RESUMO
This study used the geostatistical Kriging methodology to reduce the spatial scale of a host of daily meteorological variables in the Department of Cauca (Colombia), namely, total precipitation and maximum, minimum, and average temperature. The objective was to supply a high-resolution database from 01/01/2015 to 31/12/2021 in order to support the climate component in a project led by the National Institute of Health (INS) named "Spatial Stratification of dengue based on the identification of risk factors: a pilot study in the Department of Cauca". The scaling process was applied to available databases from satellite information and reanalysis sources, specifically, CHIRPS (Climate Hazards Group InfraRed Precipitation with Station Data), ERA5-Land (European Centre for Medium-Range Weather Forecasts), and MSWX (Multi-Source Weather). The 0.1° resolution offered by both the MSWX and ERA5-Land databases and the 0.05° resolution found in CHIRPS, was successfully reduced to a scale of 0.01° across all variables. Statistical metrics such as Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Person Correlation Coefficient (r), and Mean Bias Error (MBE) were used to select the database that best estimated each variable. As a result, it was determined that the scaled ERA5-Land database yielded the best performance for precipitation and minimum daily temperature. On the other hand, the scaled MSWX database showed the best behavior for the other two variables of maximum temperature and daily average temperature. Additionally, using the scaled meteorological databases improved the performance of the regression models implemented by the INS for constructing a dengue early warning system.
RESUMO
Quantifying distributed lateral groundwater contributions to surface water (GW-SW discharges) is a key aspect of tracking nonpoint-source pollution (NPSP) within a watershed. In this study, we characterized distributed GW-SW discharges and associated salt loading using elevated GW specific conductance (SC) as a tracer along a 38 km reach of the Lower Merced River in Central California. High resolution longitudinal surveys for multiple flows (1.3-150 m(3) s(-1)) revealed river SC gradients that mainly decreased with increasing flow, suggesting a dilution effect and/or reduced GW-SW discharges due to hydraulic gradient reductions. However, exceptions occurred (gradients increasing with increasing flow), pointing to complex spatiotemporal influences on GW-SW dynamics. The surveys revealed detailed variability in salinity gradients, from which we estimated distributed GW-SW discharge and salt loading using a simple mixing model. Modeled cumulative GW discharges for two surveys unaffected by ungauged SW discharges were comparable in magnitude to differential gauging-based discharge estimates and prior GW-SW studies along the same river reach. Ungauged lateral inlets and sparse GW data limited the study, and argue for enhancing monitoring efforts. Our approach provides a rapid and economical method for characterizing NPSP for gaining rivers in the context of integrated watershed modeling and management.
