del aire y el brillo solar. El análisis de componentes principales permitió identificar que la fortaleza de las relaciones entre los parámetros de calidad de agua y la hidrometeorología cambia dependiendo de la localización del punto de monitoreo. Finalmente, La implementación del modelo de vectores auto-regresivos mostró unas métricas de desempeño menores que uno para el pronóstico de oxígeno disuelto utilizando series de tiempo hidrometeorológicas y de calidad de agua como entrada. | This article presents the results of an investigation that seeks to establish the relationship between water quality variables and hydrometeorological variables in a tropical reservoir. "La Miel I" hydroelectric plant, located in Colombia, is used as case study. Statistical analysis was used applying univariate and multivariate techniques as well as stochastic modeling tools based on vector autoregression. For the analysis, five water quality monitoring sites and two hydrometeorological monitoring stations located inside the study area were used. The results obtained from the analyzes allow establishing that the water quality time series do not present a normal distribution unlike the hydrometeorological series. The correlation analysis showed a close relationship between water quality and hydrometeorology, specifically in variables such as precipitation, flow rate, air temperature, and solar brightness. Principal component analysis allowed us to identify that the strength of the relationships between water quality and hydrometeorology change depending on the location of the monitoring site. Finally, the implementation of the vector auto-regression model showed performance metrics lower than one for the dissolved oxygen forecast using hydrometeorological and water quality time series as input | Magíster en Hidrosistemas | Maestría | https://orcid.org/0000-0002-8756-9886 Aguirre-Duran, Cesar

The determination of space–time variation in groundwater accumulation in Colombia’s Eastern Llanos foreland basin from 2003 to 2014 was done using terrestrial water storage (TWS) anomalies identified in two versions of the Gravity Recovery and Climate Experiment (GRACE) data—from the Global Data Center for Space Research (CSR) at the University of Texas at Austin (USA) and from the Institute of Geodesy at the Graz University of Technology (ITSG, Austria)—and also soil moisture storage (SMS) data from the Global Land Data Assimilation System (GLDAS). These data were compared to changes in groundwater storage obtained using the water-budget equation, calculated based on recorded data from hydrometeorological stations. This study confirmed the viability of using satellite information to understand and monitor temporal variation in groundwater recharge in the study area. Temporal variations in TWS, SMS, and groundwater level were shown to correspond to regional rain and drought periods, which are sensitive to climate phenomena such as El Niño and La Niña. Comparing changes in TWS and groundwater level to changes in infiltration and recharge revealed correlation coefficients of 0.56 and 0.98 with CSR data and 0.71 and 0.86 with ITSG data, respectively.