Generation of spatial grids for hydrometeorological data for estimation of groundwater recharge in tropical aquifers

Climatology and hydrology depend significantly on the precise spatial characterization of their variables on continuous surfaces. The interpolation of hydrometeorological data emerges as an essential component, allowing for the estimation of values between observation points and facilitating the creation of detailed and comprehensive datasets. This process is crucial for understanding the climatic and hydrological conditions of watersheds. This work describes applying a methodology for the spatial estimation of meteorological variables through geospatial models.

The methodology includes data cleansing and validation to identify and correct outliers or errors, the assessment of model accuracy through cross-validation techniques, and a detailed analysis of the spatial and temporal variability of the data based on data availability in hydrometeorological stations.

A geostatistical analysis is conducted, adapted to the peculiarities of each measured hydrometeorological variable, considering relationships with other meteorological variables and secondary information such as altitude, latitude, longitude, and terrain aspect, using multivariable regressions. This improved the data estimation quality due to high correlations between variables.

The generation of data grids and their subsequent interpolation allow the creation of detailed maps of hydrometeorological variables in areas without monitoring stations, providing a more comprehensive and detailed view of environmental conditions. The uncertainty associated with the results is evaluated and presented to interpret the generated maps properly.

This study, conducted in a Colombian watershed, highlights the applicability of this methodology in basins with limited information. For this purpose, data and maps of temperature, evapotranspiration, and precipitation were generated at different space-time scales of interest, in addition to estimating multi-temporal potential recharge maps.