Intuitive method for the calculation of water surpluses: Case study of the central area of Córdoba, Argentina

Abstract

Changes of land use in hydrographic basins can lead to serious water imbalances, resulting in floods or barrages. Soils in central Cordoba have undergone significant changes over the past five decades as a result of land conversion to agricultural, industrial, or residential uses, and other anthropic disturbances such as periodic forest fires or invasive alien species. The development of expeditious methods for estimating water surpluses would facilitate rapid data gathering to estimate the extent of hydrological risks. The objective of this article is to develop a methodology for calculating water surpluses based mainly on empirical data on infiltration capacity and storms, designed for different recurrence times. This can be applied to geographic information systems to obtain graphical information on critical areas based on water surpluses. The starting hypothesis is that the method discussed would allow detecting potential hydrological risks derived from anthropic disturbances such as forest fires, invasive alien forest species, extensive grazing, and agricultural activities in central Cordoba, Argentina. The study and methodological approach comprised three hydrographic sub-basins located in the La Calera Defense Nature Reserve (RNDLC, in Spanish), Cordoba (Argentina). We selected one sub-basin used for agriculture, one with a history of fires, and one influenced by invasive alien forest species and extensive grazing, based on land-use and vegetation maps, fire events, and areas used for agriculture within the RNDLC. Based on these maps, different soil characteristics were defined in relation to infiltration capacity. From infiltration curves for the different types of soil, land use, and alterations, instantaneous rates were calculated at 10-minute intervals over 120 minutes; these values were then assigned to raster layers. The instantaneous infiltration capacity was corrected considering the mean slope. Intensity-Duration-Frequency curves were constructed for recurrence times of 5, 10, and 25 years; these values were corrected according to tree interception and tree coverage coefficients. Maximum instantaneous precipitation (calculated at 10-minute intervals) was subtracted from effective instantaneous infiltration values to define the instantaneous excess water that might potentially drain as runoff. Effective instantaneous water surpluses were then integrated over a period of 120 minutes. In this way, water surplus rasters were derived for each land-use scenario. Excess volumes were calculated for each pixel in each runoff category. In the agriculture sub-basin, 3-year closures resulted in reductions of up to 42% in water surpluses, and 8-year closures achieved reductions approaching 80%. In the fire sub-basin, no significant changes were observed when a non-fire situation was contrasted versus 3-year-old fire events, or versus a recent fire before the rainy season, although there was a significant 1.5 to 2-fold increase in water surpluses after the first rains, possibly as a result of the impact of rain drops and sediment accumulation from runoff. In the sub-basin with invasive alien forest species and grazing, reductions of up to nearly 50% in water surpluses were recorded when the replacement of invasive alien forest species by native species was simulated in scenarios of shorter recurrence times (Tr = 5 years) while maintaining livestock activity. On the other hand, restricting grazing through closure reduced water surpluses in about 47% under high rainfall (Tr = 25 years), but caused no effect under low rainfall (Tr = 5 years). Changes of land use such as agriculture, forest fires, invasion of alien species, or cattle ranching produced basins with lower water infiltration capacity; this may lead to significant runoff volumes that could cause flooding and sediment runoff with adverse environmental, social, and economic effects. Empirical infiltration capacity data, combined with data from variables that could also be empirically obtained, such as soil coverage, are the main inputs for these models, allowing the rapid prediction of potential flood events in hydrographic basins.