Master's degree investigates rational management of water resources in agriculture

Accurately quantifying and monitoring soil energy and water flows over large areas, in order to improve the management of water resources in agriculture, is a challenge that goes beyond the traditional approach based on point measurements of hydrological and agri-environmental parameters. To provide decision makers and other professionals with more accurate information, research combined satellite images and climate data, in spatial and temporal analyzes across different land uses, in order to model the evolution of groundwater levels in 39 wells located in a cerrado area in the Ecological Station of the city of Águas de Santa Bárbara (SP).

The study was carried out by environmental engineer César de Oliveira Ferreira Silva in his master's thesis defended in September at the Faculty of Agricultural Sciences at Universidade Estadual Paulista Júlio de Mesquita Filho (Unesp). 

Through the association of remote sensing data with geoprocessing tools, the survey considered data collected between 2014 and 2018, when there was a period of El Niño that changed the rainfall regime and generated impacts on the vegetation.

The dissertation was supervised by professor Rodrigo Lilla Manzione, from Unesp, and co-supervision was carried out by researcher José Luiz Albuquerque Filho, from Water Resources and Geoenvironmental Assessment Laboratory of the IPT, within the New Talents Program of the institution. “I realized since graduating that I would have to insert my sustainability studies in the context of agribusiness because it is the area with the greatest interface and the most openness to this issue at the moment”, says Silva.

The management of underground water resources, he explains, is strategic in public water security policies in Brazil: “For the rational use of underground water in rural supply, it is necessary to consider seasonal fluctuations in recharge and the natural behavior of aquifers, in addition to the its insertion within a territorial plan in which agrarian expansion respects the environmental characteristics of the region when granting the right to use water is requested”.

The growing pressure on water resources increasingly requires knowledge of where, when and how water is or will be used. To achieve this, it is essential to understand the behavior of parameters such as evapotranspirometric demand – evapotranspiration can be described as the sum of water evaporation from the soil surface plus plant transpiration, passing into the atmosphere in the vapor state.

Combined technologies

Remote sensing, associated with agrometeorological data, is a large-scale water resources management tool that provides essential information for rational water management. It is a tool that allows monitoring both the impacts of climate change and those caused by intensive agricultural activities on environmental conditions and the incremental water consumption of crops.

To monitor water consumption by vegetation and calculate its impact on aquifers, the environmental engineer studied a series of mathematical models in order to quantify the parameters necessary to calculate the water balance, which is the difference between the volume of incoming and outgoing water. of a basin. “When it rains, a portion of the water volume will run off, another will evaporate and a third will penetrate the soil, which has different characteristics: some allow greater infiltration while others have covers that provide greater drainage,” he says.

A vegetated slope, for example, allows little runoff and more infiltration, which results in a less dramatic response in the case of rain; on the other hand, on a slope without vegetation, a storm can even cause a landslide.

Evapotranspiration, which is part of a research universe that is still little explored in Brazil and difficult to calculate (even with the installation of sensors in the soil), constitutes one of the main variables of the hydrological cycle, being fundamental for the calculation of the water balance. “That’s why I talk about recharge in my work and the parameter inserted is the modeling of the water table – knowing how much it has recharged, I can then conclude whether there has been an increase or decrease in its level”, he adds.

The SAFER model (Simple Algorithm for Evapotranspiration Retrieving, or Simple Algorithm for Obtaining Evapotranspiration) was used by Silva to obtain evapotranspiration values ​​through an implementation built in the collaborative computational environment R. Surface runoff was modeled using the Rational Method, while aquifer recharge was simulated by the relationship of the parameters previously elaborated within the water balance. Groundwater levels were modeled on a daily scale through an adaptation of the method Water Table Flotation (WTF), which uses fluctuations in groundwater levels over time to estimate recharge in free aquifers, that is, unconfined.

“As it is a complex parameter, I had the idea of ​​building an open code so that other researchers could have a script ready for application and collaborate with new resources in the evapotranspiration modeling stage”, says Silva. “Within the R environment, it was possible, using simple commands, to 'hide' more than a hundred lines in a single line, which are the calculations necessary to insert the satellite image files and run the commands, in order to obtain the maps based on different variables. The script that needed 40 minutes to run now takes around 30 seconds.” The developed package, called agriwater, was approved and published in the official R repository.

The model was applied to images from the Sentinel-2 sensors (launched by the European Space Agency, which provides a resolution of 10 meters) and Modis (from NASA), generated between 2014 and 2018, to calculate the aquifer recharge potential and model the variation in the depth of the water table within the basins studied. The modeling was validated with data measured in the 39 wells and indicated the feasibility of using satellite images to monitor impacts on aquifers.

The results indicated that the model's uncertainty is sufficiently understood for decision-making in public policies. The conclusions may support the rational management of water resources in areas of agricultural development in scenarios of climate change and land use, in order to guarantee the best use of rainwater and minimize deficiencies and waste of irrigation water.

This information, adds Silva, is essential for water management, especially in places economically dependent on irrigated agricultural production: “The use of orbital remote sensing techniques to conduct hydrology mapping makes spatial diagnosis possible quickly and cheaply, and provides tools for integrated planning between environmental, agricultural and social resources at different scales”.

Three scientific articles have already been published as a result of the research: the first of them in the Revista Horticulturae, from the Swiss publisher MDPI, the second in Environmental Earth Sciences, from the German Springer and the third in Environmental Modeling & Software, from the Dutch Elsevier.